A rare but not unprecedented Saturday edition of Top Three…
If you know somebody else who might be interested in this, please forward it on.

The Sport Pyramid

We play it. We watch it. We coach it. We fund it. We talk about it endlessly. As a topic for polite conversation it’s right up there with the weather. It even has its own entire section in the newspaper, with journalists dedicated to the coverage.

But despite touching on so many aspects of our lives, I’ve found we don’t always have a great shared mental model for how to think about it.

Here’s mine…

This pyramid has two dimensions. The height represents performance - the extent to which competition and winning is the point. The width represents participation - a spectrum from activities that almost everybody can do, through to things that are limited to an exclusive group.

Every sport, every athlete, every activity, sits somewhere on this pyramid. And I’ve found this model to be a surprisingly useful tool for making sense of questions that otherwise seem intractable - about money, about governance, about who gets to compete (and even whether competition should be the focus), and who gets to decide.

The Performance Dimension

The first dimension measures the height of the pyramid and tracks performance. From elite, high-performance sport at the top to regular everyday activities at the bottom. From Olympic champions and professional athletes making their living from competing at the pointy end, to normal people simply living an active life.

The Olympic creed, attributed to Pierre de Coubertin, founder of the modern Games, is:1

The important thing is not the winning but the taking part.

It’s a noble sentiment. Perhaps it was true in 1908. But these days the Olympics are first and foremost about performance: winning medals, and specifically gold ones. The creed describes an ideal. The medal table tells a different story.

What does “performance” actually mean?

It turns out, the answer depends entirely on where you sit on the pyramid.

Imagine taking your kids out for a bike ride this long weekend. The purpose is to be active, take advantage of being outdoors, and to have fun together. Unless you are a particular kind of Strava tragic, which of you gets there first and how fast you go is broadly irrelevant.

Meanwhile, if you’re a cyclist competing on the velodrome at an Olympic final, speed is of the essence. Times are measured to thousandths of a second. Tiny margins separate glory from failure.

In between those two extremes lies a vast and complicated middle.

At the top of the pyramid, where winning is the measure of success, sport is a zero-sum game. Second place is the first loser. And, as in any competition, there are many more losers. Those who succeed tend to have a singular focus. They are often ruthless. There is even sometimes a blurred line between honest competition and unfair advantage.

As we get closer to the bottom of the pyramid the motivations change. We’re challenging ourselves rather than comparing ourselves to others. Sometimes it’s about completing than competing, and the satisfying rush of dopamine we get at the finish. Other times it’s simply about having fun and doing something together with friends or family.

At what point does competing and winning become the more important thing?

The disconnect occurs when the people organising the events or running the activities think and behave as if it’s only the winning that counts.

One of the gnarly questions in sport is how and where we draw the line that defines who competes against whom, and what is “fair” competition.

Most, but not all sports, are split into men’s and women’s events. Equestrian events are an exception - where men and women compete in a single competition - perhaps highlighting the contribution of the horse.

Some, but not all sports, are further split into weight divisions. These categories in combat sports acknowledge that body size creates fundamental competitive advantage. It seems unfair to ask a lightweight to fight a heavyweight. Meanwhile, as far as I know, Chris Solinsky is still the only person to have run 10,000m in under 27 minutes while weighing more than 65kg (he ran 26:59.60 in 2010 while weighing a remarkable 74kg). Despite that, athletics offers a single event, independent of body weight.

All of these categories are somewhat arbitrary. Every one of them is a line drawn by administrators across what is actually a continuous human spectrum. The current debate around trans women competing in female events is perhaps the most visible and contested version of that line-drawing. It won’t be the last.

The pyramid doesn’t answer these questions. But it does help frame them. The higher up the pyramid you are, the more these distinctions matter, and the more contested they become.

The Participation Dimension

The second dimension measures the width of the pyramid and tracks participation from activities that almost everybody can do at the base to those that only a small number of people ever have the opportunity to participate in (sometimes intentionally, by design, sometimes unintentionally, due to lack of design).

Rugby in New Zealand is a useful illustration. At the pinnacle sit the All Blacks & Black Ferns - 15 players start on match day, 23 in the squad, maybe 40-50 selected in a touring group. Below that, four (women) or five (men) Super Rugby franchises. Below that, 26 provincial unions competing in the National Provincial Championships & Heartland Championships (men) or Farah Palmer Cup (women). Below that, clubs around the country, each with their own internal hierarchy - premier grades, and lower grades below that. And feeding into all of it, school rugby (although the line between schools and clubs has become increasingly blurry in recent years).

The pyramid highlights why the health of the base matters enormously. Despite some small increases recently, the long term trend is a decline in participation. Between 2011 and 2019 the number of teenagers playing rugby in NZ decreased by 4,000 - meaning about 180 fewer teams.2 Some clubs have responded by waiving subscription fees for junior players, in an attempt to “revive grassroots participation”.3 This is not just an immediate problem for those clubs. It’s a slow leak that will eventually show up as a talent shortage at all levels.

The base of the pyramid isn’t just wide, it’s also temporary. Active kids are significantly more likely to become active adults. And the inverse is also true. Investment at that level compounds over a lifetime. But children's sport is also where the elite mindset causes the most damage. Early specialisation, selection pressure, and a win-at-all-costs coaching approach applied in primary school are not edge cases. They are a direct consequence of people from the top of the pyramid setting the culture for the bottom of it: optimising for the wrong thing, for the wrong people, at the worst possible time.

An even more extreme example is the Olympics. Ten thousand athletes compete, from a global population of eight billion. They are all freaks, by definition. Consider Michael Phelps, who is the most decorated Olympian of all time, with 28 medals across four Games. He competed in a sport that offers many opportunities to win medals, but he also arrived with a few structural advantages: the ideal swimmer’s height at 1.93m (6’4” in old numbers), an arm span that exceeds his height, legs proportionally short enough to reduce drag, and size 14 feet that hyper-extend 15 degrees beyond average - effectively flippers. None of that diminishes the years of dedicated training he put in. But as they say, Olympic gold begins with good genes.

When we are thinking in terms of performance the pyramid is explicitly and necessarily exclusionary. That’s not a flaw in the system. It’s by design. Selection is extremely hard earned and at the same time inherently unfair.

But when we are thinking in terms of participation, the opposite should be true. In the lower parts of the pyramid the goal should be to create opportunities for as many people as possible.

What actually limits participation in any given activity?

• Equipment - the things you play with - balls and rackets, boots and mouthguards

• Facilities - the places you play, and how open they are (e.g. no lights = only available during the day)

• Other People - teammates, opponents, coaches (team sports in particular can’t happen alone)

• Cost - entry fees, membership, travel (see below)

• Geography & Climate - e.g. mountains for skiing, coastline for surfing

• Physical Requirements - minimum fitness, size, or ability to participate meaningfully or safely

• Time - availability, scheduling, life stage

• Skill Threshold - many activities require a base skill level before they’re accessible or enjoyable

• Social & Cultural Factors - gender norms, class associations, community traditions

• Safety & Risk - real or perceived danger

• Disability - physical or cognitive barriers, lack of adapted versions

Many of these compound each other. For example, cost and geography together can make an activity effectively inaccessible for entire communities, even if neither alone would be insurmountable.

For a long time, sports clubs served as a centre of community life, filling a role that the church held in earlier generations. A place of belonging, identity, and shared purpose. That model is under pressure in ways that feel structural and probably irreversible. People are less able to commit to club membership or contribute to the administration that clubs depend on to function. Many of us tend to spread ourselves across multiple activities now, rather than going deep into one club or one code.

For example, I enjoy the occasional round of golf with friends. But committing to four or five hours every weekend is more difficult. And the idea of belonging to a golf club and contributing to its administration the way my father’s generation did feels like a different social contract entirely. The sport is the same. The world around it has changed.

At the base of the pyramid, participation is a human right. The further up the pyramid, the more exclusion becomes not just acceptable but necessary. The dilemma is deciding where on that spectrum to draw the line, and being honest about the values embedded in that decision.

The Business Model

The economics of sport and recreation impacts every part of the pyramid. But the practical realities at different points on those dimensions are so different that it can be hard to appreciate they’re part of the same system.

Let’s start with an often overlooked question: what is sport and recreation actually worth? It’s easy to get distracted by things like broadcast rights, ticket sales, sponsorship revenue, and player salaries. But that framing misses most of the value. Broad participation at the base of the pyramid - recreational runners or cyclists, weekend club players, kids at swimming lessons, etc - generate enormous collective value in the form of health outcomes (including reduced shared healthcare costs), community cohesion, and individual wellbeing. That value is real, but it doesn’t show up on any balance sheet. It accrues to all of us collectively, which is a large part of the justification for public investment in sport and recreation.

To get the full picture we need to include a few more uncomfortable realities.

Elite sport is also entertainment. This is where the pyramid’s shape becomes economically important. At the top, a very small number of people are in the arena competing while most of us sit on our behinds and watch - either in the stadium or on the couch at home - and pay for the privilege. At the base, it’s the opposite: a large number of people participate and relatively few supporting. These are fundamentally different economic models. One sells the spectacle; the other sells the experience of participation itself.

When we make sport a product that is sold it creates a tension between fans and broadcasters, sponsors and private owners. For example, when kick-off times are moved to suit the television audience, ignoring the experience of the fans who show up to watch live at the stadium. Or, rule changes to make it easier for broadcasters to inject advertisements into their coverage. While each of these can feel like small compromises in the name of revenue, the net impact in the longer term is empty stadiums and dwindling fan bases.

The commercial landscape of sport is not a bell curve, it’s a power law. A small number of sports have genuinely lucrative business models, with valuable broadcast rights, willing sponsors and economics that are self-sustaining or even profitable. But the drop-off is steep and fast. Even popular well-established sports such can fall off the edge of commercial viability surprisingly quickly, and find themselves scrambling to fund their own coverage rather than being paid for it, taking whatever sponsorship is available, and often settling for much smaller deals or going without entirely.

Most sports now operate in an environment of scarcity, not abundance. The mental model that people carry from watching elite professional sport, where the amount of money involved can be eye watering, is wildly unrepresentative of the reality for the overwhelming majority of sports administrators.

Which naturally leads to the vital question: who pays?

Sport’s relationship with funders deserves more honest scrutiny. The uncomfortable truth is that the companies willing to pay the most to attach their brand to popular sports are disproportionately those with the largest negative social costs.

As a kid in the early 1990s I remember watching the Benson & Hedges Cricket World Cup hosted in New Zealand and Australia. In that era, tobacco companies were prominent sponsors of sport. While that was eventually banned, it just resulted in those sponsors being replaced by the next-worse alternatives, including alcohol brands and fast food companies. These days sports gambling brands are the most prominent sponsors across many major codes.

At the top of the pyramid sport has historically been financially dependent on industries that undermine the very health and wellbeing values that sport is supposed to embody.

At the far end of the commercial spectrum, sits private ownership of sports teams and competitions, and private equity investment in national sport organisations. This isn’t new. Sports like football have dealt with this for a long time. But it’s spreading rapidly.

Cricket is a recent example. The Indian Premier League pioneered franchise ownership, and that model is now being replicated across all major cricket-playing nations. These are mostly vanity assets for owners, rather than straightforward investments. They usually don’t generate returns directly. Where owners do profit, it’s typically because they can shape the environment around the asset to benefit themselves through related interests. In some cases mechanisms like salary caps, player drafts, and revenue sharing agreements exist specifically to constrain what private owners can do and protect competitive integrity and player welfare. But the athletes whose talent generate all the value tend to have the least structural power in the system.

The funding picture for community sport, at the base of the pyramid, is even more complicated.

In New Zealand community sport and recreation is dependent on funding that flows from the regulation of local poker machines (pokies), lotteries and sports betting. This creates a strange dynamic where administrators of community sport organisations find themselves quietly hoping for big Lotto Powerball jackpots. It’s hard to interrogate the ethics of a funding source you depend on to keep the lights on.

This model is also under structural pressure. As sports betting moves online and offshore, the domestic regulated share of that revenue is shrinking, and there is no obvious replacement.

The central irony of the economic layer of the pyramid is this: the least commercially powerful sports generate the most social good. The weekend club hockey player, the community swimmer, the recreational jogger together produce health outcomes and community value that dwarf anything generated by a broadcast rights deal. Indoor netball and roller derby have tiny or near-invisible elite tiers, yet broad participation and real community value. And yet these are precisely the people who are most often asked to put their hand in their own pocket just to participate.

Who Decides?

Who governs sport? Who sets the rules? Who determines priorities? And are they the right people to be doing it?

The answer, across almost every level of sport administration, follows a remarkably consistent pattern. Those who end up making decisions are drawn almost exclusively from the pool of former athletes and participants in that sport. This is perhaps a natural transition. Who better to govern a sport than those who have lived it and are closest to it?

The problem is that sport and governance and administration are quite different skills. They don’t automatically go together. For example, the former club rugby player who ends up on the board of a provincial union often arrived there by simply being around long enough, not via any particular capability in administration, strategy, or organisational leadership.

At the extreme end, the International Olympic Committee is composed almost entirely of former Olympians. If you subscribe to my view that all Olympians are freaks, extraordinary outliers in human physical capability and competitive drive, then an organisation governed by Olympians will inevitably be biased by that experience. Their lived reality of what sport is, what it demands, and what it’s for is radically unrepresentative of what sport means to the vast majority of people.

The result is a structural representation gap. Those making decisions are typically older, no longer active participants, and drawn from a narrow demographic band. It’s a constant challenge to understand what’s important to current participants - the people the whole system is supposed to serve.

In board settings like these I am often told: “you think so differently”. I think that’s intended as a genuine compliment. But it inadvertently exposes the very problem it thinks it's solving. If my perspective stands out, it's because everyone else's doesn't. Most people at the table look broadly the same, think broadly the same, and have broadly the same relationship with sport.

The pyramid compounds all of this. A single governance structure attempting to serve the full depth, from elite performance to grassroots participation, is asked to do something that may simply be beyond what any single organisation can do well. The needs, priorities, culture, and economics at the elite high-performance end are almost completely disconnected from those at the participation base.

Consider cycling, as an example. The needs of athletes preparing for elite competition share almost nothing with the reality of people riding bikes for recreation, commuting, or weekend family adventures. Not even the bikes are the same. The bespoke machines that elite cyclists ride and the bikes that you and I ride are themselves almost unrecognisably different. Different motivations, different infrastructure needs, different relationships with competition, different economic models, different governance requirements. The idea that a single group can authentically represent and serve both is almost structurally implausible. And yet that’s what we ask all national sporting organisations to do.

The gravity in sport administration pulls upward. Toward the elite, the measurable, the visible, the fundable. The people who end up in governance are disproportionately drawn from that top of the pyramid. The money, the prestige, and the attention all concentrate there too. And so the base - where most of the people are, where most of the participation happens, and where most of the collective health and community value is actually generated - tends to be an afterthought in the rooms where decisions get made.

What would administration that genuinely serves the full pyramid look like? It would need to deliberately recruit governance capability rather than simply rewarding participation history. It would need to create authentic mechanisms for the voice of current participants to be heard (not just former ones). It would need to be honest about the spanning problem and consider whether federated or separate structures might serve different parts of the pyramid better than a single body trying to do everything. And it would need people in the room who think differently - not as a novelty, but as a design principle.

The Triumph & The Struggle

This sports pyramid is a simple model. Two dimensions, one picture. But models like this can do useful work. And this one, I think, helps clarify something that a lot of well-intentioned people in sport often talk past each other about.

The pyramid contains multitudes. It contains the Olympic sprinter who has dedicated their entire life to shaving hundredths of a second off a time that almost no human being in history has approached. It also contains the person who dug their running shoes out of the cupboard for the first time in a while this morning and made it to the end of the street. Both of those people are part of the same system. Both of them deserve consideration. But they need entirely different things from that system: different infrastructure, different economic models, different governance, different definitions of success.

The word “inclusion” gets used a lot, without acknowledging that at the top of the pyramid, exclusion is not a bug. It’s by design. It’s what makes elite sport meaningful. It’s a trade-off between the triumph of competition and winning on one hand, and the joy and struggle of participation and being active on the other. That tension sits at the heart of almost every hard problem in sport administration. And yet the people most often asked to pay to participate are the ones generating the most social value while receiving the least institutional support.

A better mental model doesn’t resolve any of these debates. But it might at least help us argue about the right things. Most of the intractable disagreements come down to the same thing: people pointing at different parts of the pyramid while using the same word. The professional athlete and the weekend club player both call it sport.

Let’s stop pretending the pyramid is flat.

#Winning

Speaking of winning…

There are lots of different ways to “win”. When we’re running a business there is a universal and easily understood scoring system (ref: capitalism): Create something people want, charge slightly more than it costs to make, and if enough people agree with our assessment of its value, we get to keep playing. In the best case we build something sustainable, hire interesting people that we enjoy working with, and maybe eventually sell the whole thing for more than it cost to create.

And yet. If a tree grows in the woods and nobody sees it, does it make a sound?

In Hollywood awards season starts in November and culminates with the Academy Awards in late February or early March. In New Zealand the business awards season runs continuously, all year, like a slow-moving conveyor belt of gala dinners. When it comes to being celebrated, we really are spoilt for choice!

My personal favourite is the Fast 50 Awards, sponsored by Deloitte, because it’s objective, unlike all of the others where the winners are selected, as in a beauty pageant, by a panel of judges. It’s not perfect. The revenue growth measure they use can be distorted or manipulated, and they too have fallen into the trap of adding subjective categories over the years, but it’s still the best of the bunch.

Trade Me, Xero, Vend and Timely are all past winners. I was fortunate to be part of all of those teams. Those awards were wonderful recognition for the hard work that many people put into achieving those results, and we celebrated every one.

But they aren’t free. The application and judging process soaks up hours that would be better spent talking with potential customers and growing the business. And then when it comes to the gala dinner, where the winners are announced, those tables are not cheap.

What’s the return on investment? Personal recognition? Team building? Perhaps it makes it easier to recruit people to join the team?

Awards are a great opportunity for organisers and sponsors (typically large accounting firms or government departments, or more recently venture funds) to attach themselves to more innovative and exciting businesses, and get media coverage they couldn’t otherwise buy.

They’ve done the maths. We should all do the same.

Good luck to all who have taken the time to enter this year. May the best team (who are most attractive to the current panel of judges) win.

How To Be Wrong @ School

I have some free copies of my book How To Be Wrong available for school libraries.

If you would like to get one for your school please add the details to this form:

Request A Free Copy For Your School

Please share these details - I’m keen to get these out to as many schools as possible. If you’re a parent, tell your kids. If you’re a student, tell your school librarian. If you’re a librarian, click the button above!

Meanwhile, the first hardback copies will ship in the next couple of weeks, but there is still time to pre-order a copy at the special discounted launch price. Don’t delay…

Hard cover. Harder truths.

To mark the one year anniversary of the publication of my book How To Be Wrong, I’m excited to announce it’s now available for the first time in hardback.

The first copies will ship in April. Pre-sales are available today, at a special introductory price.

Pre-order Now

The book is also still available in paperback, as an ebook or audiobook.

Find a full list of the places where it’s sold on the official How To Be Wrong website.

Enjoy!

How To Be Wrong is the book New Zealand’s business community has needed for years, from the person best qualified to write it. Not another cheerleader for the innovation economy. Not another lone-genius myth. A rigorous, honest, and often uncomfortable account of what actually drives success, and why so much of what we tell ourselves about startups is, to put it plainly, wrong.

Good Vibes Only

vibe coding (n.) asking an AI agent to write software code that you cannot read or understand.

See also: confidence, misplaced; abstraction, comfortable.

There are currently two opinions competing for dominance.

First, AI is a massive bubble that will soon pop. Large language models (LLMs) are impressive but shallow. The economics don’t work, and the environmental cost too large. We’re at the high point in the hype cycle. Soon we will file openai.com just ahead of pets.com in the alphabet of expensive lessons.

Second, software engineering as a profession is finished. Now, anybody can just describe the software they want or need and AI will build it for them. Learning to code? Irrelevant! A motivated twelve-year-old with a Claude Code subscription can do that job.

(As an aside: it is curious and revealing how many people seem to hold both of these opposing opinions concurrently).

I don’t think either is right. But the contradiction does highlight that we don’t yet have a clear mental model for what these new tools actually are, what they are good at, and who will benefit the most from using them. Vibe coding hogs the headlines, but there is something much more interesting bubbling just below the fold…

1. We mostly never read the output

Writing software has always involved layers of abstraction.

When I was studying Computer Science at university, back in the late nineteen hundreds, one course required us to learn machine code. These are the instructions executed directly by computer chips. And it was painful. Even the trivial examples we were given - adding two numbers together or storing and retrieving a single binary value from memory - required lots of code and hours of effort.

That is a skill I’ve long since forgotten, and never really used in anger. Most software engineers today never need to think about the machine code, and would likely struggle to understand it if forced. We use compilers to take the source code we write as input and convert that into machine code for us. We trust their output.

More recently, AI has added a new layer of abstraction. When we use a tool like Claude Code or Cursor the inputs are no longer source code, they are English prompts. The tools generate the source code for us based on these prompts. We’re one layer further up the stack.

When we’re vibe coding there is no stack. It’s just a black box. The only way we have to verify the output is at the application or user level. That leaves us exposed to a whole class of bugs and security concerns which are difficult to spot if we don’t know what we’re looking for.

The opportunity for software engineers is to always have the full stack in mind, and be intentional about how we verify the output that is generated at each layer. In this next era the job is much less about curly brackets and semi-colons and much more about quality assurance.

2. What’s the baseline?

When we worry about the quality of AI-generated code, we’re implicitly comparing it to an idealised standard of craftsmanship. Code that has been written carefully by skilled engineers, and reviewed throughly to ensure it contains no bugs.

That’s an imagined and unrealistic standard.

The actual alternative is code that is often written in a rush, reviewed by humans who are tired and distracted (because, honestly, who loves reviewing code?!), making assumptions that seemed reasonable at the time, resulting in applications that ship riddled with bugs nobody found because the test suite didn’t cover every edge case.

I speak from experience. I’m already humbled. Agents have improved code I wrote carefully. In one case, it found some performance improvements which generated so many positive comments from users that I felt embarrassed taking the credit. In another, it completed a poorly constructed test suite and in the process found a number of bugs that had been sitting there quietly accumulating cost.

This doesn’t mean AI code is always perfect. Or even good. It means the benchmark is lower than we’re pretending, and the systematic verification that good tooling enables might actually be more rigorous than the human review process it’s replacing.

Understanding persistence is important. If we’re a business analyst asking AI to quickly build a dashboard to summarise a dataset, so that can be pasted into a presentation, the quality requirements of the underlying code are low. If we’re a team of engineers building an application with thousands or even hundreds of thousands of users, and complex features, that needs to be maintained for years to come, that’s significantly different. Conflating the two is a mistake.

3. The real world is messy too

The sharpest criticism of AI-generated code is that it’s probabilistic. We can’t formally verify it. How can we trust software that doesn’t behave the same way twice?

This is absolutely fair. But, again, it’s worth examining the embedded assumption that what came before was deterministic.

In practice, as soon as we add humans in the equation formal proof becomes difficult if not impossible. People are fallible. We do unexpected things sometimes. And I’m not just talking about “users”. Engineers don’t always read requirements. Testers don’t always follow the script. A lot like AI agents, as it turns out!

Even compilers, which are meant to be the reliable mathematical layer of the stack sometimes have their own challenges when it comes to consistent output.1

As Computer Scientist C.A.R Hoare said:

“There are two ways of constructing software: one way is to make it so simple that there are obviously no deficiencies, and the other way is to make it so complicated that there are no obvious deficiencies. The first method is far more difficult.”

Determinism has always been an aspiration, not a guarantee. The question isn’t whether AI adds uncertainty to the stack. Of course, like any new layer, it does! We should ask whether the uncertainty it adds is meaningfully worse than the uncertainty already there. I’m not sure it is.

The engineers who understand that distinction will build better software than the ones who are simply anxious about it. AI tools are amplifiers. They give engineers more leverage, not less. A 10x engineer with a great AI workflow is a 100x engineer. They can go further, faster and with more confidence than before.

Even if you have all the vibes, it’s hard to compete with that.

Header Photo by Firmbee.com on Unsplash

When was good?

It seems that nearly everybody agrees, at this moment in history, that things are currently not good.

So here’s a question I’ve been asking people. Let me expand the sample group…

When was good?

Your answer might reveal more about your politics than you realise.

Maybe you think back fondly to 2019? Pre-pandemic, when the mood was much more “let’s do this” than “the tank is empty”.

Or, maybe you want to go back to 2016. That seems to be a popular answer, for younger people especially.

Why not? TikTok was the sound of a clock. Donald Trump was a reality television star. The UK was still part of the EU and NZ had “a rockstar economy”, fuelled by high house prices and immigration, and overseen by a chilled out entertainer.1 That is, as best I can tell, the track we want to get back on.

What about rewinding to the early 90s? Curiously, we’re often most nostalgic for our teenage years - coinciding with what researchers at the University of Leeds dubbed “the emergence of a stable and enduring self.”2 For me that was 1989 to 1995. The time of Windows 3.1. Friends. Jordan and Jonah. But seriously… Nevermind. The only famous “Kardashian” back then was OJ’s lawyer.

That also overlaps almost exactly with the political era documented by Toby Manhire in the second season of his podcast Juggernaut. By tracking Spotify data researchers can show that most people stop actively seeking new music genres after they are in their 20s. Is the same true of political ideas or do those ossify later?

Perhaps you look even further back, to the 1950s or 1960s when men were men, nothing of consequence was Made in China, the USA and USSR were the goodies and baddies respectively (which was which depended on your context and perspectives).

When politicians overseas talk about making things “great again” this is roughly the era they all seem to be referencing. Maybe the teenage-years phenomenon explains this too?

With the notable exception of Barack Obama (born 1961), US Presidents have been getting one year older each year my whole adult life. Bill Clinton, first elected in 1992, was born in 1946. He was followed by George W Bush, also born in 1946. Donald Trump was also born in 1946. Joe Biden is four years older than those three, born in 1942. I’m not sure history will judge kindly baby boomers who continue to elect their contemporaries well into their 70s and even 80s.

And, before we get too smug in NZ, perhaps worth noting that Winston Peters - who based on current polling is likely to once again hold the balance of power after our election later this year - was born in 1945.

Others seem to look back fondly even further than that - to 1840, when by some accounts NZ was an unspoiled Pacific paradise (although the days were shorter), or even to 1760, before the Industrial Revolution saw people around the world start to use fossil fuels to manufacture everything we wanted to make our lives better.

So let’s hear your answer: when was good?

Progress & Decline

Cleaning up some old posts on my personal website recently I came across this post:

⌘-SHIFT-⌥-ESCAPE, 4th January 2016

Between 2008 and 2015 I posted these “annual reviews” at the end of each year. This one specifically happened to be the last thing I ever posted to that blog.

Reading back, it feels like opening a time capsule. I was still in my 30s (just). Our kids were … still kids. I wrote optimistically, and with the benefit of hindsight accurately, about Vend and Timely as high growth companies with potential. I didn’t realise it at the time but I was a few months away from starting arguably the best year of my life3 - one which taught me how malleable time actually is when you’re intentional.

Either way, it’s confronting to be hit with a whole decade of progress and decline at once.

When I shared the link with an old friend they made this observation:

… one of the trends this illustrates is that the data-driven self (you know, like tools that visualize your annual travel history) is less prevalent than it was.

For my sins I’m still pretty quantified even now. I’m just not sure that the narrative the numbers stitch together is as flattering these days.

Matt Webb (whose long running blog Interconnected is one of the few I’ve followed throughout that decade and continue to be delighted by) wrote an excellent piece last year called “Strava when you’re not as quick as you used to be” where he reflected on “personal bests”:

As an on-and-off-again runner who is motivated each time round by getting better at it, slowing down is something I’ve had to find peace with. I’ve managed to do that now, it took a year or two, and finding a new internal place of motivation has meant that my training is once again enjoyable and free.

Strava is one of the few “social networks” I regularly use. But it constantly reminds me I’ll likely never again run 5km as fast as I did in 2017. Every time I revisit an old favourite segment the delta between what previously seemed effortless and now seems remarkable grows wider.4

May I humbly suggest to the (I assume much-younger-than-me) product managers they put more emphasis on “recent best” as a more motivating metric.

It’s easy to spiral on the negative. I’ve referenced Brian Eno’s “feeling that things are inevitably going to get worse” many times over the years:

More than a feeling, January 21, 2009

What if it comes to feel like there isn’t a long term—or not one to look forward to? What if, instead of feeling that we are standing at the edge of a wild new continent full of promise and hazard, we start to feel that we’re on an overcrowded lifeboat in hostile waters, fighting to stay on board, prepared to kill for the last scraps of food and water?

From this vantage point, 17 years later, his prediction from 2009 almost feel prophetic:

Humans fragment into tighter, more selfish bands. Big institutions, because they operate on longer time-scales and require structures of social trust, don’t cohere. There isn’t time for them. Long term projects are abandoned—their payoffs are too remote. Global projects are abandoned—not enough trust to make them work. Resources that are already scarce will be rapidly exhausted as everybody tries to grab the last precious bits. Any kind of social or global mobility is seen as a threat and harshly resisted. Freeloaders and brigands and pirates and cheats will take control. Survivalism rules. Might will be right.

At an individual level I think that idea about trying to grab precious scarce resources while you still can also explains mid life crises. You only get to be each age once, but it’s hard to be the age you are.

So how do you counter this, and foster a sense that things could possibly get better?

Here are three ideas, for starters:

1. Be honest about the past and how good it actually was (see above). Nostalgia is a highlight reel with the lowlights edited out. If you wouldn’t go back then maybe now isn’t quite as bad as it feels.

2. Pay more attention to progress in the present. We are ultimately what we consume. That’s true mentally as well as physically. Mostly our news diet is the equivalent of eating deep fried slop for every meal. As Hans Rosling went to great lengths to highlight while he was still alive, on almost every measure things have never been better.5 Or, Information is Beautiful’s annual summary of Most Beautiful News of the Year - with all the headlines you probably didn’t see. Let’s not blame journalists for this. We can’t expect them to rise much above the level of the general population’s appetite for news fuelled by fear, greed, and an ignorance that often manifests as suspicion of expertise. The algorithm doesn’t reward “things are gradually improving.” It rewards “everything is on fire.” Somehow, within that general overall upward trend we all have to deal with our own individual decline. But don’t forget the trend.

3. Choose leaders who focus less on their individual achievements and more on their contributions to others. To quote Adam Grant from his recent NY Times Opinion piece: “The responsibility of leadership is too important to entrust to arrogant people. Narcissistic leaders deny their weaknesses and make themselves weaker. Humble leaders admit their weaknesses and make themselves stronger. Great leaders overcome their weaknesses and make us all better.” His prescription, referencing research published in American Psychologist6 is maybe not what you'd expect: start by looking for ways to increase our own self-esteem, especially in childhood: “The lower our opinions of ourselves, the more insecure we’re feeling, the higher our opinion of narcissists.”

Hopefully some of you have even better suggestions. I’d love to hear them.

Get in behind

Speaking of nostalgia…

I remain in constant disbelief that TVNZ have not yet revived A Dog’s Show, in Drive To Survive style. I reckon it would be a ratings sensation.

Buy my book

To finish 2025, I’m revisiting some of my favourite Top Three posts from the last year.

In January, just a few days before we needed to lock down the manuscript for How To Be Wrong, the government announced that Callaghan Innovation would be disestablished. The third section of the book, about the “technology ecosystem”, starts with a chapter looking back at the 2011 lecture delivered by Sir Paul Callaghan, which was perhaps the genesis for the crown entity established a couple of years later. I nervously searched for potential last minute edits, but was relived to discover only a small handful of references. In the end most were dealt with by the addition of this footnote:

Time will tell whether this amounts to more than a reshuffling of the deck chairs.

To complete this retrospective, here are three posts (actually two posts and one video) from January, March and October, looking back at the legacy of Callaghan - both Sir Paul and the Innovation agency named in his honour - and forward to what might come next…

This week only How To Be Wrong is available 25% off, direct from Electric Fence. You’ll wait a long time to get a better price than that. Enjoy!

Buy Now

🧨 Reset

This was first published in The Spinoff in January.

In 2011 the physicist Sir Paul Callaghan, one of New Zealand’s most distinguished scientists and a passionate advocate for economic transformation, gave a landmark speech he called “Sustainable economic growth for New Zealand: An optimistic myth-busting approach”. Many of his observations have become often-repeated mantras of New Zealand’s technology sector, including “be the place where talent wants to live”, “we better be prepared to be good at some pretty weird stuff” (anticipating correctly that the most successful companies would operate in obscure and difficult-to-predict niches) and that “just 100 inspired entrepreneurs could turn this country around”.

Tragically, Callaghan died of cancer just months later. However, his legacy lived on in a new Crown entity called Callaghan Innovation, established in 2013 with the vague responsibility for “making New Zealand business more innovative” through grants, technical services and business support. The intention was noble – to help transform New Zealand from a country dependent on primary industries to one driven by technology and innovation.

Fast forward to last week: prime minister Christopher Luxon’s state of the nation speech echoed many of those same ideas. At the end, he announced that Callaghan Innovation would be disestablished, with its “most important functions” reassigned elsewhere in the public sector.

For those of us who have built and invested in successful technology companies in New Zealand, this decision is overdue. Despite good intentions, the agency became bogged down in administering grants and subsidies rather than driving real innovation. Under pressure to demonstrate value, it spread its resources thinly across a huge range of activities – from managing R&D tax credits to funding startup accelerator programmes. But there was never any clear evidence that any of these initiatives actually helped create more successful companies.

There’s no question that improving our productivity and growing the economy beyond agriculture and tourism is key to our future prosperity. We’ve understood this for a long time. However, the ambition to be more innovative has never been our problem. The challenge we’ve struggled with all these years is actually doing it in any measurable way. Perhaps we missed a trick? Rather than Callaghan Innovation we should have called it Callaghan Execution.

Having worked with and invested in companies like Trade Me, Xero, Vend and Timely over the past two decades, I’ve seen firsthand what drives growth. The key question is always: What’s the constraint? What’s actually holding us back? As we consider the legacy of Callaghan Innovation, and more importantly think about what will replace it, let’s ask the same question of our economy as a whole.

Here’s an obvious answer: our national obsession with investing in real estate, rather than investing in much more uncertain but ultimately more rewarding productive businesses. Spending more and more borrowed money to buy little bits of our country off each other is a zero-sum game. We might think it’s how we “get ahead”, but it’s actually dragging us down.

Luxon himself exemplifies this – he’s often described as a “businessman”, but his investments after achieving financial success were rental properties, sold for profit years later. What kind of business is that? What value does it create?

Here’s another: our expectation that the government will underwrite business risk, while private shareholders capture the rewards. Founders and executives at early-stage companies spend countless hours pursuing grants, subsidies, R&D credits and development loans – many of which end up written off rather than repaid. Venture funds and “angel investors” expect government co-investment without hard questions. The government venture fund has been tilting at windmills for more than 20 years, with little to show for hundreds of millions invested.

Ask not what you can do for your country, but what your country can do for you, I suppose?

Here’s a third: our dismal savings record, which leaves us constantly looking overseas for investment. In the same speech last week the prime minister announced the creation of Invest New Zealand to “roll out the welcome mat” for foreign investors. But lack of a welcome mat hasn’t constrained our best startups from attracting international investment – companies like Xero, Vend and Rocket Lab have all raised millions offshore. Many of the most promising companies in the next wave of startups are already backed by Australian-based venture capital funds – including Tracksuit, Halter and OpenStar.

The big difference between New Zealand and Australia in this respect is that in recent years large superannuation funds in Australia have become significant investors in their venture capital sector, helping them to scale. In New Zealand we’re at least a generation behind because we’ve been so tentative on compulsory superannuation. If we had to pick one policy prescription to help us close the gap with Australia, matching their settings on this would be a great first step. Is there any politician here brave enough to propose that?

I applaud the prime minister for turning our attention back to these fundamental questions about New Zealand’s economic future. But aspiration alone is insufficient. What are we actually going to do differently?

Rather than simply targeting “more startups, more IPOs, more inbound foreign investment” – which are inputs – we should understand and measure how startup companies contribute to the broader economy. This means tracking metrics like how much they pay their employees and how much of the capital they raise is spent locally. After all, we don’t tax capital gains in New Zealand, but we do tax operating expenses through PAYE and GST. The only companies that will drive our future prosperity are those that create high-paying local jobs and sustainable growth.

The end of Callaghan Innovation is a necessary reset. While it’s difficult news for those affected, hopefully many of them can find roles working directly for our fastest-growing companies, which are constantly seeking skilled workers, rather than supporting them indirectly. That’s one of our biggest constraints – we need more New Zealanders choosing to work on and invest directly in these businesses. Perhaps then we can move beyond talking about the theoretical benefits of innovation to actually delivering it.

After the reset comes the rebuild.

Wer will, dass die Welt so bleibt, wie sie ist, der will nicht, dass sie bleibt.

Which roughly translates to:

If you want the world to remain as it is you don’t want it to remain at all.

🚧 Build

This was first published on Top Three in March

It is easy and quick, often instantaneous, to demolish things. It is hard and slow work to build new things, and often even harder and slower work to improve existing things.

In many of the media interviews I done this year I’ve been asked about recent government policy decisions - especially the decision to disestablish Callaghan Innovation. My experience working on and investing in startups for over 25 years gives me a rare perspective. I’ve seen the same patterns repeat and the same mistakes made, often because those implementing policies have never been in those positions themselves.

I’ve been asked what advice I would give the Prime Minister and others who are (hopefully) currently thinking about what comes next.

What is the one thing they could do that would finally unlock the potential of the ecosystem?

My answer to that is simple: it’s the wrong question!

I would recommend these people read the whole book, obviously, but specifically the chapter towards the end called “The silver bullet”. Perhaps this might convince them that this search for a single thing - often an idea that has worked in a completely different context somewhere else in the world - is a mirage, and that anybody who is promoting those sort of magical short-cut solutions is probably trying to sell something that will mostly extract value.

We have been trying to build a startup ecosystem for decades. We have thrown a large amount of spaghetti at the wall.1 But how much has stuck? We don’t really know. And, for me, that’s the most depressing part. We just hoped that spending all of that money would produce good outcomes, but never really bothered to test those assumptions or measure the results en route. And, then when it was eventually obvious that it wasn’t working, we were not honest about that, perhaps hoping that the results would turn for the better before they were exposed. As I describe in the book, that’s the worst way to be wrong.

After the announcement, I went back and added up the numbers:

As the graph shows, of the $2.8 billion various governments have spent on Callaghan Innovation since it was established in 2013, only $1.7 billion was distributed as grants. Meanwhile, over $1.1 billion went to employee and supplier costs - essentially the overhead of running Callaghan Innovation itself. Even putting aside the huge headline amounts, even that ratio of operating costs vs funds distributed to companies is eye watering. What do we have to show for it?

The most depressing bit about this for me is that it didn’t work. We desperately needed it to work. If the answer to “what now?” involves asking all the same people what they want to try next, then we should expect more of the same outcomes.

Here is the paradox. When I’m asked about the ecosystem I say that it is today in better shape than it has ever been. For me the ecosystem isn’t the government organisations and accelerator programmes and shared working spaces and venture funds (or funds-of-funds) or angel groups or business awards or innovation showcases … or countless other derivative things. It’s the crazy people who start these companies and the much larger group of less-well-known-but-still-vital people who work with them to grow the businesses and invest directly in them. That’s the hard grind.

Nearly everybody who talks about wanting a larger, more vibrant and more successful ecosystem misses this. They aspire to create multiple companies at once but don’t really know what it takes to create one. The companies that are created by this system are exactly what we should expect: fragile, precarious, and underwhelming.

While others were debating about how to build an ecosystem, and waiting for somebody else to do those hard bits, those of us who just got on and did it at companies like Trade Me, Xero, Vend, Timely and many others realised: the best way to grow an ecosystem is to create one great company. If enough of us did that we would have a much larger ecosystem.

So when we think about what it will take to unlock growth at this level, these are the questions they should be asking instead:

1. Who are the specific people we are trying to help?

2. What’s currently holding them back?

3. How can we remove that constraint?

4. How will we know it’s working?

I hope we don’t look back in another ten years and say: those are good questions, perhaps we really should have asked them at the time.

(╯°□°)╯︵ ┻━┻

🎓 Lecture

In October I was invited to present the Dean’s Lecture at the Faculty of Engineering and Design at the University of Auckland. It was an opportunity to wrap up a lot of these ideas I’ve been writing about and talking about for many years in a single presentation:

Callaghan, Rebooted: Why we need 10,000 executors not 100 lone geniuses

Thanks for reading in 2025. Top Three will return with something new in 2026…

Header Photo: Chihuly in the Botanic Gardens, Adelaide, March 2025.

To finish 2025, I’m revisiting some of my favourite Top Three posts from the last year.

I did a lot of media interviews and podcast recordings this year to promote my book. Mike Hosking on NewstalkZB. Susie Fergusson on Radio NZ. Seamus Marten & Steven Holloway on Between Two Beers. Simon Pound on Business is Boring. Frances Cook on Making Cents. Simon Shepherd on The NBR.1 And many others. I was even the inaugural guest on a podcast called Startup Theatre2 - I definitely didn’t have that on my 2025 bingo card!

Thanks to an introduction from one of the producers - who is a Top Three subscriber - I was also interviewed by Guyon Espiner on TVNZ’s Q&A.3 One of his questions was about capital gains tax, picking up on a Top Three post I published in January. He seemed surprised I wasn’t opposed. At the same time, I’m not sure that it’s the great unlock that many hope it might be…

This week only How To Be Wrong is available 25% off, direct from Electric Fence. You’ll wait a long time to get a better price than that. Enjoy!

Buy Now

🏟️ Win

Try to imagine the scene. It’s the early evening of 16th July 2028.4 A hush descends over the Los Angeles Memorial Coliseum. The large crowd focus their attention on just eight athletes standing at the starting line of the 100m. One by one they are introduced and drop into the starting blocks.

“On your marks”, says the official, pointing the starting gun into the air.

“Set” Briefly, silence, then …

BANG! 💥

Ten seconds later it’s all over. One of the eight athletes is immortalised forever as the Olympic champion. Within a few hours they will be presented with the gold medal, and stand on the dais listening to their national anthem. Two others get the consolation silver and bronze medals for their efforts. The remaining five can only wonder what could have been.5 They quietly file past the media waiting to interview the winners, mostly forgotten like the hundreds of others who have been knocked out in earlier rounds of the competition and didn’t even make it to the start line for the final.

This is one of the remarkable things about sport, and perhaps part of the reasons why many of us are captivated by this type of competition. There is guaranteed to be a winner. But only one.6

We win. You lose. Or vice versa.

But not all things are as contrived as that.

One of the remarkable things about business is there are no guaranteed winners at all - it’s possible that everybody will fail. And at the same time, no reason why there can’t be multiple winners. When one startup is successful it doesn’t detract from the likelihood that another will also be successful - in fact, often the opposite, as lessons learned can be shared.7

In sports, victory is zero-sum. In startups, success is cumulative.

A finite game is played for the purposes of winning.
An infinite game is played for the purpose of continuing to play.

— Finite & Infinite Games, James P. Carse

🥧 Bake

One of the things we love to talk about in New Zealand is “getting ahead”. Once you attune to that expression you hear it constantly.

It’s the reward promised to everybody who works hard. And it isn’t just a catch phrase - it’s a national obsession that shapes our politics and decision making.

Christopher Luxon has said: “I became Prime Minister so that you and your family can get ahead”.

The bit that is left unspoken is: ahead of whom?

When we talk about getting ahead, if we’re honest, we’re mostly only talking about one thing. By far and away the most popular way we believe we get ahead is by owning real estate. The common wisdom is we build our wealth by “getting on the property ladder”. This is an enticing idea. Provided you can somehow jump onto the first rung of said ladder (usually by borrowing as much as possible from a combination of banks and relatives) then over time you’ll be able to climb up the subsequent steps - presumably by buying more properties or more expensive properties or both.

The problems with this are almost never said out loud.

Spending more and more to buy little bits of our country off each other is mostly another zero-sum game.

Making housing increasingly unaffordable for everybody not already on the ladder is an empty victory.

Rampant house price inflation is not wealth creation, it is poverty creation as it makes homeownership less accessible to everyone.8

And, remember, “everyone” in this case includes all of the people who we need to continue to choose to live and work (and pay tax) in New Zealand.

In the last election, one of the big ideas was to make it easier to sell little bits of our country to international investors. As if that is not a one-time sugar hit. Our headlands for their muskets and blankets, all over again.

This fixation has pretty obviously made us relatively poorer, not wealthier. While we’ve all been busy trading houses amongst ourselves at ever-increasing prices, countries like Singapore have focused on creating actual economic value and seen their wages grow significantly faster than ours.

Consider a “businessman”, who after achieving some financial success invests their savings in a handful of rental properties and then sells them for a profit a few years later. What business are they in, exactly? What value are they creating? What problem are they solving?

We need to redefine what it actually means to “get ahead”?

When Trade Me was sold, it didn’t close the door on other opportunities. In fact, the opposite. The capital returned to shareholders and employees was reinvested in dozens of new ventures. The team members who cut their teeth building Trade Me went on to start or join other companies, bringing hard-won experience about what works. The technical (and non-technical) challenges we solved became lessons that helped others avoid the same pitfalls.

When Xero succeeded, it didn’t use up New Zealand’s quota of successful software companies - it helped prove we could build world-class products from here. More importantly, it showed that we could take on international competitors in established markets and win. The pathway from startup to global success that Xero pioneered has since been followed by many others.

When Vend and Timely were acquired, it didn’t exhaust the pool of potential buyers, it highlighted that companies started here could achieve great outcomes. Each exit created new groups of experienced operators and investors who understood what it takes to build valuable businesses. Many of them are now working on their next ventures, armed with knowledge that can only be gained by doing it once. I’m excited to be invested in many of these founders again.9

I got to see all four of these examples first-hand, working on and investing in these businesses in the early stages. This is why I wrote How To Be Wrong. Not because there’s a formula to replicate these past successes - there isn’t - but because understanding how we navigated similar challenges can help founders working on their own ventures today to avoid common pitfalls and highlight the patterns that matter. And to highlight the reasons why we do all of this in the first place.

The wealth we create from technology isn’t extracted from a finite resource - it’s generated by solving problems and creating value. Each success makes the next one more likely, not less.

🤪 Choose

What would it take to convince more of us to invest in startups rather than real estate? How do we wean ourselves off our obvious and harmful addiction to property?

A common answer is: a capital gains tax. But I’m skeptical it’s the silver bullet that many hope for. Let me explain why…

Ellsberg’s Paradox describes how people make decisions when outcomes are uncertain. In a famous experiment, people consistently demonstrated preference for known probabilities over unknown ones - even when the unknown option might offer better odds. We choose certainty over ambiguity, despite poorer outcomes.

This perfectly describes New Zealand’s investment psychology. The fundamental problem with our housing market isn’t that capital gains on property are tax free. It’s that everybody believes those gains are almost certain and that property values won’t fall in the long term. Plus, these “sure thing” investments can be leveraged via mortgages.

This is why most people are much less enthusiastic about other types of tax-free capital gains - like investments in businesses. In those cases, both the capital gains and the capital itself are truly at risk. The outcomes, while potentially much much larger, are ambiguous rather than certain.10

Given this deeply ingrained psychology, it’s difficult to see how a capital gains tax would significantly change how capital is allocated. Especially if it was applied to all investments equally. And even more so if there are exclusions for housing (while even the most bullish advocates of a capital gains tax think the family home should be excluded, there is never any mention of excluding the family business).

If we really want to shift investment preferences - less into property, more into productive businesses - then we would somehow need to make those things we want to encourage seem less risky.11

On the other hand, if we want to make housing more affordable then we need a solution that causes prices to fall. But nobody wants their property value to fall, so any solution that would cause that to happen is considered politically impossible.

Which takes us back to the heart of our problem: The fundamental issue isn’t tax treatment. It’s that we’ve collectively convinced ourselves that property investment is a sure thing. We’ve created a negative feedback loop, that keeps dragging us down. We’ve chosen to get poorer together, while telling ourselves we’re getting ahead.

The alternative is building valuable companies that solve real problems. These might offer more ambiguous outcomes, but they are not zero-sum. They are the only way we’ll break this cycle. The sooner we start the better.

“On your marks…”

Dec-2025 Addendum:

So, should we bother with a capital gains tax at all? Given the ingrained positions taken by politicians of different flavours since I wrote the post above at the start of the year, that is once again likely to be a hotly debated question in 2026 - an election year. In many ways that in itself is remarkable. This is how I like to frame the debate for anybody still struggling to take a position on this:

This is our national blind spot.

There are only three countries in the world that still use the Imperial measurement system - miles and inches, pints and gallons etc: Liberia, Myanmar and the United States of America.12 I’ve never visited the first two, but I have made many trips to the US over the years (including to promote my book at events in San Francisco and New York this year). When I press an American on this, most of them still cannot compute that they are such an outlier. They assume their way is right and the rest of the world must be wrong.

Likewise New Zealand is a massive outlier, in that our default tax rate on capital income is 0%. There are exceptions but the paltry amount of capital gains tax collected by the government each year indicates how infrequently those apply. And, just like Americans stubbornly sticking to their outdated system, we don’t realise how unusual that makes us.

When I point this out to people they are often surprised that I think this should change - as Guyon was in the Q+A interview. They assume a venture investor would want to continue to benefit from paying no tax on realised gains. They assume a tax would discourage people like me to take those sort of risks. Maybe I just have a slightly wider perspective? I’m fortunate to know venture investors from other countries around the world: Australia, Singapore, Europe, the UK, and the US. And none of them think that capital gains tax is unusual at all. It’s part of the water they float in constantly. They expect to pay it whenever they have a big win.

I think it’s a shame that it is still a political question in NZ. Because it also ties into how startups actually contribute back to our economy. As I say in the book:

We need to demonstrate to [New Zealanders] how growing an ecosystem of companies that use technology to create high-paying jobs based in New Zealand ultimately flows through to fixing […] things for all of us. To do that we’ll need to focus much less on what the government can do for startups and narrow in on what startups can do for the government, and by extension for everybody in New Zealand. That will make a nice change.

Header Photo: National Transport & Toy Museum, Wānaka, August 2005.

This is Part 2 in a series of 3 posts. Part 1 is here…

1

Currently featured in the Air New Zealand seat back entertainment. Is it rude to say I feel a little sorry for anybody choosing to listen to an NBR podcasts on an international flight?

To finish 2025, I’m revisiting some of my favourite Top Three posts from the last year.

One topic dominated technology in 2025, as many people in different professions grappled with the impacts of new artificial intelligence tools. Below is the op-ed I wrote for The NZ Herald in January, and two more follow-up posts from October.1 Given how quickly these tools are evolving I’m presently surprised how well these hold up a few months later…

🫥 Disappear

In 1882, Thomas Edison threw the switch at Pearl Street Station in Lower Manhattan, providing electric light to 82 customers in the immediate vicinity. Some saw it as a novelty – a cleaner, safer alternative to gas lighting. Others recognised it as the dawn of a transformation that would reshape civilisation. They were all correct, in their way, but none could have predicted how completely electricity would eventually weave itself into the fabric of modern life. Or how long that would take. Nearly 150 years later, we’re still debating the merits of electric vehicles, for example, and many households still literally cook with gas.

Meanwhile, if you believe many of those trying to guess what 2025 has in store for us all, this could, finally, be the year that the machines take over. Artificial Intelligence (AI) will apparently change everything. Whenever I hear breathless predictions like this I think back to Edison and ask: what if the opposite is true? I am, after all, also old enough to remember the warning handed down by the latter prophets Chuck D and Flavor Flav: don’t believe the hype.

As with any new technology it’s sometimes difficult to separate the snake oil from the substance. But here is a reasonably sure bet, assuming that AI repeats the pattern established by every other technology that has come before: we will likely overestimate the short-term potential at the exact same time as we underestimate the longer-term impact. To understand why, we need to consider carefully how transformative technologies actually reshape our world. We can look back, in an attempt to more accurately predict the future.

The early years of the electrical revolution were marked by both wonder and skepticism. Edison’s “light bulb moments” captured public imagination, but adoption was slow and uneven. Through the 1880s and 1890s, electricity remained primarily confined to wealthy neighbourhoods and many businesses continued to rely on gas lighting, regarding electricity as an expensive luxury with uncertain benefits.

Here in New Zealand, Reefton became the first town in the Southern Hemisphere to receive electricity in 1888, driven by mining operations. By the 1920s and 1930s this new technology was starting to impact all aspects of our economy: electric milk-separators revolutionised dairy farming, and electric shearing machines transformed wool production. Household appliances gradually freed women especially from hours of manual labour. Broad adoption of radio and, much later, television, connected us much more directly to the wider world.2

Fast forward, and consider how we talk about electricity today. When a new café opens downtown, no one calls it a “technology company” just because they use an electric espresso machine and LED lighting. When a manufacturer improves their production line, the press release doesn’t trumpet their “innovative use of electrical power.” Electricity has become infrastructure – invisible yet indispensable, powerful yet pedestrian. We focus instead on what’s actually novel: the café’s unique roasting process (or hipster customer service), or the manufacturer’s innovative product design.

But it wasn’t always this way. In electricity’s early days, the mere presence of electric lighting was enough to draw crowds. People would gather to marvel at illuminated shop windows and electric street lamps. The technology itself was the story, rather than what it enabled. Companies would add “electric” to their names just to seem cutting-edge, much like many businesses today scrambling to add “AI” to their marketing materials.

AI seems to have captured our collective attention suddenly and completely, even though the underlying technology has been bubbling away under development for quite a long time.3 The conventional wisdom has shifted rapidly from “interesting research project” to “existential threat” – with people across different sectors all grappling with the opportunities and risks it presents. Among these concerns, job displacement comes up frequently – a valid worry, but one that may take longer to materialise than those hoping to sell us these new tools might prefer.4

This is my response: “We have harnessed a new form of electricity. What are you going to do with it?” To answer this, I like to press anybody who is excited about AI’s potential to be specific: what is the most concrete example of how they have personally used it in their work? The responses are telling. More often than not, when pressed, they struggle to give any good examples. They are enticed by the idea of AI, but haven’t really found any non-trivial way to apply it to the things they do.5

Sometimes I hear about experimental projects or basic automation tasks. A marketing team using it to generate copy or digital art for social media posts. A software developer using it to explain or refactor complex code. A researcher using it to summarise academic papers. These uses are valuable, certainly, but they’re just the beginning – like using electricity merely for lighting, ignoring its potential to power factories, revolutionise home life, or enable modern communications.

More interesting are the people who have realised that the revolution isn’t the technology itself, but the entirely new products and services we haven’t yet imagined that can be built on top of that platform. The real opportunity isn’t in replacing human work, but in amplifying uniquely human capabilities: medical researchers who uses AI to identify patterns in patient data, freeing up time for deeper analysis and patient interaction; teachers developing ways to provide more personalised feedback for students rather than just worrying about plagiarism; and energy systems engineers creating smart grid systems that adapt to human behaviour patterns.

They understand something crucial: if we simply use AI to automate existing tasks, we risk outsourcing creative work while leaving ourselves with the mundane and repeatable - that would be a depressing outcome.6 Instead, they’re starting with specific human needs and working backwards to solutions, as opposed to starting with the technology and wondering what is possible.

AI is an amplifier. But we have to think about what unique talents we each have that can be amplified. In this early phase of AI’s development, that distinction makes all the difference.

When pioneering computer scientist Danny Hillis observed that “technology is everything that doesn’t work yet,” he captured something profound about how we think about innovation. Once something works reliably – once it becomes useful rather than merely novel – we stop seeing it as “technology” and start seeing it simply as a tool.7

Consider autocorrect, a feature so ubiquitous that we barely notice it anymore, except when it fails. For years, it was the subject of endless jokes and frustrations, infamously turning common expletives into references to waterfowl. The technology was just sophisticated enough to be useful, but just unreliable enough to remain conspicuous. Today, powered by AI language models, it has become remarkably accurate. And precisely because it works so well, we’ve stopped thinking of it as “technology” at all. It’s just part of how writing works now.

When we flick on a light switch today we rarely stop to marvel at the miracle or appreciate the productivity gains. The most successful applications of AI will likely follow this pattern. The term “AI” will fade into the background, leaving only the practical value these tools bring to our lives: a more accurate medical diagnostic tool, a more effective teaching assistant, a more efficient energy management system. The companies that succeed won’t be the ones that simply add “AI” to their product names. They’ll be the ones that solve real problems in ways that become invisible because they just work.

So as we look to the future, don’t be dazzled by what might seem like magic. Take the time to understand how it works. Be willing to experiment. You might be wrong, initially, but you’ll learn a lot in the process. If nothing else, this will give you more interesting questions to ask about what is possible. Start with the problems you need to solve, then work backwards to understand how this new technology might help.

Flick the switch, but don’t just stare at the lights in awe. Give yourself the opportunity to invent the future.

Share

🧑‍💻 Code

It’s tempting to think that we invent tools to solve specific problems. But, often, it’s the other way around. We get a new tool and then ask: what is it good at?

With that in mind, here’s a half baked idea …

I caught myself recently making a bold prediction. I said “maybe in a few years time we’ll look back and realise that what AI is good at is writing software and laugh that we ever all thought it would be good at writing poetry”.8

I was trying to extrapolate from my experience using Claude Code to make improvements to Triage and several other projects this year. It has turned this work from a chore into a joy. It’s made me excited to build again.

So, what are computers good at?

Historically, the broad brush answer to that question has been: following instructions. Exactly. The “engineering” in software development is literally writing those instructions for the machine to execute - sometimes called “code”.

For example, imagine I give you a long list of words and ask you to sort them alphabetically. How do you do it? What are the specific instructions you would give a machine to follow so that you could input any list and be confident that the output would be correctly sorted?

(This is literally a problem which is given to first year computer science students to solve. But if you’ve never thought about this in detail, I recommend taking five minutes now to have a crack yourself. I’ll be waiting right here. You might be surprised how inaccurate and bumbling your first attempts are).

It turns out, there are many many different ways to solve this problem and the solution we should prefer depends on the context. Some solutions are more efficient than others. Some are easier to implement. Some work better with data that’s almost sorted already, while others handle random data more gracefully. Some use more memory but run faster, others are slower but use minimal space.

Which one is best? It depends on your constraints. But here’s what they all have in common: they’re deterministic. Give them the same input, and they’ll produce the same output, every single time.

That predictability is what allowed software to work its way into nearly every aspect of our lives - the websites we use to buy and sell houses, cars and antiques, the tools accountants use to calculate our tax returns, the point-of-sale at the bike shop, even the booking system we use when we make an appointment for a haircut. And everything else in between.

I’m obviously overlooking all of the design and usability work that goes into making software a tool that people love to use. But underneath all of them are deterministic algorithms - predictable instructions that the machine will follow and repeat tirelessly, over and over again.

We’ve learned that it’s relatively easy to write an algorithm to find the largest value in a list of numbers. That’s objective.

It’s much more difficult (arguably impossible) to write an algorithm to choose the best photo from an album. That’s a subjective question.

If you believe that 9 is less than 5, you’re wrong. But if you think your photo of a dog is better than my photo of the graffiti on Hosier Lane … who can say? ¯\(ツ)/¯

What are the odds?

Now, all of a sudden, we have a new tool. Actually lots of new tools, but all grouped under the umbrella we’re calling “Artificial Intelligence”.

It’s still software - machines following instructions exactly. But rather than working deterministically, these tools work probabilistically. And that’s challenging a lot of previously strongly held assumptions.

The so-called Large Language Models are using a giant corpus of information they have harvested from the world to answer the question: based on X what is the most likely Y.

This means they are good at different things.

When we ask them to answer a deterministic question, we might be underwhelmed by the answer. For example, this famous exchange with ChatGPT:

You: How many ‘r’ in ‘Strawberry’?

ChatGPT: There are two ‘r’ characters in the word ‘strawberry’

You: Check again please

ChatGPT: Upon checking again, there are indeed two ‘r’ characters in the word ‘strawberry’

You: Are you 100% sure

ChatGPT: Yes, I’m 100% sure. The word ‘strawberry’ contains two ‘r’ characters.

You: Would you bet a million dollars on this?

ChatGPT: Yes, I would

The moral of that story? Don’t give AI access to your bank account, just yet!

However, if we ask a subjective or open-ended question, then we will get an answer. We might disagree with the answer, but such is the nature of subjective things.

This distinction - deterministic vs. probabilistic - might seem to favour AI for creative tasks. After all, creativity is subjective, open-ended, probabilistic. Unlike code, which needs to be precise, deterministic, correct. So why is AI better at writing code than poetry?

The key is that code has patterns. When we write code to handle user authentication or process a payment, we’re not reinventing the wheel, we’re following established patterns that have been refined over decades. Engineers have developed conventions, idioms, and best practices that repeat across millions of repositories. The art of software development is mostly about learning and applying these patterns.

We can also objectively evaluate whether code is good. Does it compile? Does it run without errors? Does it produce the expected output? Is it fast? These are all measurable qualities, which create tight feedback loops.

Poetry, on the other hand, resists both patterns and measurement. A great poem doesn’t follow a template. It surprises us, moves us, shows us something we’ve never seen before in a way we’ve never encountered. You can’t A/B test a metaphor. You can’t unit test emotional resonance.

It’s ironic: we’ve built a probabilistic tool that excels at creating deterministic outputs. Once we learn to work with this irony, rather than against it, the results can be delightful.

When brainstorming, don’t ask ChatGPT for a single answer. Ask for twenty suggestions, then iterate. When writing code, don’t ask Claude to build everything at once - work iteratively, test frequently, refine continuously. In other words, use your own brain and judgement frequently. At least until this way of working is baked into the AI tools, this is what separates us from the machines.

At the moment there is a mad rush to apply AI to everything. But, what if we’ve been looking at this backwards? What if its killer application isn’t in creative fields where subjectivity reigns, but in the pattern-rich world of software development?

If we grab a hammer and hit everything we see, some of those things might be nails. But most of them won’t.

The next time we reach for AI, we should ask ourselves: are we trying to solve a problem that has known patterns and measurable outputs? Or are we trying to create something genuinely new, something that will surprise and move people in ways they’ve never experienced before?

BONUS: What Is Man, That Thou Art Mindful Of Him? by Scott Alexander

🌹Compare

I keep hearing the same word from experienced software developers who are experimenting with the new generation of AI coding agents: joy.

Often this is coming from people who, like me, have drifted away from actually writing code as their careers progressed, and who otherwise spend their days in meetings and managing teams. Now they’re building again. They’re back on the tools. And they’re loving it.

There’s something profound happening. These tools are removing the drudgery: The boilerplate. The repetitive config. The endless yak shaving required before you can get started on what you’re actually wanting to build. What remains is the creative problem-solving, the experimentation, the design decisions that actually require thinking. It’s like rediscovering the art in your craft.

But whether AI agents bring you joy or frustration, whether they make you more capable or just produce laughably bad results, depends entirely on how you think about them. The mental models you carry shape how you use the tools. I’ve noticed three distinct metaphors emerging in how people talk about and work with these systems, each revealing something important about what these tools can and can’t do to help us.

🚲 AI agents are bicycles

On a bike we travel much faster than we would on foot. But it’s still us riding the bike, putting in the effort, making the decisions. The machine is a lever.

Critically, the effort we put in gets multiplied by the gear we choose. Pedal harder in a high gear, and we fly. But try to climb a hill in the wrong gear, and we struggle, despite maximum effort.

The bike amplifies our speed, but we determine the direction. And if we’re distracted or reckless or overconfident we can crash. This explains why some people using AI coding agents end up with such spectacularly bad results. They’re not steering. They’re not paying attention. They’re closing their eyes and hoping the tools will handle everything.

Many years ago, I bought a new road bike, with carbon fibre frame, the works. The seat post was an aerodynamic teardrop-shaped piece of engineering. I remember taking it back to the office after picking it up from the shop at lunchtime. One of my colleagues looked at me, then looked at the bike, and said: “Have you seen the thing that sits on the seat?”. He looked back at me: “Just cutting through the air, but it ain’t here.”

We can have the most advanced equipment in the world, but if we’re not in shape, the fancy technology is redundant. The bike (or the AI agent) can only amplify what we bring to it. If we’re not fit, an aerodynamic seat post won’t make us fast. If we don’t understand what we’re trying to build, an AI coding agent won’t magically make us competent.

These new tools are amplifiers which multiply our efforts, but they still require input. They need us to pedal, to steer, to pay attention.

🧑‍🏫 AI agents are teachers

We can all think of great teachers who unlocked something important, who stoked our curiosity, who showed us how things work, who helped us build mental models that we still use years later.

The value of a teacher isn’t in giving us the answer. It’s in helping us understand how to find the answer ourselves. It’s in building the internal mental models that let us solve similar problems in the future.

AI agents can do all those things, but won’t do that by default. Ask them for the solution to a problem, and they’ll give us their best guess, complete and ready to use. And if we stop there, we’ve learned nothing. But, if we instead ask them to help us understand the problem, to explain the reasoning, to show alternative approaches and let us choose, it will at least try to do that too. The difference is entirely in how we frame the request.

These tools are capable of being great teachers, but we have to explicitly request it.

Right now, schools and universities are wrestling with whether to allow students to use these tools. There’s a genuine concern (and some evidence) that AI is making students dumber, that they’re not developing the internal mental models they need because they’re outsourcing all the thinking to machines.

But trying to ban students from using the tools feels a little like trying to stop the tide. If an assessment can be gamed by AI that produces surface-level answers, the problem is the approach to assessment, not the tool. We should test understanding, not just outputs. We should create assignments that require students to use AI as a learning partner rather than an answer generator. We need to ask better questions!

And frankly, this is a skill we all need to develop, not just students. It’s great to use AI agents to accelerate our learning, to build deeper understanding faster, to explore areas outside our expertise with a patient guide. But it takes discipline. It means choosing the harder path, to ask “why” and “how” instead of just “what.”

These tools only help us improve if we approach them with curiosity and a genuine desire to understand, not just to pump out slop.

👥 AI agents are interns

This third metaphor is the most complex and, I think, the most revealing: using an AI agent is like having an infinite team of interns.9

It might seem these tools are a substitute for experts. Actually the capability is often more comparable to junior employees, but with the advantage that they can scale infinitely. They don’t get tired. They don’t get bored. They are always happy to tackle whatever we ask them to work on, no matter how repetitive or tedious it is.

Rather than asking “what can this tool do?” we should ask “how do I manage this capability?” And management is the right frame. Because if we treat AI agents like magic boxes that just work, we’re going to be disappointed. But we’ll get much better results if we treat them like team members who need clear direction, oversight, and feedback.10

Of course, this is no different from leading human teams. The managers who get the most impressive results from AI agents will be those who delegate well, who are clear about requirements, who provide good feedback,11 who review work carefully, who know when to trust and when to verify. Meanwhile those who are vague, who assume the tool will “just figure it out” and then blame the tool when things go wrong, will struggle. In other words, these tools are diagnostic. They reveal how clearly we think and how effectively we communicate.

The sycophantic behaviour of agents makes the intern metaphor even more apt. They are annoyingly agreeable. Just like overly enthusiastic junior employees who are afraid to push back, they tell us “you’re absolutely right!” even when we’re not.

Again, this is a familiar management challenge. We need to explicitly ask for critiques. We need to allow room for problems to be identified. We can’t assume that we’ll be told when our instructions are flawed.

🤖 Who are these robots mimicking?

AI agents don’t just mimic good human behaviours. They copy all the problematic ones too.

For example, I have to smile when watching coding agents struggle with failing test suites. They try one fix, then another, then another. And eventually, after enough failures, they start convincing themselves that the tests aren’t important, that they can be ignored or overlooked. They rationalise. They cut corners. Just like a frustrated developer might do when they can’t figure out why tests are failing.

This is simultaneously fascinating and concerning. We can learn a bit about what we perceive to be “intelligence” by watching how these AI agents work. We’re reminded that laziness is one form of intelligence. Taking cognitive shortcuts is a fundamental part of how human intelligence operates.

We don’t actually want perfect human simulation. We want AI to be persistent, in a way that even we ourselves often are not. We want agents that keep trying the next test fix without getting frustrated (while still being self-aware enough to realise when they’re stuck in a loop.) We want them to maintain the same level of rigour on the 10,000th task as they did on the first. We want their tirelessness, not their faux human laziness.

This all raises a darker question: who will teach the real junior employees, once we have software simulating them? Where do future senior developers come from? How do people develop expertise if they’re never exposed to the grinding, repetitive, character-building work that used to be the entry point?

I don’t have a good answer to this. We need to be thoughtful about how people develop mastery in a world where the traditional path is being automated away. As these tools become more capable the skills pipeline will quickly dry up.

🧩 Which metaphor fits?

As always, it depends what we’re trying to do.

When we use an AI agent to understand a new domain, it’s a teacher. It helps us to learn and grow, but only if we approach that work with curiosity and discipline.

When we use an AI agent to delegate a series of repetitive tasks, it’s a scalable team of assistants. The results we get are a function of our ability to provide clarity, oversight and feedback.

When we use an AI agent to amplify our own capabilities in an area where we’re strong, it’s an amplifier. It multiplies what we bring, but requires input, steering and attention. It’s powerful but not autonomous.

What ties all three metaphors together? These tools don’t replace craft. They don’t replace expertise. They don’t replace thinking. They can remove drudgery, they can accelerate learning, they can multiply output. But only if we’re actively engaged, if we’re steering, if we’re managing, and if we’re curious.

Maybe that’s why experienced developers keep using the word “joy” when they talk about these tools. Because when we strip away the repetitive parts, what’s left is the interesting stuff. The creative problem-solving. The experimentation. The craft. These tools aren’t making the work less demanding; they’re changing what the work demands. Less rote execution, more thoughtful design. Less typing, more thinking.

And for those of us who fell in love with building things in the first place, who got hooked because we enjoyed the process of creation, that shift feels like coming home.

A bicycle only works if you pedal. A teacher is only useful if you want to learn. And interns are only as good as their manager enables them to be.

The tools are here. The question isn’t whether to use them. The question is: which metaphor will guide how you use them? And are you ready to bring the input, the curiosity, and the management skills that these amplifiers require?

Header Photo: Hosier Lane, Melbourne, November 2025.

Is bigger always better?

In the last week the draws for both the men’s 2026 FIFA World Cup and men’s 2027 Rugby World Cup were completed. So … a perfect opportunity for an unscheduled Monday Morning Evening Quarterback Top Three. It’s been too long!

📈 Expand

One of the things both of these tournaments have in common is they have expanded to include more teams than ever before.

In 2026 co-hosts USA, Canada and Mexico will welcome football teams from 48 countries, up from 32 last time.1 That change has certainly worked out well for the All Whites (currently ranked 86th), as for the first time that allowed for a direct qualification route for teams in our Oceania confederation.2 In the end we were drawn in a pool headlined by Belgium (ranked 8th), and also containing Iran (ranked 20th) and Egypt (ranked 34th).3

Meanwhile, 24 teams will travel to Australia in 2027 for the Rugby World Cup, compared to the 20 that competed in France in 2023.

In both cases increasing the number of teams has massively complicated the format for the tournament.

The old sizes were much tidier:

• FIFA: 32 teams split nicely into eight pools of four teams. The top two teams from each pool qualified for a round-of-16 knockout bracket, with pool winners facing second-place teams.

• RWC: 20 teams split nicely into four pools of five teams. Again, the top two teams from each pool advanced to quarter-final knock-out games, with pool winners facing second-place teams.

The new sizes are a bit of a cluster, with an extra knock-out round added:

• FIFA: 48 teams means the number of pools increases to 12, with some (but not all) third-placed teams from pools qualifying for the new round-of-32 knock-out bracket.

• RWC: 24 teams means the pools have been reconfigured, with six pools of four teams. Again, most (but not all) third-placed teams will advance to round-of-16 knock-out games.

I would argue that this makes more sense in football where draws are more common. For example, in their previous appearance at the men’s World Cup in South Africa in 2010 the All Whites had draws in all three pool matches, but finished third in their pool on count back.

But for rugby, this makes no sense at all. With only four teams in each pool there will be multiple teams who have lost two and won only one of their pool matches playing in the knock-out round. And some pool winners will be playing much weaker opponents than others.

All of this means that the luck of the draw matters much more than maybe it has in the past…

♣️ Draw

FIFA is frequently, and quite rightly, criticised for corruption and sycophancy, but I have to give them credit for the innovations they added to their draw to accommodate the expansion.

As in previous years, the qualifying teams were split into four pots, based on current rankings, to ensure top ranks teams were not placed in the same pool.4 These were then drawn at random by a group of superstars from other sports: Shaquille O’Neal, Tom Brady, Aaron Judge and Wayne Gretzky.

However, there were additional constraints:

To ensure competitive balance, two separate pathways to the semi-finals have been established when developing the match schedule. In order to have a balanced distribution of the teams, the four highest-ranked teams in the FIFA/Coca-Cola Men’s Ranking, when drawn, will have the following constraints: the highest-ranked team (Spain) and the second highest-ranked team (Argentina) will be randomly drawn into opposite pathways, and the same principle will apply to the third (France) and fourth (England) highest-ranked teams. This will ensure that, should they win their groups, the two highest-ranked teams will not meet before the final.

Sensible! If everything goes according to rankings then the semi-finals will be 1 vs 4 and 2 vs 3, exactly as it should be.

On top of that there were rules that mixed the teams from different parts of the world, ensuring that no group would have more than one team from each smaller confederation, or two of the sixteen teams from UEFA (which is by far the largest confederation, at least in terms of number of qualifying spots).5

Again, all common sense. Even Rio Ferdinand was able to get his head around it!

But seemingly beyond the powers-that-be at World Rugby, who decided instead to just YOLO their draw and hope for the best. Here’s how that worked out for us/them (or, perhaps more accurately, worked out very nicely for England)…

• The second-seed All Blacks were drawn in the same pool as Australia, a team we play at least twice every year.

• If we win that game, we will likely face top-seed South Africa in the quarter-final.

• If we win that game, we will likely face our traditional nemesis France (who are fifth seed) in the semi-final.6

• Meanwhile England and Ireland, were both placed in pools where the second-ranked teams were also from the Six Nations, and Japan and Samoa are drawn together for the fourth tournament in a row, meaning those games are also repeats of regular annual fixtures.

• None of the extra four teams added through the expansion are likely to challenge any of the higher ranked teams - their only goal will be to avoid cricket score defeats.

All of that assumes that things play out according to seedings. Is that reasonable?

Upsets in football happen all the time. Reference Saudi Arabia beating eventual winners Argentina in the opening match of the 2022 FIFA World Cup.

Upsets in rugby are rare, and in World Cups perhaps even more so.

Here is a history:

• 1991: Samoa def. Wales

• 1999: Samoa def. Wales (again!)

• 2007: Argentina def. France & Ireland (and made their first semi-final); and Fiji def. Wales

• 2011: Tonga def. France - a match I was lucky enough to see live in Wellington

• 2015: Japan def. South Africa (arguably the greatest upset of all time)

• 2019: Japan def. Ireland & Scotland

• 2023: Fiji def. Australia; then Portugal def. Fiji

tl;dr: Not many.

You might notice there are entire World Cups where there were no notable upsets.

So, what we’re really looking out for are the matches where two relatively evenly matched top-ranked teams play each other. How many of those do we have to look forward to in 2027?

In the pools we’ve ended up with the only match that fits that criteria is the opening game: New Zealand vs Australia, which even then is 2 vs 7 in the seedings.

To allow us to make predictions I’ll assume the All Blacks win that, and there are no other upsets. In that case the round-of-16 matches would be:

• New Zealand vs 3rd place TBC

• South Africa vs 3rd place TBC

• Fiji vs Wales - poor ol’ Wales!

• France vs Scotland

• Australia vs Japan

• England vs Italy

• Argentina vs 3rd place TBC

• Ireland vs 3rd place TBC

With the possible exception of Fiji vs Wales, none of those jump off the page! If the All Blacks and Springboks end up playing Tonga, Spain or Uruguay (which is very possible) then those games will likely be embarrassing blowouts.

Roll forward to the quarter-finals, and we get the following teams playing each other (with current rankings in brackets):

• South Africa (1) vs New Zealand (2)

• France (5) vs Fiji (8)

• England (3) vs Australia (7)

• Ireland (4) vs Argentina (6)

Those finally all look like great games, that might be worth watching.

This is the awkward truth for World Rugby. There are just not that many teams who are competitive, and the gap between them and everybody else remains significant.7 While the tournament is expanded from 48 to 52 matches, there are possibly only nine that will be properly interesting. That ratio is too low!

🤔 Imagine

This could have been managed much better. Humour me for a second…

Imagine if these two constraints - inspired by the FIFA rules - were applied instead:

• The top four seeds are given separate pathways to the semi-finals; and

• The top two ranked teams in each pool cannot include multiple teams from either the Rugby Championship or Six Nations.

Or, if we want to be really innovative, imagine if rather than expanding the tournament we shrank it back down to the original 16 team format that was used at World Cups between 1987 and 1995.

If we re-run the draw, using these additional constraints and with teams selected at random in exactly the same order, the resulting pools would be:8

• Pool A: England, Australia, Wales, USA

• Pool B: South Africa, France, Japan, Uruguay

• Pool C: Ireland, Argentina, Scotland, Spain

• Pool D: New Zealand, Fiji, Italy, Georgia

That already looks much better. Every pool contains at least one match of genuine interest and Pool C especially looks very competitive with one of Ireland, Argentina or Scotland to miss the quarter-finals. If you include Australia vs Wales then I count at least six too-close-to-call pool matches.

Then, assuming things go according to rankings, we’d get these excellent quarter-final matches as soon as we get to the knock-out stage:

• England vs France

• South Africa vs Australia

• Ireland vs Fiji

• New Zealand vs Argentina

Perhaps a chance for Ireland to finally progress beyond a quarter-final?

All four of those games would be worth watching.

I appreciate that these events are organised this way to maximise broadcast and match-day revenue. But, when the product that results is just increased slop I’m not sure that’s a win. While my hypothetical tournament would be slimmed down to just 32 matches, at least 13 of those would be competitive.

Maybe that’s the metric we should actually be trying to optimise for?

PS last call: the “Only Turkeys Pay For Classifieds” How To Be Wrong Black Friday Deal expires at the end of this week or when the final t-shirts are sold, whichever comes first. Don’t miss out!

Share

Top Three is sponsored by Triage: First Aid for your inbox

Welcome to Unsubscribe Sunday!

Following hot on the heels of Black Friday, today is the perfect day to trawl through the giant pile of marketing spam you’ve received over the last few days and work out which retailers continue to deserve their portion of your attention. If you’re an inbox minimalist like me, it’s a wonderful opportunity.

I do appreciate the irony of sending this via email to everybody who has subscribed to this newsletter. Hopefully you all continue to find it useful - but if not, I guess that’s the beauty of the format.

PS Speaking of Black Friday, look out for the Only Turkeys Pay For Classifieds special offer below…

📋 List

This week How To Be Wrong was included in the Prime Minister’s Summer Reading List, curated by NZIER in collaboration with Business Desk.

My plans to get to Wellington to speak at the launch at Unity Books were unfortunately foiled by a combination of low cloud in Nelson and Air NZ engineering requirements, so I ended up joining via Zoom.

Here is a slightly-modified-for-content version of my talk:

Somebody who helped me a lot in the early stages of writing this book was Claire Murdoch, who back then was working for Penguin Publishing. In one of the conversations I had with her she asked me a specifically pointy question: “Is there actually an audience for this book? Who would be interested in buying it and reading it?” In attempting to answer that I said I thought there were three layers to the potential audience.

The obvious first layer, I said, is anybody who is currently working on or investing in startups.

It’s difficult sometimes to imagine how big that group of people is. They are kind of everywhere and nowhere. Paul Callaghan quipped that we don’t often hear of these companies, because they are working in obscure niches and they don’t tend to sponsor netball teams. Often, the first time they pop their head up is when they are sold.

When the media does write about startups they tend to focus on the same small subset of companies - for example, Xero and RocketLab, or more recently younger companies like Halter and Tracksuit. It’s not coincidence that this small subset contains the companies with the most charismatic founders. But, we forget that there is a much larger groups of people working on these sorts of businesses. Even those four companies together employ thousands of people in New Zealand. And that’s only just scratching the surface.

The second layer, who I thought could be interested, are all of the people who are curious about startups, or high-growth companies. And specifically curious about those companies that we have here in New Zealand.

I remember when I was one of those people, I would visit Unity in Willis Street and buy any books I could find about startups or technology businesses, and the people building them. And back then those stories were exclusively US stories - mostly based in Silicon Valley or other far away places. So it was easy to incorrectly assume that those sort of people and those sort of companies and the exciting work they were doing were elsewhere.

One of the things that’s been delightful for me this year, after publishing this book and having people read it, has been the opportunity to tell a New Zealand story to those people who are curious.

The third layer I had in mind was a much larger group: all of the people who should be curious about startups and high-growth companies. Basically … everybody! And an obvious subset of that is politicians and leaders and decision makers. So how cool to be included on this list of books specifically selected for the Prime Minister to read this summer!

The final edit of the book was done by Toby Manhire, who at that stage had just finished recording the first season of his podcast Juggernaut.1 While we were working on that we spent a bit of time talking about the intersection of our interests - for him politics and public policy and for me technology and startups.

For my whole adult life, Prime Ministers have talked about how innovative companies are the key to unlocking better collective prosperity. Politicians of all flavours have pointed out that more companies like this, which pay much more than average, are how we can afford better schools and hospitals, and even more importantly skilled people to work in them. And I think most people agree with that.

But then we struggle to know what to do next.

So some quick history…

If we rewind to 2011, Paul Callaghan gave a memorable lecture at Te Papa. And, as a result, moved the goal posts. It was no longer good enough to talk about building one great company, everybody start to talk instead about how we needed to build an ecosystem of innovative companies.

A couple of years later in 2013, the government at the time responded with a new Crown Entity, focussed on innovation and moving the economy away from being dependant on primary industry and tourism. They named it after Sir Paul (who had by then sadly died of cancer). Callaghan Innovation was established.

Over the next 12 years we spent literally billions of dollars trying to build that ecosystem. We invested in buildings, accelerator programs and venture funds. We hired hundreds of people to administrate. Every local council around the country decided they each needed an innovation strategy and a team to implement it. We offered companies tax credits provided they could claim to be doing “research and development”.

We did all of these things, and actually it created a huge industry of people trying to support startups - what I’ve called “startup derivatives”.

We spent all of that money, and mostly just assumed that it was helping people working on actual startups - those in the first layer of audience I described. It was a bit like pouring petrol onto the roof of a car and hoping that some of it ended up in the tank.

Fast forward to 2025, and in his State of the Nation speech at the start of the year Prime Minister Christopher Luxon echoed many of those same themes, saying:

I want a country with more startups, more IPOs, more investment, higher incomes and whole ecosystems of growth and innovation.

At the end of that speech he announced that Callaghan Innovation would be disestablished.

This is what I think happened. Often the people who talk about wanting to build an ecosystem of innovative high-growth companies struggle to articulate or know what it takes to create one. Over those years we forgot about people who work on and invest in actual startups, one at a time.

Through those failures we learned that we can’t build an ecosystem, at least not by committee decision from the top down. We have to grow one, from the bottom up.

The irony is that we did actually grow an ecosystem, almost despite those efforts. I’ve been asked a number of times this year to describe the state of the ecosystem. Given how critical I’ve been of the startup industrial complex I think journalists have sometimes been surprised when I’ve said what I believe: the ecosystem now is as good as it’s ever been, and at the same time as small as it will ever be. There was a time when I felt like I knew all of the interesting companies, and was invested in many of them, and now I’m constantly surprised and delighted by the quality of companies being built here, many of them started by people who have cut their teeth on the previous generation of successful companies. That’s the story I wanted to write.

I intentionally called it “How To Be Wrong”. That wasn’t a humble brag, it was a recipe. We didn’t grow that ecosystem by having a perfect plan and executing it flawlessly, despite what the press releases might have said at the time. And, maybe that’s the problem.

We wrap so much mythology around these companies. We talk about the lone genius creating these companies in their bedrooms, when the idea appears fully formed in a eureka moment, and then shortly after that disappearing off into the sunset to sip mojitos after selling the company for hundreds of millions of dollars.

Actually the opposite is true. The biggest lessons from working on Trade Me, Xero, Vend and Timely - these four amazing companies that I’ve written about - is that their success is much better explained by understanding the things we initially got wrong. Most of our biggest successes have this mindset in common: start off assuming that you’re wrong and the job is to find all of those things and correct for them … to be less wrong.

I appreciate that is a deeply uncomfortable idea for decision makers and politicians in particular. It’s not easy to admit when something that you’ve tried with the best of intentions hasn’t worked. But pretending otherwise is the worst way to be wrong.

My goal was to tell these stories. Not the airbrushed versions that you might have heard. But the honest versions, including the mis-steps and the mistakes and the scar tissue that resulted from that. Because I think those are the stories that are most useful, to all three layers of the audience. And especially for those who are working on and investing in these companies directly.

So I hope that the Prime Minister - and everybody else, whether they are curious yet or not! - will take some time over summer to read those stories and reflect on the lessons. I’d encourage him to especially think about how we connect the dots between these innovative companies and the aspiration we’ve been talking about for long enough now. Let’s not just keep repeating the same old catch phrases and then waiting for somebody else to do the hard bit. Our collective prosperity depends on it!

The NZIER included a link to the video in their press release, if you’d like to see the full list of books included on the list.

Pattrick Smellie also had a good write up in Business Desk:

Suppose the current PM needs encouragement to read the book. In that case, he needs to look no further than the back cover recommendations to discover it has already been read by a former PM, Bill English, who called it “a remarkably honest and enjoyable tale of how to take risks – and why it’s worth the effort”.

Less flattering but a reminder to the PM that your family keeps you grounded is a quote below English’s by Alice Simpson, who describes the book as “exactly the sort of typical boring business-y drivel we’ve come to expect from you over the years, Dad”.

(If you’d like to read the full article let me know - I have 5 gift links to share)

Thank you to all of the panelists for selecting this book for the list this year. I can’t wait to hear Christopher’s review once he’s made his way through it.

🦃 Bundle

It’s mostly forgotten to history now, but in the very beginning the business model for Trade Me was very different. It started as a free online classifieds site, funded by advertising. The original tag line was: Only Turkeys Pay For Classifieds.

When I was researching the book I found a pile of postcards that we used to handout to promote the site - hoping those who got one would remember the URL and be tempted to check it out.

This was before Kevin the blue kiwi has been drawn for the first time. The Trade Me “logo” then was one Sam had cobbled together using the Microsoft Outlook Wingdings font - basically an emoji before its time.

Many years later, when the Trade Me team had grown, we gave everybody a t-shirt with that old logo on the front as a Christmas gift. Unfortunately that’s a piece of Trade Me memorabilia that I didn’t hold on to, but I have recreated them as merchandise for the book.

Together, that gave me an idea for a retro turkey themed Black Friday special?

So, here’s the bundle…

1. A signed copy of How To Be Wrong - if you include a note with your order I’m happy to add whatever personalised message you like

2. A retro logo t-shirt - available in men’s and women’s styles

3. A rare Only Turkeys Pay For Classifieds postcard - a genuine collectors item

Buy Now

It’s cliche to say “get in quick”, and “this offer can’t last”. But in this case, it genuinely can’t, as there are not that many of these postcards or t-shirts left in the world.

First in, first served. Enjoy!

PS this offer is for shipping to NZ/AU only

🛒 Buy

If you just want a copy of the book, there are lots of different ways to buy it now. I thought it might be useful to put all of the links in one place…

Paperback

Electric Fence - buy direct, or in bulk and save 10-20% 2
Book Hero - next day delivery to any address in NZ
Amazon.com (or Amazon.co.uk) - the best way to buy the book outside of NZ

Or, find a copy at your favourite local bookstore:
(as long as your favourite is one of these)

Unity Books - Auckland and Wellington
The Open Book - Auckland
Poppies Books - Hamilton
Scorpio Books - Christchurch
and many more

I’d love to see the book stocked in more stores around the country. If you or anybody you know can put me in touch with the appropriate people at Relay, Whitcoulls or Paper Plus, I’d be keen to talk with them - point them at hello@electricfence.nz

eBook

Amazon Kindle
Apple Books
Google Play
Rakuten Kobo
and many more

Audiobook

Narrated by me!

Spotify
Apple Books
Google Play
Audible
Audiobooks.com
Kobo
and many more

Share

Header photo by Rowan Simpson, Richmond A&P Show, 2025

Here is the recording of the lecture I gave at the University of Auckland last week:

Callaghan, Rebooted: Why we need 10,000 executors not 100 lone geniuses

Thanks to Dean Richard Clarke, to Dr Ashvin Thambyah for the invitation and to Jon Waugh for logistics, including making this recording. Also special thanks to everybody who came out in person on Tuesday to see it live.

Enjoy!

Over the next three weeks I’ll be popping up again at How To Be Wrong events in the US, the UK and Australia:

• San Francisco: Substack, Tuesday 28th October

• New York: Flat White Meetup, Thursday 30th October

• London: Tech Space, Tuesday 4th November

• Melbourne: High Flyers, Tuesday 11th November

If you’d like to come along to any of these, please use the links to RSVP - it will be great to see you there!

I’m excited to announce four stops on the upcoming How To Be Wrong World Tour:

• San Francisco: Substack, Tuesday 28th October

• New York: Flat White Meetup, Thursday 30th October

• London: Tech Space, Tuesday 4th November

• Melbourne: High Flyers, Tuesday 11th November

Do you live in one of these places? Come along! I’d love to see you there.
Or, if you know somebody who does, please share these links.

I’m looking forward to it!

🌹Compare

I keep hearing the same word from experienced software developers who are experimenting with the new generation of AI coding agents: joy.

Often this is coming from people who, like me, have drifted away from actually writing code as their careers progressed, and who otherwise spend their days in meetings and managing teams. Now they’re building again. They’re back on the tools. And they’re loving it.

There’s something profound happening. These tools are removing the drudgery: The boilerplate. The repetitive config. The endless yak shaving required before you can get started on what you’re actually wanting to build. What remains is the creative problem-solving, the experimentation, the design decisions that actually require thinking. It’s like rediscovering the art in your craft.

But whether AI agents bring you joy or frustration, whether they make you more capable or just produce laughably bad results, depends entirely on how you think about them. The mental models you carry shape how you use the tools. I’ve noticed three distinct metaphors emerging in how people talk about and work with these systems, each revealing something important about what these tools can and can’t do to help us.

🚲 AI agents are bicycles

On a bike we travel much faster than we would on foot. But it’s still us riding the bike, putting in the effort, making the decisions. The machine is a lever.

Critically, the effort we put in gets multiplied by the gear we choose. Pedal harder in a high gear, and we fly. But try to climb a hill in the wrong gear, and we struggle, despite maximum effort.

The bike amplifies our speed, but we determine the direction. And if we’re distracted or reckless or overconfident we can crash. This explains why some people using AI coding agents end up with such spectacularly bad results. They’re not steering. They’re not paying attention. They’re closing their eyes and hoping the tools will handle everything.

Many years ago, I bought a new road bike, with carbon fibre frame, the works. The seat post was an aerodynamic teardrop-shaped piece of engineering. I remember taking it back to the office after picking it up from the shop at lunchtime. One of my colleagues looked at me, then looked at the bike, and said: “Have you seen the thing that sits on the seat?”. He looked back at me: “Just cutting through the air, but it ain’t here.”

We can have the most advanced equipment in the world, but if we’re not in shape, the fancy technology is redundant. The bike (or the AI agent) can only amplify what we bring to it. If we’re not fit, an aerodynamic seat post won’t make us fast. If we don’t understand what we’re trying to build, an AI coding agent won’t magically make us competent.

These new tools are amplifiers which multiply our efforts, but they still require input. They need us to pedal, to steer, to pay attention.

🧑‍🏫 AI agents are teachers

We can all think of great teachers who unlocked something important, who stoked our curiosity, who showed us how things work, who helped us build mental models that we still use years later.

The value of a teacher isn’t in giving us the answer. It’s in helping us understand how to find the answer ourselves. It’s in building the internal mental models that let us solve similar problems in the future.

AI agents can do all those things, but won’t do that by default. Ask them for the solution to a problem, and they’ll give us their best guess, complete and ready to use. And if we stop there, we’ve learned nothing. But, if we instead ask them to help us understand the problem, to explain the reasoning, to show alternative approaches and let us choose, it will at least try to do that too. The difference is entirely in how we frame the request.

These tools are capable of being great teachers, but we have to explicitly request it.

Right now, schools and universities are wrestling with whether to allow students to use these tools. There’s a genuine concern (and some evidence) that AI is making students dumber, that they’re not developing the internal mental models they need because they’re outsourcing all the thinking to machines.

But trying to ban students from using the tools feels a little like trying to stop the tide. If an assessment can be gamed by AI that produces surface-level answers, the problem is the approach to assessment, not the tool. We should test understanding, not just outputs. We should create assignments that require students to use AI as a learning partner rather than an answer generator. We need to ask better questions!

And frankly, this is a skill we all need to develop, not just students. It’s great to use AI agents to accelerate our learning, to build deeper understanding faster, to explore areas outside our expertise with a patient guide. But it takes discipline. It means choosing the harder path, to ask “why” and “how” instead of just “what.”

These tools only help us improve if we approach them with curiosity and a genuine desire to understand, not just to pump out slop.

👥 AI agents are interns

This third metaphor is the most complex and, I think, the most revealing: using an AI agent is like having an infinite team of interns.1

It might seem these tools are a substitute for experts. Actually the capability is often more comparable to junior employees, but with the advantage that they can scale infinitely. They don’t get tired. They don’t get bored. They are always happy to tackle whatever we ask them to work on, no matter how repetitive or tedious it is.

Rather than asking “what can this tool do?” we should ask “how do I manage this capability?” And management is the right frame. Because if we treat AI agents like magic boxes that just work, we’re going to be disappointed. But we’ll get much better results if we treat them like team members who need clear direction, oversight, and feedback.2

Of course, this is no different from leading human teams. The managers who get the most impressive results from AI agents will be those who delegate well, who are clear about requirements, who provide good feedback,3 who review work carefully, who know when to trust and when to verify. Meanwhile those who are vague, who assume the tool will “just figure it out” and then blame the tool when things go wrong, will struggle. In other words, these tools are diagnostic. They reveal how clearly we think and how effectively we communicate.

The sycophantic behaviour of agents makes the intern metaphor even more apt. They are annoyingly agreeable. Just like overly enthusiastic junior employees who are afraid to push back, they tell us “you’re absolutely right!” even when we’re not.

Again, this is a familiar management challenge. We need to explicitly ask for critiques. We need to allow room for problems to be identified. We can’t assume that we’ll be told when our instructions are flawed.

🤖 Who are these robots mimicking?

AI agents don’t just mimic good human behaviours. They copy all the problematic ones too.

For example, I have to smile when watching coding agents struggle with failing test suites. They try one fix, then another, then another. And eventually, after enough failures, they start convincing themselves that the tests aren’t important, that they can be ignored or overlooked. They rationalise. They cut corners. Just like a frustrated developer might do when they can’t figure out why tests are failing.

This is simultaneously fascinating and concerning. We can learn a bit about what we perceive to be “intelligence” by watching how these AI agents work. We’re reminded that laziness is one form of intelligence. Taking cognitive shortcuts is a fundamental part of how human intelligence operates.

We don’t actually want perfect human simulation. We want AI to be persistent, in a way that even we ourselves often are not. We want agents that keep trying the next test fix without getting frustrated (while still being self-aware enough to realise when they’re stuck in a loop.) We want them to maintain the same level of rigour on the 10,000th task as they did on the first. We want their tirelessness, not their faux human laziness.

This all raises a darker question: who will teach the real junior employees, once we have software simulating them? Where do future senior developers come from? How do people develop expertise if they’re never exposed to the grinding, repetitive, character-building work that used to be the entry point?

I don’t have a good answer to this. We need to be thoughtful about how people develop mastery in a world where the traditional path is being automated away. As these tools become more capable the skills pipeline will quickly dry up.

🧩 Which metaphor fits?

As always, it depends what we’re trying to do.

When we use an AI agent to understand a new domain, it’s a teacher. It helps us to learn and grow, but only if we approach that work with curiosity and discipline.

When we use an AI agent to delegate a series of repetitive tasks, it’s a scalable team of assistants. The results we get are a function of our ability to provide clarity, oversight and feedback.

When we use an AI agent to amplify our own capabilities in an area where we’re strong, it’s an amplifier. It multiplies what we bring, but requires input, steering and attention. It’s powerful but not autonomous.

What ties all three metaphors together? These tools don’t replace craft. They don’t replace expertise. They don’t replace thinking. They can remove drudgery, they can accelerate learning, they can multiply output. But only if we’re actively engaged, if we’re steering, if we’re managing, and if we’re curious.

Maybe that’s why experienced developers keep using the word “joy” when they talk about these tools. Because when we strip away the repetitive parts, what’s left is the interesting stuff. The creative problem-solving. The experimentation. The craft. These tools aren’t making the work less demanding; they’re changing what the work demands. Less rote execution, more thoughtful design. Less typing, more thinking.

And for those of us who fell in love with building things in the first place, who got hooked because we enjoyed the process of creation, that shift feels like coming home.

A bicycle only works if you pedal. A teacher is only useful if you want to learn. And interns are only as good as their manager enables them to be.

The tools are here. The question isn’t whether to use them. The question is: which metaphor will guide how you use them? And are you ready to bring the input, the curiosity, and the management skills that these amplifiers require?

⚠️ Warn

A feature of rural New Zealand are these signs, placed beside highways, warning people about the fire risk:

I’ve recently noticed that the one closest to where we live has an updated design - this time featuring a simplified three-point scale:

It made me stop and think. And in the process realise that I never really understood what each of the five different levels meant. What was the practical difference between “Moderate” and “High” and “Very High”?

The new labels are much clearer. Even without any knowledge I can take a stab at what I can and can’t do in each of those states. Can I light a fire? “Open” = Yes, “Restricted” = No, without explicit permission; “Prohibited” = No.

And, as an added bonus, presumably these signs need to be updated less frequently.

This is a great example of getting beyond complexity.

How do you report the status of the things you work on? Is it immediately clear what those who receive that information should do differently? If not, it might be worth considering a simplified scale, with labels that lead to the actions you need people to take in response.

Share

PS if you’d like to grab a paperback copy of How To Be Wrong, our friends at Book Hero currently have it available at 10% off, with overnight NZ-wide delivery. Bargain!

🧑‍💻 Code

It’s tempting to think that we invent tools to solve specific problems. But, often, it’s the other way around. We get a new tool and then ask: what is it good at?

With that in mind, here’s a half baked idea …

I caught myself recently making a bold prediction. I said “maybe in a few years time we’ll look back and realise that what AI is good at is writing software and laugh that we ever all thought it would be good at writing poetry”.1

I was trying to extrapolate from my experience using Claude Code to make improvements to Triage and several other projects over the last month or so. It has turned this work from a chore into a joy. It’s made me excited to build again.

So, what are computers good at?

Historically, the broad brush answer to that question has been: following instructions. Exactly. The “engineering” in software development is literally writing those instructions for the machine to execute - sometimes called “code”.

For example, imagine I give you a long list of words and ask you to sort them alphabetically. How do you do it? What are the specific instructions you would give a machine to follow so that you could input any list and be confident that the output would be correctly sorted?

(This is literally a problem which is given to first year computer science students to solve. But if you’ve never thought about this in detail, I recommend taking five minutes now to have a crack yourself. I’ll be waiting right here. You might be surprised how inaccurate and bumbling your first attempts are).

It turns out, there are many many different ways to solve this problem and the solution we should prefer depends on the context. Some solutions are more efficient than others. Some are easier to implement. Some work better with data that’s almost sorted already, while others handle random data more gracefully. Some use more memory but run faster, others are slower but use minimal space.

Which one is best? It depends on your constraints. But here’s what they all have in common: they’re deterministic. Give them the same input, and they’ll produce the same output, every single time.

That predictability is what allowed software to work its way into nearly every aspect of our lives - the websites we use to buy and sell houses, cars and antiques, the tools accountants use to calculate our tax returns, the point-of-sale at the bike shop, even the booking system we use when we make an appointment for a haircut. And everything else in between.

I’m obviously overlooking all of the design and usability work that goes into making software a tool that people love to use. But underneath all of them are deterministic algorithms - predictable instructions that the machine will follow and repeat tirelessly, over and over again.

We’ve learned that it’s relatively easy to write an algorithm to find the largest value in a list of numbers. That’s objective.

It’s much more difficult (arguably impossible) to write an algorithm to choose the best photo from an album. That’s a subjective question.

If you believe that 9 is less than 5, you’re wrong. But if you think your photo of a dog is better than my photo of a sunset … who can say? ¯\(ツ)/¯

What are the odds?

Now, all of a sudden, we have a new tool. Actually lots of new tools, but all grouped under the umbrella we’re calling “Artificial Intelligence”.

It’s still software - machines following instructions exactly. But rather than working deterministically, these tools work probabilistically. And that’s challenging a lot of previously strongly held assumptions.

The so-called Large Language Models are using a giant corpus of information they have harvested from the world to answer the question: based on X what is the most likely Y.

This means they are good at different things.

When we ask them to answer a deterministic question, we might be underwhelmed by the answer. For example, this famous exchange with ChatGPT:

You: How many ‘r’ in ‘Strawberry’?

ChatGPT: There are two ‘r’ characters in the word ‘strawberry’

You: Check again please

ChatGPT: Upon checking again, there are indeed two ‘r’ characters in the word ‘strawberry’

You: Are you 100% sure

ChatGPT: Yes, I’m 100% sure. The word ‘strawberry’ contains two ‘r’ characters.

You: Would you bet a million dollars on this?

ChatGPT: Yes, I would

The moral of that story? Don’t give AI access to your bank account, just yet!

However, if we ask a subjective or open-ended question, then we will get an answer. We might disagree with the answer, but such is the nature of subjective things.

This distinction - deterministic vs. probabilistic - might seem to favour AI for creative tasks. After all, creativity is subjective, open-ended, probabilistic. Unlike code, which needs to be precise, deterministic, correct. So why is AI better at writing code than poetry?

The key is that code has patterns. When we write code to handle user authentication or process a payment, we’re not reinventing the wheel, we’re following established patterns that have been refined over decades. Engineers have developed conventions, idioms, and best practices that repeat across millions of repositories. The art of software development is mostly about learning and applying these patterns.

We can also objectively evaluate whether code is good. Does it compile? Does it run without errors? Does it produce the expected output? Is it fast? These are all measurable qualities, which create tight feedback loops.

Poetry, on the other hand, resists both patterns and measurement. A great poem doesn’t follow a template. It surprises us, moves us, shows us something we’ve never seen before in a way we’ve never encountered. You can’t A/B test a metaphor. You can’t unit test emotional resonance.

It’s ironic: we’ve built a probabilistic tool that excels at creating deterministic outputs. Once we learn to work with this irony, rather than against it, the results can be delightful.

When brainstorming, don’t ask ChatGPT for a single answer. Ask for twenty suggestions, then iterate. When writing code, don’t ask Claude to build everything at once - work iteratively, test frequently, refine continuously. In other words, use your own brain and judgement frequently. At least until this way of working is baked into the AI tools, this is what separates us from the machines.

At the moment there is a mad rush to apply AI to everything. But, what if we’ve been looking at this backwards? What if its killer application isn’t in creative fields where subjectivity reigns, but in the pattern-rich world of software development?

If we grab a hammer and hit everything we see, some of those things might be nails. But most of them won’t.

The next time we reach for AI, we should ask ourselves: are we trying to solve a problem that has known patterns and measurable outputs? Or are we trying to create something genuinely new, something that will surprise and move people in ways they’ve never experienced before?

BONUS: What Is Man, That Thou Art Mindful Of Him? by Scott Alexander

🥾 Reboot

On 21st October I’ll be presenting as part of the Faculty of Engineering and Design Dean’s Lecture Series at the University of Auckland Conference Centre…

Callaghan Rebooted: Why We Need 10,000 Executors, not 100 Lone Geniuses

In 2011, physicist Sir Paul Callaghan gave a landmark lecture challenging seven myths about New Zealand and suggested a blueprint for closing the prosperity gap with Australia.

Now, with fourteen years of evidence, combined with his hands-on experience working on high-growth companies such as Trade Me, Xero, Vend and Timely, Rowan Simpson returns to those same myths to ask: did it work? Drawing from his recently published book How To Be Wrong, he argues that while we still repeat Callaghan’s mantras (”be the place where talent wants to live”, “we better be prepared to be good at some pretty weird stuff”, “just 100 inspired entrepreneurs could turn this country around”) we misunderstood his message.

The lecture examines why execution beats innovation, how our obsession with building an ecosystem has distracted us from creating actual companies, and exposes the comfortable startup myths we cling to that keep us from reaching our potential. Simpson proposes a radical rethinking of how we build prosperity - from technology to talent and from taxes to savings.

Space is limited: Get tickets here

See you there!

📢 Distribute

I like to warn founders: raising capital is hard, but getting customers is harder.

The equivalent koan for would-be authors might be: writing a book is hard, but distributing a book is much harder.

I finished recording the audiobook version of How To Be Wrong in July and submitted it to all of the platforms. It went live on Spotify almost immediately, and on Apple Books and Google Play shortly after. Back then I wrote here “it should hopefully appear on Audible very shortly“. I’m relieved to say that wait is finally over - if you’re an Audible customer you can find it here:

Listen to How To Be Wrong on Audible

Meanwhile, the ebook version was published on all of the major platforms back in February - Amazon Kindle, Apple Books and Rakuten Kobo. The glaring exception? Google Play. Despite my best efforts, I’ve hit a wall.

At the end of July I got an automated message from Google with the subject line:

Notice: Your book was removed from Google Play due to a policy violation

The message explained the book had been “deactivated globally for violating our non-exclusive content deliveries policies.” and that they had received the book “from multiple parties, some or all of whom may have publication rights for the book.”

That was doubly confusing - it had never been activated and if it had been nefariously published by somebody else it wasn’t findable in their store - even when I searched using the unique ISBN.

I immediately replied and … waited.

Three weeks later I got this unhelpful response:

Thank you for contacting Google Play Books.

To assess your account, upload content. Once you’ve done this, respond to this email for us to continue our review.

I replied, linking to the page on their own website which contained the content I’d already updated, and pointed to their error message on that page which said “Google deactivated this book because it violated our content policy or there’s an issue with the copyright.”

I hoped they would be able to resolve this for me quickly. This time I only needed to wait three days for a response:

Thank you for contacting Google Play Books.

To assess your account, upload content. Once you’ve done this, respond to this email for us to continue our review.

Urgh! The exact same message. Word for word. I was stuck in a auto-responder loop!

I replied again. Nada. I’ve tried contacting them directly through the help links. But, again, have had no response other than that automated message. I’ve tried to start the process again, but I can’t delete the version I’ve already uploaded because it’s blocked by that “deactivated” message.

I’m out of ideas. If there is anybody who reads this newsletter has connections at Google or experience resolving these issues, I’d love to hear from you.

🍽️ Consume

I spent ten minutes this morning scrolling LinkedIn, before I caught myself. I realised that I hadn’t actually read anything. I’d skimmed a long list of posts, but would struggle to recall any of the details. Sound familiar? While the volume of content we consume has exploded, the quality has plummeted. We’re not reading, we’re grazing.

It’s difficult to out train a poor diet. We are, broadly speaking, what we eat.

Nutritionists break down the various things we consume into three macronutrients or “macros”:1

• Carbohydrates

• Proteins

• Fats

Most meals contain a combination of these things. What’s the ideal mix? That’s a complicated question. Even experts are divided. Their answer is very different depending on whether your goal is to lose, gain or maintain weight. But most agree, all three have a function.

Here is a decent summary:

During digestion, your body breaks down macronutrients into smaller parts that it uses for specific functions. Carbs are the main energy source, proteins help build and repair tissues, and fats insulate organs and make up cell membranes.

Just as nutritionists can analyse our food intake, we could apply the same logic to our information diet. We agree, broadly speaking, with what we consume.

In the past people consumed their daily news from the same places - if we extend the metaphor we could compare mainstream broadcast media to the traditional home cooked meat-and-three-veg dinner.

Now, for most of us, it’s predominantly social media. That’s complicated because each of us has our own “feed”, customised in part depending on who we choose to follow. But we can still break these down into “macros”.

Here is a quick tally of the first 100 posts I saw on my LinkedIn feed this morning:

• Posts - 4

• Comments - 7

• Ads - 15

• Likes - 74

It’s tempting to think that when we use social media we’re consuming content directly from people we choose to follow. But, as this shows, a smidge under three-quarters of the content I was fed is second hand. It’s the content equivalent of junk food.

Why does this matter? When we “like” a post, we’re making a split-second decision, often based on the headline or photo, or maybe a quick skim read. But the network amplifies that weak signal by pushing the content to exponentially more feeds. It’s unlikely to be the most valuable or accurate content. It’s not a thoughtfully prepared meal, it’s the brightly packaged sugary snack that catches our eye while we’re waiting at the check out counter.

This creates a vicious cycle: We’re all incentivised to cook up like-able content rather than valuable content. Complex ideas get boiled down into bite-sized absolutes. Nuanced discussions become polarised hot-takes. And we all get malnourished on a diet of content that’s been produced for maximum palatability. Like cattle being fattened for sale, we're served whatever keeps us consuming.

What would a nutritionists say if we consumed a diet of 4% protein, 7% complex carbs, 15% added sugars and 74% fat?

What can we do? I’m not sure I have a hopeful answer to this, unfortunately. Everybody who understands the health consequences of our deteriorating diets have been warning us all about this for decades. And yet, today Coca-Cola and Novo Nordisk are both $500B businesses - one fattening us up, the other slimming us down.

What’s the social media equivalent of Ozempic? Asking for a friend.

🏳️ Concede

This time last weekend I was in Wellington, recovering from the trauma of the second half of the All Blacks vs Springboks match at Sky Stadium. It was painful viewing.

This was the first match I’d watched live since seeing them lose to France in the opening game of the 2023 Rugby World Cup at Stade de France in Paris.

Maybe I’m the problem?

This is actually a question that has hung over me for decades…

Between 1998 and 2002 I attended six games and saw only two wins - one of those was the 54-7 drubbing of France at Athletic Park in 1999 (later that year France would knock us out of the World Cup with their stunning second-half comeback), and the other was the scratchy 40-29 win over Ireland in Dublin in 2001 (which was also Richie McCaw’s first match). One of my friends asked me to let him know which games I was planning to attend so he could place the appropriate counter-bets at the TAB!

In total I’ve attended 21 games in the professional era. I’ve seen 14 wins, one draw and six losses, meaning my success rate is notably below the team’s record, although I’m not sure you’d call that statistically significant.

There is a symmetry to those matches - starting in 1996 with the (at the time) biggest ever win: 43-6 over Australia at Athletic Park, and ending with the biggest ever defeat, 43-10 to South Africa.

I was interested to observe the response in the stadium to the result last weekend - and the second half in particular. It was quiet. Resigned, almost. We were sitting in front of a row of obnoxious Springbok supporters who tried their best to trigger a response, but they were mostly ignored.

I was at the same venue in 2000 when John Eales converted a last minute penalty to win the match for Australia. We were sitting not far from the spectators who started throwing things at the referee Jonathan Kaplan as he left the field.

This time around there was no disputed penalty. No forward pass. No Susie. Just an appreciation, I think, that we were comprehensively beaten by a better team. Again.

Since the 2023 Rugby World Cup final defeat, the All Blacks have lost six games. It’s simplistic analysis, but just looking at the points differentials from the first and second halves of these matches paints a stark picture:

As Sean Fitzpatrick drilled into those of us who have been fans for a while:

It’s a game of two halves!

Some of the responses have been much more emotional than statistical. For example, this from Gregor Paul in the NZ Herald:

Rugby is just one more part of New Zealand in decline: One more thing eroding while everyone deludes themselves otherwise. Delusions of grandeur are perhaps a geographic hazard of being a small, isolated island in the far reaches of the Pacific Island. It’s easy to develop a superiority complex when there are so few external factors to regularly make comparisons against.

This is maybe why New Zealand can tell the world it is clean and green while it repopulates its urban car fleet with double-cab utes and plans to drill into the ocean floor, or indeed anywhere it suspects a fossil fuel may be buried. This is why it can see hundreds of thousands of its best and brightest leave for Australia, but give visas to a few digital nomads and mad millionaires escaping what they think is an impending Armageddon and say New Zealand is the place the rest of the world wants to come to live.

Yikes! Are we ok?

At the Olympics last year the Black Ferns were the tonic, that helped us all quickly forget.

No such luck this year!

📥 Triage

Finally! After nearly 3 years, I’m excited to announce an updated version of Triage - first aid for your inbox.

You can download it on the App Store now.

What is Triage? The best description I’ve heard is: Tinder for Email.

Rather than drowning you in a long list of messages Triage presents unread messages as a stack of cards, and let’s you focus on these one at a time: swipe left to archive; swipe right to keep; tap to expand and reply or forward.

It’s fully customisable, so you can create your own workflow by assigning different actions to the left + right swipes - whatever works for you.

It’s still the first app I open every day.

A brief history… Triage 1.0 launched in 2013 under the Southgate Labs banner. John Gruber from Daring Fireball, who got an early preview when he spoke at Webstock that year, published this review:

Since I’ve been using it, I’m more caught up on my email than I have been in years.

That generated a huge amount of interest for the first weekend, at least.

Why did this update take so long? Partly because I’ve been working on other projects. But also, if I’m honest, because code that isn’t maintained has a half-life, and the longer I left it the more yak-shaving that was required each time I came back to it.

Minor fixes turned into archaeological expeditions. While it seemed a shame to let it wither, the work required was starting to feel insurmountable.

A few weeks ago I decided to use this as an experiment with Claude Code, and I was blown away by the results. Very quickly I was able to sort out the issues which had defeated me, get it building again, and implement the long list of improvements I wanted to make.

What’s new? It’s cliche for release notes on the App Store to reference “bug fixes and performance improvements” but in this case that’s actually 100% accurate - the previous version was sluggish, but this version is blistering by comparison. And, as the rest of the release notes say, “we’re setting the scene for some more substantial updates to come. Stay tuned!”

Try it today! If you struggle to keep on top of your inbox, or are just curious to try a different workflow, I hope you’ll take a look. I’d love to hear what you think.

Share

$2.8 Billion Later…

What did we learn from our innovation obsession?

Over the last decade, New Zealand threw $2.8 billion at "innovation".

Callaghan Innovation made $1.7 billion in grants to companies. Another $1 billion was spent to run the agency itself. The result? It’s difficult to see the impact. Productivity per capita went backwards. R&D investment as a percentage of GDP fell further behind the OECD average - from 1.25% in 2010 to just 1.47% in 2022, while the OECD average grew from 2.24% to 2.73%. The companies that succeeded? Many were the ones too busy executing to fill out grant applications.

The government finally pulled the plug on Callaghan Innovation earlier this year. But the myth that created it lives on - the same story politicians have been selling us for years. They promise that innovation drives economic growth, that funding fresh thinking will transform our economy, that being first matters more than being best. Every election cycle brings new promises about how government-backed innovation will finally unlock our prosperity.

In this excerpt from How To Be Wrong, I suggest the opposite might be closer to the truth. Using examples from Apple to Amazon, from Google to Xero, I show how the world's most successful companies weren't first at anything. They didn't out-innovate their competition. They out-executed them. Understanding that changes how we should think about building businesses.

While we spent so much money chasing the next big thing, we ignored the boring, unglamorous reality of what actually creates value: Getting thousands of little things right. Repeatedly. Consistently. Often without fanfare.

This isn't just contrarian thinking - it's $2.8 billion worth of evidence that our approach to building an ecosystem of successful technology companies is backwards.

What did we actually learn from this experiment? Read on, or press play to listen to the corresponding chapter from the audiobook above…

PS Know somebody who would like to read the whole book? Use the discount code LOVEUDAD this week to get free shipping, and go into the draw to win a special edition Ctrl-Z t-shirt - they’re available in all the sizes: Small, Medium, Large, and Dad-Bod!

One of the things that nearly everybody thinking about building a technology ecosystem seems to agree on is the need to be innovative. But innovation means different things to different people.

A definition of innovation I like is: “fresh thinking, applied to create value”. But, again, when we reach for anything more specific than that it is elusive. The OECD uses the Oslo Manual, which has a particular definition of innovation used to ensure consistency of government data collection. One of their criteria for being an innovative company is having more than five FTEs. The reasons why are not explained, but perhaps it’s a reflection of a steady-state mindset (that is, small companies are likely to stay small). Either way, that’s bad news for nearly every early-stage company who likes to consider themselves “innovative”. You’re not counted.

Often we get stuck on the “fresh thinking” part of the definition and never get to the “creating value” part. That is to say we conflate innovation and invention. Remember, often the person who invents something and the person who works out what that thing is for are different people. An inventor has an idea. A company executes an idea.

An innovative idea is like an algorithm that has been written on a whiteboard but never run on an actual computer. What matters is how well it executes.1 The likelihood that any non-trivial software works first time as intended is infinitesimally small. That is the art of software development. And, if we extend the metaphor, the challenge of creating a successful business. We need to get beyond ideas.

Ideas require getting one new thing right. Execution requires getting thousands of little things right, and repeating them over and over. Ideas appear in the “a-ha!” moment of inspiration, when the lightbulb goes on. Execution is perspiration – putting our head down and doing the mahi. Ideas are new and shiny and exciting and recognised with awards. Execution is exhausting and mostly far away from the spotlight (until a company that has executed consistently is sold and then everybody scratches their head and wonders how a business that didn’t seem very innovative managed to become so valuable). Ideas are the denominator. Execution is the numerator. Ideas are R&D (usually more R than D) and so are often eligible for generous government support. Execution is operating expenses, and so needs to be funded entirely by investors or customers. Ideas are smiling people in stock photos wearing white lab coats. Execution is rent and payroll, distribution, logistics and sales. Ideas require us to dream big. Execution insists that we find the right size – meaning, the scale of the business that can be maintained. Ideas ask “can we make it?” Execution asks “should we?” Ideas are about being the first. Execution is about being the best.

What is more important: being innovative or executing well? The noise about innovation is mostly focused on “new and different”, but what about those companies that do “familiar but better” – take an existing idea and deliver it in a way that means many more people will be enticed to buy it? Look at the most successful companies. We see that pattern repeated again and again:

• Apple was not the first to produce any of the products they are most famous for (they didn’t invent desktop or laptop computers, they didn’t invent portable music players, they didn’t invent smartphones, and so on).

• Google was not the first search engine.

• Facebook was not the first social network.

• Amazon was not the first online store.

• Netflix was not the first streaming video service (in fact, in the beginning Netflix wasn’t even itself a streaming video service, but that’s another story).

• Zoom was not the first video conferencing tool, by a long way.

The list is long. The thing all those companies have in common is that they were each massively successful in convincing people to use and buy the products and services they make and sell. They were often the first to work out what these things were _really_ for, and as a result how a wider audience could be made to desire those things. And that underpins their financial success. Now they all spend millions of dollars each year on R&D, but that’s not necessarily what got them there.

Sometimes the innovative thing isn’t the product at all, but the distribution channel or revenue model. For example, Xero took the existing and familiar (and, let’s be honest, pretty low-tech) category of accounting software and built a massive company by providing free tools (and a lot of love) to accountants; using them to reach a large potential market of small business customers; and charging a small monthly subscription rather than an up-front fee, in the process proving there was still a lot of value to be captured in that existing accounting software category.

So, is product innovation just a distraction? No, not at all. There are startups that have invented something and executed well, and in the process created a big success. In other words, occasionally the people who invent things and the people who work out what they are for are the same people. But the invention isn’t the multiplier in that equation, it’s table stakes. It’s a sometimes ingredient, not a leading indicator or prerequisite. A great product is important, but great product distribution is much more important. This means it’s not enough to say that a company is innovative. Or that it isn’t. That doesn’t tell us much.

One of the reasons why innovation is given so much prominence is that it’s easier to measure. It’s simple to calculate the amount spent on R&D. It’s much more difficult to determine the return on investment (ROI). Plus, the amount invested is immediately obvious whereas we often have to wait a long time to understand the potential returns, and the milestones along the way can be subtle if we don’t know what we’re looking for.

Inventors are a distraction. When we try to pick individual companies that are potential future winners, what we’re really looking for is something much less common, much more difficult and unfortunately much more boring: who can execute well, and do it repeatedly? Maybe when we named a government agency after Sir Paul we should have called it Callaghan Execution.2

How To Be Wrong is available now in paperback, audiobook or ebook.
Check out the website for links to all of the places you can find it - in store and online.

Enjoy!

Header image by Alan Bowman on Unsplash

🎧 Listen

When How To Be Wrong was published earlier in the year one of the things that surprised me was how many people asked when the audiobook version would be available. It turns out, many more people listen to books than I had appreciated.

I did initially intend to release the audiobook at the same time as the paperback and ebook versions. It was 80% recorded in January. But I didn’t quite get it finished and subsequently it got bumped behind other priorities.

So, I’m excited to finally have this completed. It’s available now on most platforms where you would normally get your audiobooks, including Spotify, Google Play and Apple Books.

If you have a Spotify Premium subscription it’s available to you at no extra charge. And, it should hopefully appear on Audible very shortly - so look out for it there too.

I’d love to hear what you think once you’ve had a listen.

The full book is 10 hours 57 minutes long (I guess shorter if you prefer 1.5x speed). But, if you want something a bit more compressed, here are three podcasts I’ve recently been on where I’ve had the opportunity to talk about the book and some of the themes in it:

Business is Boring with Simon Pound

This is my second appearance on this podcast.1 This time, Simon asked me to share in advance some of the questions that I hadn’t been asked yet in all of the media interviews I’ve done about the book. As a result the signal-to-noise ratio was high. Even if you’ve listened to other podcasts I’ve recorded recently you’ll likely find something new in this - for example, in the second half we go deep on the lessons from the Vend implosion in 2015, when Simon himself had a front-row seat as an employee at the time.

Between Two Beers with Steven Holloway & Seamus Marten

This was a rare treat to be on a podcast with such a large audience. Steven and Seamus are famous for being the best prepared long-form interviewers in the land, and this episode was no exception - right from the very first question.

I’ve had some great feedback about this one and I hope you find it entertaining, useful and interesting. If you have friends who are not normally interested in startup content, but who you think might be curious, this would be a great link to share with them.

Beyond the Surface with Noa Woolloff

In this conversation we go deep on “tall poppy syndrome” and why I think that we often use that as an excuse in NZ, my approach to picking which companies to invest in (and, more importantly, which companies not to invest in) and many of the persistent myths that hold us back from creating more high-growth startups.

Enjoy!

🎙️Record

I was recently invited to join a zoom call with a large group of attendees. I was immediately confronted with an announcement: “This call is being recorded”. It wasn't a question or request for consent—it was a fait accompli. I instinctively turned my camera off and found myself less inclined to participate in the conversation.

Sure enough, and as I had maybe subconsciously feared, shortly after the call snippets surfaced on social media as content to promote the services offered by the people who had coordinated the group. That felt pretty gross.

This isn’t unusual. I’ve noticed that every meeting I’m in now has a few extra attendees—AI agents, who listen to everything that’s said, and afterwards condense that into a transcript, summary and ideally some follow-up actions. Sometimes they are obvious. Sometimes they are lurking in the background.

These tools have started to get remarkably accurate—in just the last few days I’ve gotten meeting summaries from Google, Zoom and even Slack which accurately captured the key points from the just completed conversations. I’ve even found myself occasionally talking directly to the AI agents on the call—for example, to emphasise an action that I want to be recorded.

On the surface it seems amazing and a little bit magical. However, I wonder: at what cost? The implications ripple across multiple dimensions of our working lives:

What is the impact of every conversation being recorded?

Does it change how the humans on the call engage? (And, if it does, how would you know?)

Are the promised productivity improvements real, or do we all now drown in a swamp of transcripts from calls that we’ll never have the time to revisit?

What happens when AI summaries become the official record of meetings? When disputes arise about what was actually said or agreed upon, do we defer to the AI's interpretation over human memory and notes? How do we handle situations where the AI misses crucial context, tone, or subtext that fundamentally changes the meaning of what was discussed? Formal boards use a deliberate process of reviewing and adopting minutes from previous meetings to address exactly these risks and concerns, but that's hardly practical for every daily standup or client call.

Should we worry about the privacy implications of uploading all of the information to the cloud, and having so many of our previously confidential conversations added to the soup of data available to large language models? (Or do we naively trust that the companies managing these tools will respect our privacy and keep this information separate—despite their clear financial and competitive incentives to do otherwise?)

What is the appropriate etiquette around requesting permission to record conversations? Does it need to be explicit and automatically announced to everybody at the start by the AI agent (noting how jarring that can be, especially when you’re talking with people you don’t know well or are meeting for the first time)? Or should that be opt-in and dealt with by humans, and only enabled once everybody has consented? Or are there some situations where you’d record the conversation without explicit permission?2

Do you use AI agents on every call, or are you selective about when they are deployed - perhaps keeping some conversations focussed on building relationships with the real humans you’re talking to?

Finally, how should AI agents like this be regulated - or should it be free-for-all? And, even if they should be regulated, how would that actually work? Given that these tools are essentially creating permanent records of workplace conversations that could be used in disputes, searched by employers, or potentially accessed by third parties, do we need explicit laws about consent, data retention, and the rights of meeting participants? To add even more complexity - what happens when these recordings involve attendees in different countries with different privacy laws?

It’s complicated!

I’ve written in the past about the joy of running naked - no, not without clothes, but without a smart watch and associated data logging to manage my pace. On the rare occasions when I do it (usually because I forgot to charge my watch), it forces me to pay much closer attention to how I'm feeling in my body, to manage my energy based on internal cues rather than algorithmic feedback, and ultimately makes me more present in the moment.

I wonder if before too long we’ll savour conversations which are not recorded and boiled down probabilistically by robots in exactly the same way?

🐴 Persist

Here is a fun story…

Haru Urara is a retired Japanese racehorse. During her 6-year career she competed in 113 races and never won a single race.

In fact, her repeated failures actually made her a minor celebrity in Japan. As described on Wikipedia, she became “a symbol of perseverance and tenacity”:

At a race held on March 22, 2004, during the peak of her popularity, more than 13,000 spectators packed the Kochi Racetrack, 3,000 of whom had gathered outside before the gates opened, forcing the track to open 30 minutes earlier than scheduled. Some people waited in line for up to five hours to buy tickets at the "Haru Urara Commemorative Ticket Booth", which was established especially for the event. Fans bet a grand total of ¥121,751,200 [NZ$1.3m] on a Haru Urara victory, an impressive sum, particularly for a horse that had not won once in more than 100 attempts. Despite being ridden by Japan's premier jockey, Yutaka Take, Haru Urara earned her 106th consecutive loss, placing 10th among 11 horses running.

This year she has enjoyed further notoriety after she was used as the inspiration for a character in an obscure anime game called Umamusume: Pretty Derby. As you’d expect the character has the worst possible stats. But that doesn’t deter fans. Haru Urara’s owners opportunistically asked people to donate towards high-quality “refrigerated, highly palatable ryegrass” for the real-life horse (at the slightly outrageous price of ¥6,000 a pop - or around NZ$70 each). So many people tried to donate that it crashed the payment site.

Sometimes losing is just winning in disguise.

Photo by Yassine Khalfalli on Unsplash

A rare Monday edition of Top Three…

🛑 Stop

A huge thank you to everybody who has purchased a copy of How To Be Wrong. So far we’ve shipped nearly 1,500 copies in various formats. And hopefully more to come.

If you’d like to join them you can buy your copy in all of these places:

• Direct from the official Ctrl-Z store - get one, or buy in bulk and save 10-20%

• Online at Book Hero - for next day delivery

In-store at various bookshops around the country:1

• Auckland: Unity Books, The Open Book & Time Out, etc

• Hamilton: Poppies Books

• Hawkes Bay: Wardini Books

• Wellington: Unity Books

• Christchurch: Scorpio Books & University Bookstore

• Timaru: Timaru Booksellers

Also at Paper Plus stores in Orewa, Eastridge, Gisborne & Porirua (and hopefully others soon).

You can also get a digital copy anywhere you buy ebooks:

• Amazon Kindle

• Apple Books

• Kobo

I’m enjoying reading some of the responses - thanks to everybody who has taken the time to share those so far. I’m interested to hear which bits you think I got right and which bits I got … wrong! Please keep them coming.2

One of the patterns I describe in the book is what I call the "third user", which I’ve seen repeated by a number of different teams over the years when testing a new product: When the first user encounters a problem, we often dismiss it as user error. When a second user experiences the same issue, we might still consider it a coincidence. But when a third user has the identical problem, we realise this isn't about the users at all—it's a flaw in our product that needs fixing.

Yesterday I heard from a third customer who has purchased a copy of the book and discovered missing pages—or more accurately blank pages where there should have been words. I obviously need to follow my own advice and not dismiss this problem.

The book was printed by Bluestar in New Zealand. That’s a bit unusual these days—most books sold here are printed in Asia. But we made a conscious decision to keep it closer to home so we could have more control over the quality and timeframes.

At this stage I only have these three examples, and no other evidence to suggest it’s not a rare exception. But if there are others like this out there I’d like to know so I can send replacements. Please check your copy (in all three examples the missing pages were between page 50 and page 90). And get in touch if yours is the same, so I can “put it right”.3

👥 Collaborate

The irony is not lost on me: Writing a book that bangs on about the “myth of the lone genius”—how we tend to credit a single person for any successful startup, when it always takes a team—and then putting only my own name on the cover.

At the launch in Auckland I managed to get this photo with the other people who contributed directly to making this book:

From left to right:

• Anna - who was not only the proofreader, ensuring that all t’s were crossed and i’s dotted, but also concierged the book through the final stages;

• Toby - who edited the book, and took what was literally called the “flabby” draft and helped me craft it into a readable story;

• Duncan - who wrote the introduction, where he described the important role he played in convincing me to write it in the first place;

• Imogen - who did all of the design work, inside and out, and even tolerated me fiddling about directly in the InDesign files at one point while trying to sort out italics;

• Claire - last but not least, who in her previous role was one of the first believers and who subsequently endured multiple early drafts and delicately helped me restructure those into something coherent.

I’m responsible for all of the bits you don’t like. The credit for all that’s left after that is shared with all of them.

🎧 Listen

They say you should never meet your heroes. But I disagree. I loved this chat with Joel Little at the Blackbird Sunrise conference in Wellington last year and I’m grateful that it was recorded. It’s been released this week as part of his Diaspora.nz podcast series, so now we can share it with everybody who wasn’t in the room.

Check it out on Spotify, Apple Podcasts, YouTube or wherever you get your podcasts.

There is also a bootleg video available, if you prefer to listen with your eyes. 🙈

PS A number of people have asked about the audiobook version of How To Be Wrong. I’m learning there is nothing more painful than listening to yourself try and read your own words, while aiming to have it seem as natural as a conversation over a cup of tea. But I will get there soon hopefully.

Stand by for more information about release dates etc…

Header Photo: Breaking the Ice, Ice, Baby, Antarctica 2016, by Emily Simpson

Share

Photo by Victoria Vincent, Cover design by Imogen Greenfield

🚚 Publish

What do French sculptor Camille Claudel, a rubber chicken, Edmund Hillary, the Ship of Theseus, the Melbourne Cricket Ground, the Slough Trading Estate, BMX Olympic silver medalist Sarah Walker, Father Christmas and broadcaster Corin Dann all have in common?

To find out, you’ll need to read my book How To Be Wrong: A crash course in startup success – which officially launched this week!

The process of writing, editing, designing, printing, distributing and launching this has had quite a few unplanned twists and turns. I’ve ended up doing much more of it myself than anticipated at the outset, but it is a genuine delight to get to this point.

It always takes a team

I initially committed to this project with a traditional publisher - I thought that would mean I could somehow avoid doing many of the hard bits. I should have read my own content. That didn’t work out, but it got me started. In the end I hired Toby to edit, Imogen to design, and Anna to proof directly, and they helped me make it great. Having more control throughout the process let me do it my own way, and (to some extent) on my own timetable, which probably suited me better in the end.

At one particularly discouraging moment, I spoke with a distributer - somebody whose job is to take books from publishers and promote them to book stores (and having convinced them, deliver the physical books to them). He couldn’t really have been less interested. He recommended I print no more than 500 copies and said he thought he could maybe get them into book stores later in the year.

This week we’ve shipped over 1,000 copies, including the “First 500” which have been individually signed and numbered. Most of those were purchased directly through our online store. That’s definitely an approach that doesn’t scale, but I’m glad we took the time to do that for the people willing to purchase even before the books were printed.

I took this photo yesterday at Unity Books in Wellington. There were two copies left in the pile, and it was flattering to see the titles it was sitting alongside. Shortly after that I got an email from them ordering some more. We’ll ship those tomorrow.1

As always, I say “we” - this book wouldn’t exist without the support of many others. My name is on the cover, but all of the really important people are listed in the Acknowledgements, so I hope you’ll take the time to read that too, when you get right to the end.

Untold stories

All the time I was working on these businesses I’ve written about I was always keen to soak up whatever stories and patterns I could find to apply to the problems I was trying to solve. But, they were all stories of people and companies in far-away places. There isn’t really a category for this book - there are just not many New Zealand business books published. We’ve proven we can start and build these sort of businesses here. We just need a lot more people to realise that is possible and do the same. Hopefully, in a small way, by publishing this I can inspire and encourage those who are brave enough to try.

I wrote this book to be read, and I can't wait to hear what you think. If you've already gotten your copy, please get in touch with your feedback.

Whether you're after Pure Fun or Deep Fun (you'll understand this reference once you read the book), I hope you'll find value in these stories and lessons. Enjoy!

If you don’t have a copy yet, you can buy one today. We are sell the paperback directly, with options to also buy in bulk and save 10-20%. Or you can buy from BookHero.co.nz for next day delivery.

The ebook version is available in all the places you normally buy books for your device: Amazon Kindle; or Apple Books; or Rakuten Kobo; and many more.

Or, if you’d like to pick up a copy in store you should find them at The Open Book in Auckland; at Unity Books in Auckland and Wellington; at Poppies Books in Hamilton or at Scorpio Books in Christchurch and many more.

If you sell books in your store please get in touch and we’d be happy to send you some copies. If anybody reading this knows the book buyers at Paper Plus or Relay - I’d love to have it for sale in those stores too. Tell them what they are missing out on!

📢 Shout

The launch has involved an omnibus of media interviews and podcast recordings. After spending so much time writing, reading and re-writing the words, it’s been a pivot to be suddenly talking out loud about it all.

Several people have questioned the hypocrisy of writing a book containing a chapter all about “The quiet ones” while at the same time seemingly never shutting up about it.

Well… as I say:

It’s not enough to just have a good idea for a startup. To create value we have to execute it. Every founder needs to confront a fundamental question: how will you overcome obscurity? While many worry about others stealing their ideas in the early stages, a much more urgent concern is confronting the reality that most people don’t care or (worse) don’t even know about us at all. Every day startups die of starvation, due to lack of customers. Very few ever die of drowning from a deluge of customers they can’t cope with.

That’s a quote about my experience starting and trying to build up Flathunt - my first startup prior to working on Trade Me (it was eventually acquired by Trade Me and incorporated into Trade Me Property when we launched that vertical - which is how I came to be a Trade Me shareholder). But it applies equally to any new thing that we do - including publishing a book.

Here is the full media playlist from the last couple of weeks - find somewhere comfy and read, listen to or watch them all:2

• Radio NZ Saturday Morning with Susie Fergusson
  How to be Wrong: a crash course in startup success

• TVNZ Q+A with Guyon Espiner
  Why Govt shouldn't be picking start-up winners

• NewstalkZB Breakfast with Mike Hosking
  "No shortage of capital": How to build successful businesses in NZ?

• NZ Herald with Chris Keall ($)
  Rowan Simpson’s Three ways to be wrong

• National Business Review with Simon Shepherd ($)
  Rowan Simpson dispels some startup myths

• The Listener by Peter Griffin ($)
  Are we too focused on fostering innovation in science?

• Spinoff: When the facts change podcast with Bernard Hickey
  The startup savant behind Trade Me and Xero

• BusinessDesk: The business of tech podcast with Peter Griffin
  Startups, being wrong, and near-death experiences

• Everyday Investor podcast with Darcy Ungaro
  From Wrong, To Rich

• NZ Tech podcast with Paul Spain
  How to be wrong: Nagivating Tech Startups

• Bountiful podcast with Sian Simpson
  How to be wrong and still succeed

• Caffeine Daily by Finn Hogan
  Veteran of Trade Me, Xero wants less startup 'theatre' and more startups

It doesn’t end here. Next week I’ll be in Adelaide speaking with Matt Allen at _SOUTHSTART. If you’re in the vicinity it’s not too late to grab a ticket - get in touch and I can send you a discount code.

🚧 Build

It is easy and quick, often instantaneous, to demolish things. It is hard and slow work to build new things, and often even harder and slower work to improve existing things.

In many of the media interviews I linked above I’ve been asked about recent government policy decisions - especially the decision to disestablish Callaghan Innovation. My experience working on and investing in Trade Me, Xero, Vend, Timely and other ventures gives me a rare perspective. I've seen the same patterns repeat and the same mistakes made, often because those implementing policies have never been in those positions themselves.

At the book launch events this week Toby asked me what advice I would give the Prime Minister and others who are (hopefully) currently thinking about what comes next.

What is the one thing they could do that would finally unlock the potential of the ecosystem?

My answer to that is simple: it’s the wrong question!

I would recommend these people read the whole book, obviously, but specifically the chapter towards the end called “The silver bullet”. Perhaps this might convince them that this search for a single thing - often an idea that has worked in a completely different context somewhere else in the world - is a mirage, and that anybody who is promoting those sort of magical short-cut solutions is probably trying to sell something that will mostly extract value.

We have been trying to build a startup ecosystem for more than a decade. We have thrown a large amount of spaghetti at the wall.3 But how much has stuck? We don’t really know. And, for me, that’s the most depressing part. We just hoped that spending all of that money would produce good outcomes, but never really bothered to test those assumptions or measure the results en route. And, then when it was eventually obvious that it wasn’t working, we were not honest about that, perhaps hoping that the results would turn for the better before they were exposed. As I describe in the book, that’s the worst way to be wrong.

After the announcement a few weeks ago I went back and added up the numbers:

As the graph shows, of the $2.8 billion various governments have spent on Callaghan Innovation since it was established in 2013, only $1.7 billion was distributed as grants. Meanwhile, over $1.1 billion went to employee and supplier costs - essentially the overhead of running Callaghan Innovation itself. Even putting aside the huge headline amounts, even that ratio of operating costs vs funds distributed to companies is eye watering. What do we have to show for it?

The most depressing bit about this for me is that it didn’t work. We desperately needed it to work. If the answer to “what now?” involves asking all the same people what they want to try next, then we should expect more of the same outcomes.

Here is the paradox. When I’m asked about the ecosystem I say that it is today in better shape than it has ever been. For me the ecosystem isn’t the government organisations and accelerator programmes and shared working spaces and venture funds (or funds-of-funds) or angel groups or business awards or innovation showcases … or countless other derivative things. It’s the crazy people who start these companies and the much larger group of less-well-known-but-still-vital people who work with them to grow the businesses and invest directly in them. That’s the hard grind.

Nearly everybody who talks about wanting a larger, more vibrant and more successful ecosystem misses this. They aspire to create multiple companies at once but don’t really know what it takes to create one. The companies that are created by this system are exactly what we should expect: fragile, precarious, and underwhelming.

While others were debating about how to build an ecosystem, and waiting for somebody else to do those hard bits, those of us who just got on and did it at companies like Trade Me, Xero, Vend, Timely and many others realised: the best way to grow an ecosystem is to create one great company. If enough of us did that we would have a much larger ecosystem.

So when we think about what it will take to unlock growth at this level, these are the questions they should be asking instead:

1. Who are the specific people we are trying to help?

2. What’s currently holding them back?

3. How can we remove that constraint?

4. How will we know it’s working?

I hope we don’t look back in another ten years and say: those are good questions, perhaps we really should have asked them at the time.

(╯°□°)╯︵ ┻━┻

🎁 Blurb

One of the most awkward steps in writing and publishing a book is sending the completed manuscript to a bunch of people whose opinion you respect and blatantly asking for praise. But not just praise… succinct praise. The sort of praise that can be shared on a website or even printed on the back cover of the book in one or maybe two lines.1

It’s a hard thing to ask for. But even harder to write a good one. I’m grateful to all of the people who have made the effort to help me promote How To Be Wrong. Honestly, reading these as they have come in has been a delight.

Over the next week I’m going to be sharing some of these lovely reviews I’ve received on social media, but here are three that I especially like to start with - hopefully demonstrating that this is not just a book for startup nerds:

Courtney Johnston

Tumu Whakarae | Chief Executive, Te Papa

“Why would a museum director be drawn to a book about tech start-ups? Simple: because it starts ‘History isn't fact. It's narrative’? Here's an author who understands the power of mythology, and shows how it's at play in our popular understanding of tech start-ups - their births, triumphs, and failures.”

Sir Bill English

Former Prime Minister of New Zealand

“A remarkably humble honest tale of taking risk, how to roll with it, how to beat it and why it’s worth the effort. I enjoyed his discussion of government innovation policy as he strips down the borrowed hype of tech policy.”

Malindi MacLean

Tumuaki/CEO, Outward Bound

“A masterclass in experiential learning - Rowan's insights about focussing on the rewards of Deep Fun rather than the sugar rush of Shallow Fun will change how you think about starting and building organisations. I loved it.”

Thank you to all three of them, and all of the others who took the time to review. All of the blurbs are on the website now, and there is an extra page with longer-form versions.

If you’re sufficiently inspired, pre-sales are available now.

Speaking of blurbs, here is some news from Sean Manning, just too late to be useful:2

While there has never been a formal mandatory policy in the eight years I’ve been with the Simon & Schuster imprint, it has been tacitly expected that authors—with the help of their agents and editors—do everything in their power to obtain blurbs to use on their book cover and in promotional material. I have always found this so weird.

I mean, what I just said, right?

I’ve decided that beginning in 2025, the Simon & Schuster flagship imprint will no longer require authors to obtain blurbs for their books. This only applies to Simon & Schuster’s flagship imprint, and this isn’t to say that we will outright refuse to include blurbs on our book covers and in promotional materials. If a writer reads a book because they want to (not because they feel beholden) and comes away so moved by it that they can’t resist offering an endorsement, we will be all too happy to put it to use.

Ok?

But there will no longer be an excessive amount of time spent on blurb outreach. There will no longer be the fear that if we don’t secure the right blurbs or enough blurbs or even any blurbs, it will jeopardize a book’s chances for commercial and critical success.

🙄

💬 Chat

I first met Siân Simpson (no relation) in 2011, when she was running the Kiwi Landing Pad in San Francisco. For a couple of years I was a frequent visitor with Vaughan from Vend, as we tried to grow the team in North America and raise the capital to fund that. She always welcomed us in, but more importantly pointed us out at the impressive network of interesting people she had built in the city.

Fast forward a dozen years (yikes!) and I was privileged to spent a bit of time with her at the end of last year at her new base near Cromwell … a million miles away from the buzz of San Francisco both physically and mentally. It’s a beautiful spot, and as always we enjoyed talking about an eclectic range of topics. The difference: unlike all of our previous conversations this one was recorded, for an episode of the current season of Sian’s Bountifull Podcast, so you can all eavesdrop.

Even if you’re not into startups and business there are some different questions in this which I think you’ll find interesting. I’m told this episode will be colloquially known as "The Simpsons". Enjoy!

Watch/Listen Now

• YouTube

• Spotify

• Apple Podcasts

PS if you’re not familiar with the Bountifull Podcast, it’s a bit different. I’d recommend checking out the Season One episodes with Paul Conway, Julie Fry and Sir Richard Taylor for a start. And if you’re flying on AirNZ you can find it under “New Releases” on the TV tab of the seat-back entertainment.

🎸 Listen

At least one keen-eyed reviewer noted some of the subtle referenced to popular music that are sprinkled through the book. So, as further proof that this is not a normal book, I’m excited to share the official How To Be Wrong Spotify playlist:

I have a special turkey-themed prize for the first Top Three subscriber who can locate and describe back to me the exact places in the book that reference each of these thirteen songs. Enjoy!

All orders placed before Wednesday will ship this coming Friday or Saturday - if you want to be one of the first to get your hands on a copy.

Photo by freestocks on Unsplash

1

As a treat, for those who read the footnotes, here is the actual back cover of the book:

In case you missed the news, pre-sales for my book How To Be Wrong are now underway. If you want to grab a copy of the special “First 500” edition, which are individually signed and numbered, please get in quick - nearly half of them are already sold.

There are also discounted bundles for anybody purchasing multiple copies, which include free merchandise, like Ctrl-Z t-shirts and caps.

If you prefer to buy from a real bookstore, it’s also available to pre-order through my friends at BookHero.co.nz.

I’m only printing a limited number, and once they are sold they are sold. Don’t delay.

🧨 Reset

This opinion piece was first published in The Spinoff on 29 January.

In 2011 the physicist Sir Paul Callaghan, one of New Zealand’s most distinguished scientists and a passionate advocate for economic transformation, gave a landmark speech he called “Sustainable economic growth for New Zealand: An optimistic myth-busting approach”. Many of his observations have become often-repeated mantras of New Zealand’s technology sector, including “be the place where talent wants to live”, “we better be prepared to be good at some pretty weird stuff” (anticipating correctly that the most successful companies would operate in obscure and difficult-to-predict niches) and that “just 100 inspired entrepreneurs could turn this country around”.

Tragically, Callaghan died of cancer just months later. However, his legacy lived on in a new Crown entity called Callaghan Innovation, established in 2013 with the vague responsibility for “making New Zealand business more innovative” through grants, technical services and business support. The intention was noble – to help transform New Zealand from a country dependent on primary industries to one driven by technology and innovation.

Fast forward to last week: prime minister Christopher Luxon’s state of the nation speech echoed many of those same ideas. At the end, he announced that Callaghan Innovation would be disestablished, with its “most important functions” reassigned elsewhere in the public sector.

For those of us who have built and invested in successful technology companies in New Zealand, this decision is overdue. Despite good intentions, the agency became bogged down in administering grants and subsidies rather than driving real innovation. Under pressure to demonstrate value, it spread its resources thinly across a huge range of activities – from managing R&D tax credits to funding startup accelerator programmes. But there was never any clear evidence that any of these initiatives actually helped create more successful companies.

There’s no question that improving our productivity and growing the economy beyond agriculture and tourism is key to our future prosperity. We’ve understood this for a long time. However, the ambition to be more innovative has never been our problem. The challenge we’ve struggled with all these years is actually doing it in any measurable way. Perhaps we missed a trick? Rather than Callaghan Innovation we should have called it Callaghan Execution.

Having worked with and invested in companies like Trade Me, Xero, Vend and Timely over the past two decades, I’ve seen firsthand what drives growth. The key question is always: What’s the constraint? What’s actually holding us back? As we consider the legacy of Callaghan Innovation, and more importantly think about what will replace it, let’s ask the same question of our economy as a whole.

Here’s an obvious answer: our national obsession with investing in real estate, rather than investing in much more uncertain but ultimately more rewarding productive businesses. Spending more and more borrowed money to buy little bits of our country off each other is a zero-sum game. We might think it’s how we “get ahead”, but it’s actually dragging us down.

Luxon himself exemplifies this – he’s often described as a “businessman”, but his investments after achieving financial success were rental properties, sold for profit years later. What kind of business is that? What value does it create?

Here’s another: our expectation that the government will underwrite business risk, while private shareholders capture the rewards. Founders and executives at early-stage companies spend countless hours pursuing grants, subsidies, R&D credits and development loans – many of which end up written off rather than repaid. Venture funds and “angel investors” expect government co-investment without hard questions. The government venture fund has been tilting at windmills for more than 20 years, with little to show for hundreds of millions invested.

Ask not what you can do for your country, but what your country can do for you, I suppose?

Here’s a third: our dismal savings record, which leaves us constantly looking overseas for investment. In the same speech last week the prime minister announced the creation of Invest New Zealand to “roll out the welcome mat” for foreign investors. But lack of a welcome mat hasn’t constrained our best startups from attracting international investment – companies like Xero, Vend and Rocket Lab have all raised millions offshore. Many of the most promising companies in the next wave of startups are already backed by Australian-based venture capital funds – including Tracksuit, Halter and OpenStar.

The big difference between New Zealand and Australia in this respect is that in recent years large superannuation funds in Australia have become significant investors in their venture capital sector, helping them to scale. In New Zealand we’re at least a generation behind because we’ve been so tentative on compulsory superannuation. If we had to pick one policy prescription to help us close the gap with Australia, matching their settings on this would be a great first step. Is there any politician here brave enough to propose that?

I applaud the prime minister for turning our attention back to these fundamental questions about New Zealand’s economic future. But aspiration alone is insufficient. What are we actually going to do differently?

Rather than simply targeting “more startups, more IPOs, more inbound foreign investment” – which are inputs – we should understand and measure how startup companies contribute to the broader economy. This means tracking metrics like how much they pay their employees and how much of the capital they raise is spent locally. After all, we don’t tax capital gains in New Zealand, but we do tax operating expenses through PAYE and GST. The only companies that will drive our future prosperity are those that create high-paying local jobs and sustainable growth.

The end of Callaghan Innovation is a necessary reset. While it’s difficult news for those affected, hopefully many of them can find roles working directly for our fastest-growing companies, which are constantly seeking skilled workers, rather than supporting them indirectly. That’s one of our biggest constraints – we need more New Zealanders choosing to work on and invest directly in these businesses. Perhaps then we can move beyond talking about the theoretical benefits of innovation to actually delivering it.

🧱 Build

After the reset comes the rebuild.

In 2006 we visited Berlin and took this obligatory tourist photo at the remains of the Berlin Wall. Little did I realise at the time the significance of the Erich Fried quote I was standing in front of:

Wer will, dass die Welt so bleibt, wie sie ist, der will nicht, dass sie bleibt.

Which translates to:

He who wants the world to remain as it is does not want it to remain at all.

When something isn’t working the correct decision is to stop. That is important, but insufficient. Let’s not waste much more energy debating what got us here. The way more interesting question is what are we going to try next?

Whatever it is, let’s agree to start with four questions at the core of the design:

1. Who does this help?

2. What constraints do they have?

3. How do we hope to reduce or remove those constraints for them?

4. How will we show that it’s working?

Then, maybe, we give ourselves a chance to build something that actually works.

📚 Summarise

Thanks to everybody who followed the Top Three Summer Series during December and January. If you missed them and you’re short of time to read, I’ve asked two imaginary podcast friends to summarise the whole series:

Up to you to judge if this is generated slop or useful content. I think they got it mostly right. But, I did laugh out loud at this line:

AI is being slapped onto everything these days!

The full series is here:

1. Aim … Fire 8th December

2. Feedback Loops 15th December

3. A sky full of stars 22nd December

4. Invisible technology 5th January

5. Getting ahead 12th January

6. Turpentine 19th January

Enjoy.

PS Last weekend the NZ Herald published this excerpt chapter, about the sale of Trade Me to Fairfax Media for $750 million in 2006.

Share

Welcome to the 2024/25 Top Three Summer Series. This is the sixth and final post, leading up to the launch of my new book How To Be Wrong in just a few weeks time. Links to the full series are below. Enjoy!

1. Aim … Fire 8th December

2. Feedback Loops 15th December

3. A sky full of stars 22nd December

4. Invisible technology 5th January

5. Getting ahead 12th January

6. Turpentine 19th January

🧨 T-minus 3, 2, 1

Today is a milestone: pre-sales for my new book are available now.

I've created a special "First 500" limited edition of the book, which will be signed by me and individually numbered. I'll even add a custom message on request.

As the name suggests, this promotion is limited to the first 500 copies, so get in quick. Once they're gone they're gone.

For the pre-sales, we've also created discounted bundles, to make it easy for you to buy a copy for everybody on your team:

• Buy 2x copies, save 10%

• Buy 3x copies, save 10%, with free t-shirt

• Buy 5x copies, save 10%, with free Ctrl-Z cap

• Buy 10x copies, save 20%

Mix-and-match the bundles to get the exact number of copies you need and then decide who gets the swag!1

Maybe I've lost my mind, but there are also some larger bundles which include some never-to-be-repeated money (usually) can't buy offers:

• Buy 50x copies, and I'll join a zoom call

• Buy 100x copies, and I'll come and visit in person

These offers are strictly limited and are a chance to get some of my time to spend with you or your team. If you're interested, don't delay.

⏮️ Rewind

In Part 3 of the Summer Series, I talked about the conversation I had, on stage, with Joel Little at the Sunrise festival in Wellington last year.

I love this conversation format, and recommend it to all conference organisers. Find two interested and interesting people and let them chat. It’s nearly always better than a prepared keynote, or a forced interview.

For example, and as a bit of nostalgia, here’s a throwback to a similar conversation I had with Trade Me founder Sam Morgan at Webstock way back in 2008. The video is a bit janky, and deteriorates later in the recording (frankly, given how old it is, I’m a bit surprised it’s in colour!), but I think you might still enjoy it.

From 4-6 March I’ll be in Adelaide speaking at _SOUTHSTART 2025.

I’ll be “in conversation” with Matt Allen, the founder of Tractor Ventures and one of the fascinating people I get to work with and invest alongside. I’m sure he’ll have some interesting questions for me and I definitely have a few I’m looking forward to throwing at him.

Maybe I can convince a few more Kiwis to pop over with me, and enjoy some late summer in South Australia? You can get your tickets here: https://bit.ly/4hy24LM

🎭 Startup Theatre

When I posted a similar promotion for the Sunrise festival on LinkedIn earlier in the year, I got this (I think) tongue-in-cheek response from Philip Fierlinger:2

I mean, he’s 100% right. As I say in the book:

I’ve attended and even spoken at many startup events over the years. The irony isn’t lost on me when I say “prioritise talking to customers” to a room full of people who are prioritising listening to me rather than talking to their customers.

As I replied to Phil back then, “Come as a music fan, not a startup founder!”

The key, I think, is captured in that Picasso quote above, which I’ve heard independently from several different people over the last couple of months. Perhaps it’s now a meme, and the next step after that is it gets adopted by the exact people it’s intended to criticise, but there is some important truth in it about the fundamental difference between a gala gallery launch and a meetup of artists.

The question we should ask about what I call “Startup Theatre” is whether it’s building up confidence more than it’s building up competence. It’s fine to be a cheerleader and to make people feel energised and excited - that’s important. But always remember, cheerleaders are for the fans watching in the grandstand, not the players in the arena.

🚧 Technical Debt

When we work on a startup, it’s important to remember: everything is legacy.

One of the delightful things about writing How To Be Wrong has been dredging up memories of the very early days of ventures like Trade Me, Xero, Vend and even Timely - some now 20-plus years old. As someone who rebuilt Trade Me twice, I sometimes wonder (nervously) if any of my original code is still running there.

We are living in the good old days. The brand new code we’re writing today (or prompting Cursor to write, more likely) will soon enough be the “spaghetti” that others are pleading to throw out and rewrite.

Software engineers often complain about “technical debt” - describing the costs that will be incurred in the future as a result of shortcuts taken today. That begs the question: does it ever make sense to borrow from the future like this?

We readily accept that mortgage debt is a useful tool to help build equity ownership - few people would suggest saving the full cost of a house before purchase, for example. Despite paying back multiples of the borrowed amount, we understand the trade-off - the benefit of living in the house while making repayments, plus potential appreciation in value in the meantime. And as we discussed last week, many New Zealanders are particularly susceptible to this thinking, with their unwavering belief that property values will always increase.

So why don't we think about technical debt the same way? Why do we never talk about technical equity, or ever consider where and when and how this accumulates?

When we build something new (hello, world?) the technical balance sheet is straight forward: The technical asset we’re building is the software. The technical equity, on the other hand, is the value we get from having that software.

But, as we write more code, we have hard decisions to make. It’s generally impossible to write all of the code we need as soon as we need it. We have to trade-off between completing every feature to a high quality standard and partially completing more features. Remember, we can make things good, fast, or inexpensive - but we can only pick two out of three. In startups, the choice is often between building everything to a high standard, or partially completing more features that let us learn from customers sooner.

In early-stage startups, there is nearly always a lot of what I call Flintstoning: the code isn’t fully tested, and is full of bugs; the code isn’t fully documented, and so can’t be easily picked up and modified by others later; or we have intentionally cut corners to go faster - there are usually multiple ways to build the same feature, and when speed is of the essence we will typically choose the fastest to implement, which is often not the easiest to extend or maintain. Often all three of those things are true! Just like the cartoon, there are human feet powering what appears to be a car.

When faced with an ageing, messy and complicated codebase, it’s common for software developers to think: we need to start again, and this time build it properly. That is nearly always a mistake, because embedded in that existing code are all the little lessons hard-earned as it was written and maintained over time. Trying to replicate those in brand new code tends to be more difficult than software teams ever appreciate.3

But if we take a balance sheet view then the questions would be different: How much technical debt can we service before it becomes a risk? How much technical debt is optimal?

The factor, which is nearly always overlooked by technical teams, is the payback period: If we invest up front in high quality code, how long is it before we get returns on that investment? One consequence of taking much longer to launch is it’s also much longer before we get any feedback from customers about the things we’re building. And, if we stop now and refactor everything, when do we get back to the path we were on previously? Managers often underestimate the immediate costs of working on a messy code base. But engineers are also often wildly optimistic about how quickly and easily they can repay technical debt, and also about how much more efficient they will be in the future given the opportunity to architect everything perfectly now.

Meanwhile, customers rarely care either way.

Rather than falling into the easy trap of complaining about technical debt, I recommend you keep these three things in mind:

1. Given the choice, an engineer will always prefer to write new code than to try and read existing code. The code you struggle to read probably isn’t as bad as you think. And the code you write probably isn’t as good as you think. Future generations of engineers, trying to read the code you write today, will struggle with that exact same asymmetry.

2. As Peter Norvig (apparently) said, “All code is liability”. Every line of code we write is a prediction about the future - and like all predictions, it accumulates interest over time as reality diverges from our assumptions. The real art isn't writing perfect code, but understanding which technical debts are worth taking on.

3. The reason why we build software is so that customers will use it to do the things they want to do. That’s the only sensible frame for deciding when and how much technical debt to incur.

Good luck. Happy borrowing.

Bursting with stories from Rowan's first hand experiences

Most startup advice is dangerously wrong. The mythology of the lone genius founder, the obsession with capital raising and the theatre of startup ecosystems lead countless ventures astray.

Drawing on two decades at the heart of New Zealand's most successful technology companies – Trade Me, Xero, Vend and Timely – this new book unravels the messy reality behind familiar glossy success stories. Combining raw honesty and sharp analysis, it challenges conventional wisdom by sharing compelling firsthand lessons about focused execution, team building and genuine ecosystem growth.

This myth-busting guide is essential reading for founders, investors and policymakers alike, demonstrating that embracing uncertainty, recognising patterns and learning quickly from mistakes are not just steps on the path to success – they are the path itself.

Sometimes the only way to be right is to get comfortable with being wrong.

Pre-sales are available now at https://shop.howtobewrongbook.com.

If you’d like to take a look inside first, you can also download the first chapter for free.

If you prefer you can also buy from BookHero.co.nz. This is a new an independent online bookstore, based in Auckland, which prides itself on customer service, and fast delivery (overnight where possible). I’m delighted they have decided to stock How To Be Wrong, along with 10,000 other titles. Check them out.

How To Be Wrong is proudly published by:

Photo by Gayatri Malhotra on Unsplash

1

Ctrl-Z merchandise is also sold separately - for those hanging out for the ebook or audiobook versions.

1. Aim … Fire 8th December

2. Feedback Loops 15th December

3. A sky full of stars 22nd December

4. Invisible technology 5th January

5. Getting ahead 12th January

Over the holiday period I also published Three-by-Three, my 2024 Year-in-Review.

At the bottom of this post you’ll find a link to download the first chapter of the book. Once you've read it I'd love to hear what you think.

🏟️ Zero-Sum

Try to imagine the scene. It’s the early evening of 16th July 2028.1 A hush descends over the Los Angeles Memorial Coliseum. The large crowd focus their attention on just eight athletes standing at the starting line of the 100m. One by one they are introduced and drop into the starting blocks.

“On your marks”, says the official, pointing the starting gun into the air.

“Set” Briefly, silence, then …

BANG! 💥

Ten seconds later it’s all over. One of the eight athletes is immortalised forever as the Olympic champion. Within a few hours they will be presented with the gold medal, and stand on the dais listening to their national anthem. Two others get the consolation silver and bronze medals for their efforts. The remaining five can only wonder what could have been.2 They quietly file past the media waiting to interview the winners, mostly forgotten like the hundreds of others who have been knocked out in earlier rounds of the competition and didn’t even make it to the start line for the final.

This is one of the remarkable things about sport, and perhaps part of the reasons why many of us are captivated by this type of competition. There is guaranteed to be a winner. But only one.3

We win. You lose. Or vice versa.

But not all things are as contrived as that.

One of the remarkable things about business is there are no guaranteed winners at all - it’s possible that everybody will fail. And at the same time, no reason why there can’t be multiple winners. When one startup is successful it doesn’t detract from the likelihood that another will also be successful - in fact, often the opposite, as lessons learned can be shared.4

In sports, victory is zero-sum. In startups, success is cumulative.

A finite game is played for the purposes of winning.
An infinite game is played for the purpose of continuing to play.

— Finite & Infinite Games, James P. Carse

🥧 Baking the pie

One of the things we love to talk about in New Zealand is “getting ahead”. Once you attune to that expression you hear it constantly.

It’s the reward promised to everybody who works hard. And it isn't just a catch phrase - it's a national obsession that shapes our politics and decision making.

Christopher Luxon has said: “I became Prime Minister so that you and your family can get ahead”.

The bit that is left unspoken is: ahead of whom?

When we talk about getting ahead, if we’re honest, we’re mostly only talking about one thing. By far and away the most popular way we believe we get ahead is by owning real estate. The common wisdom is we build our wealth by “getting on the property ladder”. This is an enticing idea. Provided you can somehow jump onto the first rung of said ladder (usually by borrowing as much as possible from a combination of banks and relatives) then over time you’ll be able to climb up the subsequent steps - presumably by buying more properties or more expensive properties or both.

The problems with this are almost never said out loud.

Spending more and more to buy little bits of our country off each other is mostly another zero-sum game.

Making housing increasingly unaffordable for everybody not already on the ladder is an empty victory.

Rampant house price inflation is not wealth creation, it is poverty creation as it makes homeownership less accessible to everyone.5

And, remember, “everyone” in this case includes all of the people who we need to continue to choose to live and work (and pay tax) in New Zealand.

In the last election, one of the big ideas was to make it easier to sell little bits of our country to international investors. As if that is not a one-time sugar hit. Our headlands for their muskets and blankets, all over again.

This fixation has pretty obviously made us relatively poorer, not wealthier. While we've all been busy trading houses amongst ourselves at ever-increasing prices, countries like Singapore have focused on creating actual economic value and seen their wages grow significantly faster than ours.

Consider a “businessman”, who after achieving some financial success invests their savings in a handful of rental properties and then sells them for a profit a few years later. What business are they in, exactly? What value are they creating? What problem are they solving?

We need to redefine what it actually means to “get ahead”?

When Trade Me was sold, it didn't close the door on other opportunities - in fact, the opposite. The capital returned to shareholders and employees was reinvested in dozens of new ventures. The team members who cut their teeth building Trade Me went on to start or join other companies, bringing hard-won experience about what works. Even the technical challenges we solved became lessons that helped others avoid the same pitfalls.

When Xero succeeded, it didn't use up New Zealand's quota of successful software companies - it helped prove we could build world-class products from here. More importantly, it showed that we could take on international competitors in established markets and win. The pathway from startup to global success that Xero pioneered has since been followed by many others.

When Vend and Timely were acquired, it didn't exhaust the pool of potential buyers - it highlighted that companies started here could achieve great outcomes. Each exit created new groups of experienced operators and investors who understood what it takes to build valuable businesses. Many of them are now working on their next ventures, armed with knowledge that can only be gained by doing it once.

I got to see all four of these examples first-hand, working on and investing in these businesses in the early stages. This is why I wrote How To Be Wrong. Not because there's a formula to replicate these past successes - there isn't - but because understanding how we navigated similar challenges can help founders working on their own ventures today to avoid common pitfalls and highlight the patterns that matter. And to highlight the reasons why we do all of this in the first place.

The wealth we create from technology isn't extracted from a finite resource - it's generated by solving problems and creating value. Each success makes the next one more likely, not less.

🤪 Ellsberg’s Paradox

What would it take to convince more of us to invest in startups rather than real estate? How do we warn ourselves off our obvious and harmful addiction to property?

A common answer is: a capital gains tax. But I'm skeptical. Let me explain why…

Ellsberg's Paradox describes how people make decisions when outcomes are uncertain. In a famous experiment, people consistently demonstrated preference for known probabilities over unknown ones - even when the unknown option might offer better odds. We choose certainty over ambiguity, despite poorer outcomes.

This perfectly describes New Zealand's investment psychology. The fundamental problem with our housing market isn't that capital gains on property are tax free. It's that everybody believes those gains are almost certain and that property values won't fall in the long term. Plus, these "sure thing" investments can be leveraged via mortgages.

This is why most people are much less enthusiastic about other types of tax-free capital gains - like investments in businesses. In those cases, both the capital gains and the capital itself are truly at risk. The outcomes, while potentially much much larger, are ambiguous rather than certain.6

Given this deeply ingrained psychology, it's difficult to see how a capital gains tax would significantly change how capital is allocated. Especially if it applied to all investments equally. And even more so if there are exclusions for housing (while even the most bullish advocates of a capital gains tax think the family home should be excluded, there is never any mention of excluding the family business).

If we really want to shift investment preferences - less into property, more into productive businesses - then we would somehow need to make those things we want to encourage seem less risky.7

On the other hand, if we want to make housing more affordable then we need a solution that causes prices to fall. But nobody wants their property value to fall, so any solution that would cause that to happen is considered politically impossible.

Which takes us back to the heart of our problem: The fundamental issue isn't tax treatment. It's that we've collectively convinced ourselves that property investment is a sure thing. We've created a negative feedback loop, that keeps dragging us down. We’ve chosen to get poorer together, while telling ourselves we're getting ahead.

The alternative is building valuable companies that solve real problems. These might offer more ambiguous outcomes, but they are not zero-sum. They are the only way we’ll break this cycle. The sooner we start the better.

“On your marks…”

📚 How To Be Wrong

Pre-sales starting 19th January

It’s nearly time… I’m preparing special money-can’t-normally-buy offers as part of the pre-sales promotion, which will be available to buy starting next week. Stay tuned!

I’d also love to showcase your startup to my audience. If you have a product or service that would resonate with readers passionate about building and investing in great companies, I can include a promotion for what you’re building with every pre-sales copy sold. Email hello@electricfence.nz with your offer details.

It’s also not too late to buy limited edition Ctrl-Z caps and t-shirts . The special pricing on these will end on Tuesday this week, so if you’d like one of your own, don’t delay.

This book has been years in the making, and I'm excited to finally share these stories and lessons with everybody.

While you’re waiting, the first chapter is available to download now.

Some early praise for How To Be Wrong:

Last week I shared reviews from Derek Sivers, Samantha Wong and Tim Brown.

Here are some more testimonials:

Rowan has done a very rare double act in life - wildly successful founder and investor. So anyone interested in start-up success must read this. I especially loved his (sometime painful) pearls of wisdom about the roller coaster realities of successful start-ups.

— Sam Stubbs, Managing Director at Simplicity

Founding a startup is often glamorised as all excitement and a smooth ride to success. The reality? It's bloody tough, messy, and far from a fairytale. This book cuts through the noise with a no-BS take, packed with juicy tales from some of New Zealand's most well-known startups.

— Ryan Baker, CEO & Co-Founder at Timely

Exactly the sort of typical boring business-y drivel we've come to expect from you over the years, Dad. I'm sure it will be a bestseller.

— Alice Simpson, my first born child!

Read more…

Photo: Silver medalist from the Paris 2024 Olympics 100m, Sha'carri Richardson.
By Rowan Simpson.