Russell McMurray’s recent essay Nothing Personal is the clearest writing I have read on what AI does to wisdom. He gives us a vocabulary. The Scar Line — the personal threshold above which AI use consumes the practitioner instead of serving them. The Certainty Undertow — the soft, validating pull of a machine that never doubts. And his sharpest line:

The process isn’t the cost of getting to the answer. The process is the wisdom. Separate the output from the process and you haven’t saved time. You’ve killed the system.

His diagnosis is right. He points at a quieter question than most AI discourse engages: not what AI does to us, but what we quietly stop doing ourselves when AI starts doing it for us.

But I want to push on his answer. McMurray pivots from diagnosis to prescription — sharper individual self-awareness, then team-based wisdom practice — but the structure he is reaching for is necessary, not sufficient. Cognitive abdication does not stop at the line where you decided you would hold it.

The Scar Line is moving. Quietly, granularly, but moving. And once you see that, the prescriptions change.

The line that does not stay still

McMurray’s framing is elegant. Below your Scar Line, AI helps. Above it, AI eats you. The discipline is to notice where your line sits and refuse to pull AI across it.

There is a hidden assumption in this picture. The line is stable. You decide where it is, you hold it, and your wisdom is preserved on the upper side.

But McMurray’s own argument about Aerobics tells us the line cannot stay still. Wisdom, in his framing, is a living system that needs exercise. Knowledge persists; the system does not. Stop exercising it and it dies.

Now apply that to the Scar Line itself. Every time you use AI to do a task just below your line — analysis, first draft, synthesis, structured summary — you are not exercising the muscle that produced the line. You are just being efficient. But the muscle softens by a fraction, and the line moves down with it.

The line is not a defended border. It is the retreating front of a glacier. Nothing dramatic happens on any given day. Years later, the landscape is different.

This is the shape of what I call cognitive abdication: a slow, granular handing over of practices we did not notice were load-bearing. Hand prediction, modelling, analysis, and consequence-tracking to AI piecemeal, and you do not lose the ability to judge. You lose, by inches, the experiential substrate that made judgment trustworthy in the first place.

There is no safe zone below the Scar Line. There is only a slower zone.

The faculty that detects the problem is the one being lost

McMurray rightly dismisses the idea that sharper individual judgment is the answer. Self-assessment is unreliable at every level of experience.

But there is a deeper reason the personal-discipline answer fails, and it is the one his framework points to but does not quite say.

The faculty you would use to detect that AI has pulled you over your Scar Line is the same faculty AI is eroding.

The senior partner who would once have noticed the wrongness of a structuring decision pre-articulately — the smell before they could name the problem — developed that capacity through years of being wrong in ways that cost real people real money. If AI has been smoothing those errors for five years, the smell never developed. The work flows. By the time the failure shows up, it is downstream and unattributable.

This is why “sharpen your judgement” cannot be the answer. The judgement is being sharpened on increasingly worn material. You cannot detect the loss with the instrument that is itself the loss.

This is also what makes the Certainty Undertow so insidious. The machine is not just good company. It is flattering company. It never makes you feel the small, productive shame of having missed something. And the small, productive shame of having missed something is, on a long enough time horizon, the whole of how judgment is built.

The team in the room is real. The independence is not.

McMurray’s pivot to teams is the right instinct. He sees that the individual-plus-machine configuration cannot hold the line above the Scar Line, and argues that wisdom lives in the interactions between practitioners with different priors.

I agree with the instinct. I want to push on the construction.

The team-as-wisdom-system depends on cognitive independence — something McMurray assumes but does not name. The mechanism only works because the people in the room bring genuinely different priors. The junior sees what the senior had filed away as solved. The senior smells something off the junior cannot yet decode. None of this is possible if everyone has the same starting position.

For most of professional history, that independence was free. Different people read different things, learned from different mentors, made different mistakes. The cost of building three different views into a team was already paid by the time the team formed.

AI quietly removes that free independence.

Consider a modern board meeting. Each member prepares with the same model — the senior asks for strategic analysis, the mid-career for competitive synthesis, the junior for precedent. Three different-looking outputs. So far so independent.

But the model is the same. Its tendencies — the reference classes it reaches for, the framings it finds natural — are constant across the three queries. The three outputs are not three independent perspectives. They are three samples from the same distribution. The variance is shallow. The bedrock beneath is identical.

The three walk into the room thinking they bring different views. In truth they bring three lightly differentiated outputs from a shared latent space — the same considerations weighted slightly differently, the same gaps unnoticed in the same places. The disagreement that surfaces is the disagreement the model already knew about. The thing it does not know — the angle that only a human prior built over decades of being wrong could provide — is absent. Nobody notices, because everyone present feels they did their own thinking.

This is worse than ordinary groupthink. Groupthink emerges in the room and can be detected by the absence of dissent. AI-mediated convergence is pre-installed. It arrives dressed as independent views. The dissent that surfaces is shallow — the kind the model has already mapped. The dissent that would have mattered — the senior’s pre-articulate unease, the junior’s untrained but unspoiled glance — has been quietly painted over. And the better the model, the worse this becomes: the dissent that survives is the dissent the frame anticipated; the dissent that is missing is the dissent that would have come from outside the frame entirely.

Teams are exactly as good as McMurray says — when their independence is real. AI quietly removes that independence. Plural humans plus a shared machine produces, on hard problems, less cognitive diversity than three humans alone twenty years ago, even though it feels like more.

The team in the room is real. The independence is not. And without independence, the team is not a wisdom system. It is a coordination ritual built around a converged prior.

What stewardship actually requires

What AI is consuming is not capability but practice — the slow, cumulative, often invisible practices that produce the capabilities we credit to individuals. None survives being outsourced piecemeal. They die quietly, without anyone deciding.

McMurray’s team-level answer is real. But teams operate within institutions whose incentives pull in the opposite direction: speed, polish, output volume, productivity per head. Sustained deliberation and unaided thinking are costs. Without something explicit at the institutional level, the practices McMurray’s framework requires get sanded down within a quarter.

Cognitive stewardship is the institutional-level answer. It is the deliberate decision to preserve practice as infrastructure — not because practice produces better outputs on any given task, but because it produces the people who can do the work above the Scar Line at all. Stewardship is to judgment what teaching hospitals are to medicine: the work that builds capability has to be protected from the work that extracts value, even when the second is more efficient.

In practice, stewardship amounts to a small number of concrete commitments, all of which will appear inefficient on a quarterly basis.

Fence off the work that builds judgment. Designate categories of decision where AI is not permitted in the deliberation phase. People decisions. Strategic pivots. Ethical escalations. Irreversible commitments. Not because the AI would do them badly — often it would do them well. Because doing them is how the next generation of decision-makers earns the only training that produces real judgement: the experience of being personally responsible for hard calls.

Require unaided first articulation. The first articulation of any consequential position must be written without AI. AI can then stress-test and surface what was missed. But the position originates with the human. A view you have only ever held with the machine’s help is not a view you have built. It is a view you have rented.

Build AI-free rooms. The deliberation itself occurs without devices or model access. The team reasons together, drawing on the cognition they bring. This is how you re-introduce real disagreement — by forcing it to be produced by humans, in real time, rather than imported pre-converged from a shared latent space.

Protect early-career wrongness. Senior practitioners have scars built before AI. Junior ones will not, unless we deliberately preserve the conditions for scarring — structured periods of unaided work on real problems with real consequences. This is the only way to ensure the next cohort has anything to bring above the Scar Line at all.

Make cognitive infrastructure a governance concern. Boards should audit not just decisions and outcomes but the processes that produced them. Were positions formed independently? Was AI used to converge views prematurely? Treat the organisation’s judgement capacity as strategic infrastructure that depreciates without maintenance, and ask whether the organisation can still produce, in five years, the kind of decision-makers it relied on to get to today.

None of these will rank well on any productivity dashboard. All will be derided as friction, as Luddism. Sustained reading was once derided in the same terms. So was slow deliberation.

The whole point of stewardship is that the cost is the point. You are paying to preserve a substrate. The moment the substrate becomes cheaper to outsource, the practice of outsourcing begins, and the substrate thins. Stewardship is the institutional refusal to let that thinning happen on your watch.

Leave a comment

Below the line and beyond it

McMurray ends his piece with an image that has stayed with me. The Consequence Horizon — the frontier past which consequence becomes unknowable. The place where you shape the future. The machine maps the territory up to the edge. It cannot take the step. It takes courage to step into the unknown. It takes wisdom to know you can’t step alone.

He is right that you cannot step alone. But you also cannot step with hands that have atrophied. The resistance he calls for — reaching for the hand next to you — only works if the hands themselves are still capable of doing what hands are for.

The Scar Line is moving. It has been moving the whole time. Practices that look optional are actually load-bearing, and once they go, no amount of personal discipline or team huddling can put them back. We are not facing a choice between AI and wisdom. We are facing a choice between cognitive stewardship and cognitive abdication, and we are already further down the second path than we admit.

The atom did not split. It hollowed.

When Isaac Newton wrote to his rival Robert Hooke in 1676 — a man of notably diminutive stature, which lends the gesture a certain ambiguity — he offered what has become one of the most quoted lines in the history of science: “If I have seen further, it is by standing on the shoulders of giants.” My own observations about New Zealand’s innovation ecosystem are similarly borrowed. They are assembled from two decades in Silicon Valley, from conversations with founders, investors, and researchers on both sides of the Pacific, and most recently from a reader whose comment on my last post stopped me in my tracks.

Charles Sorensson raised something I have been circling around for years without quite landing on it so cleanly. He asked: what creates pull? Not funding, not talent, not timing — no single piece in isolation. But what if, in New Zealand’s particular case, the missing variable is a specific and nameable thing? What if the system is structurally misconfigured not because of what it lacks in quality, but because of what it lacks in direction?

I think it is. And I think the missing piece is the presence of large, R&D-oriented multinational companies with genuine innovation mandates, embedded deliberately and physically close to our universities.

Universities Are Built to Push

Let me start where Charles did: with universities. They are extraordinary institutions. New Zealanders are genuinely world-class in a number of domains — agritech, marine science, photonics, quantum technologies, materials, aerospace, and biomedical engineering. The researchers are talented, the output is rigorous, and the funding, while never sufficient, is real.

But universities are structurally designed to push. The incentive architecture — grants, citations, tenure, research excellence frameworks — rewards knowledge production. It does not reward commercialisation. Tech transfer offices exist at most institutions, and they work hard, but they are intermediaries operating without a buyer reliably on the other end. They push IP toward a marketplace that, in New Zealand’s case, is thinly populated with companies that know what to do with it.

The result is a well-documented and deeply frustrating pattern: excellent New Zealand research gets published, noticed, and then commercialised — elsewhere. The founder emigrates. The IP gets licensed offshore. The value accrues to economies with the commercial infrastructure in place to receive it.

This is not a failure of ambition or intelligence. It is a failure of system design.

The Missing Ingredient: Context

Throughout my career — and particularly as I have thought about what separates innovation ecosystems that compound from those that stall — I have become increasingly convinced that the most underrated variable is context.

Great researchers cannot be told what to do. Attempt to direct them too narrowly, and you extinguish the very quality of mind that makes them valuable. But there is a profound difference between direction and context. Provided with a rich, real-time understanding of what problems matter most — to their industry, to their community, to the world — researchers naturally gravitate toward work that is both intellectually serious and commercially relevant. They do not need to be managed to create value. They need to be informed about where value is being sought.

This distinction matters because it reframes what we should actually be asking of the commercial sector’s relationship with universities. We are not asking companies to set research agendas or to capture academic output for private benefit. We are asking them to be present — to be in the room, at the seminar, in the corridor conversation — so that the ambient signal of what is commercially urgent reaches researchers as naturally as the ambient signal of what is academically interesting. Context is the mechanism by which pull works. And it is precisely what New Zealand’s research community is, at present, largely denied.

What Pull Actually Looks Like

In the ecosystems we tend to admire — Silicon Valley, Boston’s Route 128 corridor, Eindhoven in the Netherlands, Grenoble, France, and Melbourne, Australia — the defining feature is not the universities themselves. It is the gravitational relationship between the universities and the large commercial entities that orbit them.

IBM and MIT. Genentech and UCSF. Google, Lockheed, and Stanford. These are not coincidences of geography. They are the result of deliberate co-location strategies by large corporations that understood something important: proximity to a university research environment is not just a recruiting advantage. It is an intelligence function. It lets you see what is coming before it arrives. And crucially, it works in both directions — the corporate presence continuously radiates context back into the research environment, shaping what questions feel worth asking without ever issuing a directive.

The best examples of this go further still. IMEC in Leuven, Belgium, became one of the world’s most important semiconductor research centres not because Belgium had an obvious claim to that role, but because it built a model in which large multinationals — Intel, Samsung, TSMC — co-invest in pre-competitive research alongside a public institution. The Fraunhofer Institutes in Germany operate on a similar logic: applied research conducted in close partnership with industry, solving real commercial problems, with a deliberate mandate to transfer knowledge into the economy. Singapore has spent thirty years constructing exactly this kind of architecture, and it has worked.

What these models share is pull — and the mechanism of pull is context. Researchers produce knowledge in conversation with partners who have skin in the game, can articulate which problems are commercially urgent, and can provide a pathway from the laboratory to the marketplace. The horse, as Charles put it, gets a cart with better wheels — because someone who needs the cart to move faster is standing right there, and the horse knows it.

Perhaps the most telling evidence of this dynamic, however, is the behaviour of sophisticated non-American companies toward Silicon Valley itself. Toyota, Volkswagen, Bosch, and Samsung have all established substantial R&D presences in the Bay Area — not to sell cars or semiconductors to Californians, but to be physically close to the crucible of ideas. SAP, the German enterprise software giant, runs a significant innovation lab in Palo Alto. Alibaba and Tencent, despite the political complexities, have maintained research outposts there. Nokia, when it was still a force in mobile, famously built its most forward-looking research unit not in Helsinki but in Silicon Valley, precisely because the ecosystem intelligence available there was worth more than any cost saving from centralisation.

I know this not only from observation, but from direct experience. As a member of the Bay Area Science and Innovation Consortium — a body established by the Bay Area Council to preserve and strengthen the San Francisco Bay Area as a global crucible of innovation — I was involved in meeting the leadership teams of many global companies that were considering establishing a small R&D footprint in the region. We were deliberate and active in making the case, and we were successful. Those relationships helped establish and reinforce the gravitational pull that Silicon Valley exerts on global start-ups to this day. The companies that came were not doing so on a whim. They came because a community made the argument clearly, backed it with relationships, and demonstrated that the ecosystem had something genuinely worth being close to.

These companies were not naive about the expense or the difficulty. They made a deliberate calculation: the optionality and early visibility that comes from proximity to a world-class innovation cluster is worth paying for — and, crucially, that calculation can be influenced by the quality of the welcome they receive. The same logic, applied in reverse, is the argument I am making for New Zealand. If the world’s most sophisticated companies can be persuaded to establish R&D outposts in California, it is not a stretch to imagine that the right framing — the right domain, the right relationships, the right policy environment — could bring a small number of them to Auckland, Wellington, Christchurch, or Dunedin for the same reason. What worked in one direction can work in another. What is required is not a miracle. It is the deliberate, sustained effort to make the case.

New Zealand’s Structural Gap

New Zealand’s commercial sector is dominated, for understandable historical reasons, by domestically-oriented businesses. Retail, property, agriculture, banking, telecommunications. These are not innovation-hungry industries in the research sense. The multinationals that do have a presence here — the major banks, the FMCG companies, the telcos — are largely distribution and service arms. Even the major technology multinationals with a presence in New Zealand make their R&D decisions in California, London, or New York. Their New Zealand operations have neither the mandate nor the budget to engage seriously with university research.

The consequence is a context vacuum. When a researcher at the University of Auckland develops a promising technology, or when a team at Victoria University produces a genuinely novel result in a commercially relevant domain, they look around for a commercial partner and find the room largely empty. There is no ambient signal telling them which of their possible directions leads toward a problem the world will pay to solve. There are New Zealand companies with a genuine appetite for innovation — and I am proud to back a number of them through Quidnet — but the ecosystem of large, well-resourced, technically sophisticated commercial partners that would continuously replenish that signal is simply not here at the scale required.

The consequences are compound. Without a commercial context in the room, researchers cannot calibrate their work to real-world problems — not because they lack the inclination, but because they lack the information. Graduate students see no local career pathway and emigrate, taking with them the embodied knowledge that no publication fully captures. Founders with promising technology cannot find the strategic partnerships that would allow them to scale without immediately leaving. And each departure increases the likelihood of the next departure.

I have been advocating, so far without measurable effect, for a specific intervention: encouraging large multinationals to establish small, R&D-focused satellite teams in New Zealand, located deliberately near our university research clusters. Not full subsidiaries. Not sales offices. Small, technically sophisticated teams with genuine innovation mandates — the kind of presence that restores context to our research environment and changes the gravitational dynamics of the ecosystem around them.

Why a Multinational Would Do This

The reasonable question is: why would they? New Zealand is a long way from anywhere. The market is small. The talent pool, while excellent in quality, is limited in depth.

The answer lies in understanding what these beachhead investments are actually for. They are not market-entry plays. They are intelligence and optionality plays. A small team of senior researchers embedded in Christchurch or Auckland gives a global corporation early visibility into what is emerging from New Zealand’s science base — and New Zealand produces genuinely world-leading work in several domains that matter enormously to the global economy over the next two decades. The exchange, moreover, is not one-directional: the corporate team also radiates context back into the local ecosystem, informing researchers about where the commercial frontier is moving in ways that no grant application or journal submission ever could.

There are other advantages that are underappreciated. New Zealand sits in a time zone that bridges Asia and the Americas, which is operationally useful for global R&D networks. The country has a stable regulatory environment, strong intellectual property protections, and — for the right kinds of research — access to unique natural environments that are simply not replicable elsewhere. The agritech and marine science opportunities, in particular, are tied to physical realities that cannot be reproduced in a California laboratory.

There is also the AIP — the Active Investor Plus visa programme — which has made New Zealand meaningfully more accessible to offshore capital. We have used it as a differentiator for Quidnet’s Fund II, and it has resonated. But the logic applies equally to corporate R&D investment. The policy infrastructure to welcome this kind of presence exists, or can be constructed without enormous difficulty. What is missing is the deliberate effort to make the case.

Patience, Systems, and the Longer Arc

Charles made another point that I do not want to pass over too quickly. He pushed back against the tyranny of speed — the “quick in, quick out” model that dominates much of contemporary startup discourse. He is right that some innovations require a longer arc. Regulatory environments need to mature. Technical dependencies need to resolve. Market timing is real. The venture capital model, with its fund cycles and return horizons, is not always well suited to holding that space.

This is another reason why multinational R&D beachheads matter. A large corporation with a multi-decade time horizon and a substantial balance sheet can sustain a research partnership through the fallow periods that would exhaust a venture fund. They can absorb the cost of waiting for a regulatory environment to catch up, or for a complementary technology to mature, in ways that early-stage investors structurally cannot. And crucially, they can maintain the context signal continuously — even during the fallow periods — so that when the moment of readiness arrives, the research community is already oriented toward it. The patience that Charles identifies as a reality — not a virtue, but a simple requirement of some innovation trajectories — is more available to a corporate partner than to almost any other actor in the system.

This is not an argument against venture capital. It is an argument for a more complete system — one in which different actors play the roles they are actually suited to play, rather than one in which we ask early-stage investors to carry weight that belongs on stronger shoulders.

What Would Actually Change

If New Zealand succeeded in attracting even a handful of serious multinational R&D presences — genuinely technical, genuinely embedded, genuinely engaged with university research — I think the downstream effects would be larger than they appear.

Researchers would design projects in conversation with partners who understand what problems are commercially urgent. The gap between academic output and commercial application would narrow, not because researchers became less rigorous, but because they had better information — better context — about what rigour was needed where. Graduate students would have career pathways that did not require a flight to Sydney or San Francisco. IP would be more likely to be commercialised in, or at least through, New Zealand, building local capability and local revenue. And — this is the compounding effect — multinational presence attracts multinational presence. The first serious R&D beachhead makes the second easier to justify.

None of this happens automatically, and none of it is cheap. It requires deliberate policy design, active government-to-business diplomacy, and a sustained investment of political attention over a period longer than any single electoral cycle. It requires New Zealand to be explicit about what it is offering and to whom, rather than relying on the general warmth of its reputation to do that work.

But the alternative — continuing to produce excellent science and watch it leave — is not a neutral outcome. It is a choice, made by default, with compounding costs.

Leave a comment

A Final Word to Charles

You asked what would create momentum in a complex system of competitive players with misaligned incentives and different time horizons. I do not have a complete answer. But I think momentum requires a direction — something that orients the system’s energy rather than just adding more of it. Pull provides that direction in a way that push, on its own, cannot. And the mechanism of pull, at its heart, is context: the continuous, real-time signal of what problems matter, flowing from the commercial world into the research environment and back again.

The horse needs a cart. But first, someone has to want what the cart is carrying badly enough to show up and say so — and to keep saying so, every day, until the horse knows exactly which way to run.

Mark Bregman is Managing General Partner of Quidnet Ventures, an Auckland-based deep tech seed fund. He writes Antipodean Musings on innovation, political economy, and life at the bottom of the world.

The Catalyst

A short exchange ran on LinkedIn this week that captured, in miniature, something I’ve been turning over for a long time. Lane Litz, founder of What Founders Want, posted an observation:

“There are no incubators currently running in New Zealand. 🤯”

By “incubator”, she didn’t mean the equity-taking, cohort-based programmes that have come and gone here over the last fifteen years. She meant the older, looser model: a long-term, community-based environment where founders work at their own pace, with ongoing access to advisors, no fixed cohort, no demo day, and no equity required. A year is more realistic than twelve weeks, she argued, and what founders most need is “community while getting shit done.”

Rowan Simpson responded with the comment that drew most of the attention:

Perhaps the thing you’re missing is that none of these programs are designed for founders. They are really designed for the people who run them. In a hospital, having no sick babies in an incubator would be a milestone to be celebrated, no?

And a second:

We learned that shared working spaces and underwhelming mentors do not attract the best founders — doesn’t matter how many foosball tables and bean bags you have.

An unusually candid pair of observations. Sam Minnée, reflecting on his SilverStripe days, was more nuanced — the looser incubator structure was “less likely to get in founders’ way” than the cohort accelerators that came later, and he valued the community aspect — but landed on a line that gets to the heart of it:

There’s a necessary humility in how much the programme itself is doing — it’s a place where like-minded people can be near each other (including potential advisors); what they do with it themselves is up to them.

Aaron Scott then asked the obvious next question: what does a founder-focused incubator actually look like?

That’s the question I want to take seriously — not because incubators are the right answer, but because answering it properly requires saying something about what New Zealand’s innovation ecosystem actually lacks, and why we keep having the wrong debate about it.

Three things people mean by “incubator”

The word does too much work. An accelerator is a fixed-cohort, time-boxed programme: eight to fourteen weeks, demo day, equity for cash. Y Combinator is the archetype. Most New Zealand programmes that came and went between 2012 and 2022 — Lightning Lab, Creative HQ’s accelerator, Vodafone xone — sat here. A classical incubator, in Lane’s sense, is open-ended, place-based, light on programme structure, often non-dilutive. This is what she’s asking after, and it has largely disappeared from New Zealand. A tech-transfer incubator is a third creature again, existing to commercialise university and CRI research, taking equity in exchange for repayable grant capital. The Callaghan Technology Incubator scheme — WNT, Bridgewest, Sprout — sits here, though its future is now uncertain.

Conflating these three is half the reason the debate goes in circles. Lane is asking why we don’t have the second. Rowan is critiquing what the first largely became. The third is doing its own quite different job.

What the evidence actually says

The empirical picture of accelerators is more sober than the marketing suggests. Careful work — Gonzalez-Uribe and Leatherbee on Start-Up Chile, Hallen, Cohen and Bingham on the broader population — finds real treatment effects, but modest ones. The mechanisms that matter are compressed feedback, network access, and signalling to capital. Most of the apparent magic of Y Combinator is selection: the best founders apply, get in, and would mostly have succeeded anywhere. Rowan’s line — that shared working spaces and underwhelming mentors do not attract the best founders — is empirically defensible.

Notice what’s underneath that finding. Two of the three mechanisms that do work — network access and capital signalling — are not really about programme structure. They are about proximity to the right people. The programme is a delivery wrapper. The active ingredient is the network.

The diagnosis no one quite states

What founders in New Zealand most reliably lack is not a curriculum, a twelve-week sprint, or a building with a foosball table. It is sustained, serendipitous contact with the other people who make innovation happen. Quantum physicists who never meet venture capitalists. Policy thinkers who never sit across from agritech founders. Researchers whose work could change an industry but who have no path to the entrepreneur who could carry it. The default state, in a small and geographically dispersed country, is isolation.

Innovation is a collaborative contact sport. The pattern that produces it — university towns, coffee houses, the early Royal Society, the Wagon Wheel bar in Mountain View where semiconductor engineers swapped salaries and stories — has always been about places where paths cross. Silicon Valley was not built by programmes. It was built by a permeable boundary between Stanford and industry, an open-network norm, and the geographic accident that put a critical mass of the right people inside a thirty-mile radius. The programmes came later and largely codified what the community had already produced.

This is why Rowan’s critique lands so hard. Programmes designed for their operators do not produce connective tissue; they produce throughput. Sam’s defence of the community aspect of the older incubators points at the same thing from the other direction. What was valuable was not the curriculum. It was the proximity, and the trust that built up between people who kept showing up to the same rooms.

Aotearoa’s structural problem is that the rooms have mostly stopped existing. The institutions that historically did this work have been hollowed out, repurposed, or designed for a different job. What remains is a set of programmes optimised for processing cohorts and a deep-tech scheme doing important but narrow work. The connective tissue has thinned.

What the alternative looks like

I should be careful here. I have spent the last several years working on a concept — currently iterating under the working name The Crossroads Collaboratory — that tries to instantiate one version of the answer. I want to be honest about that interest without overselling the project, because what matters is the design problem it gestures at.

The hardest part of getting something like this off the ground in New Zealand has not been money, space, or founder interest. It has been the absence of reference models the country can readily picture. When I describe a salon-style innovator community — diverse by design, anchored in trust rather than contract, measured in years rather than quarters — people pattern-match to the thing they know: a coworking space, an accelerator, an industry association. None of those is the right comparison. The closer reference points are Shack15 in San Francisco, the Long Now Foundation’s gatherings, the old Homebrew Computer Club, and the Royal Institution in its original form. These are not programmes. They are infrastructures of contact, and they produce disproportionate output precisely because they do not try to do what programmes do.

The design principles are straightforward. Place-based but not space-driven: who is in the room matters more than the WiFi password. Curated for ecosystem mix rather than cohort homogeneity — founders, investors, researchers, policy thinkers, storytellers, artists — because the serendipity premium is highest when the room is heterogeneous. Designed for serendipity rather than throughput: salon dinners, fireside conversations, briefings, showcases. Funded so operator incentives align with founder outcomes — not equity, not founder fees, but anchor membership, philanthropic underwriting, or sub-sovereign support that lets operators be measured by what the community produces over a decade. And honest about scope: this infrastructure does not replace capital, IP support, or technical mentorship. It does the one thing those cannot do for themselves. It puts the right people in the same room, repeatedly, over years.

The deep-tech qualification

Deep tech, where my own fund does most of its work, needs more than community: long-horizon translational capital, specialist facilities, regulatory navigation, and technical mentors who have shipped real hardware or biotech. But even there, the deal flow depends on being inside the connective tissue, not outside it. Cyclotron Road and Activate at LBNL are nominally programmes; they are really embedded communities with capital and labs attached. Community-plus, not community-or.

Leave a comment

Back to Aaron’s question

What does a founder-focused incubator look like? The honest answer is that the question is slightly misframed. What New Zealand most needs is not another programme. It is the infrastructure of contact that programmes, at their best, were always a partial substitute for. Less like a cohort and more like a club. Less like a curriculum and more like a calendar of high-trust gatherings. Less like a twelve-week sprint and more like a decade-long bet on a thicker network.

The Lane/Rowan exchange is a small public version of a much larger national conversation we keep not quite having. New Zealand’s binding constraint is not founders, ideas, or capital. It is the institutional scaffolding that turns research into export-capable companies, and that scaffolding depends on connective tissue that takes decades to build and minutes to damage. The disestablishment of Callaghan Innovation is a moment of risk for exactly this reason: whatever replaces it will either invest in connective tissue, or it will not.

The right question is not whether incubators went out of style. It is whether we are willing to fund the infrastructure of contact, an innovation ecosystem actually requires, and to measure it by the patience it demands rather than the cohorts it processes.

There is a quality to the current moment that most people I speak with seem to recognise, even if they cannot quite name it. A low background hum of unease. A sense that the ground is moving faster than the legs can carry the body. Things one used to feel certain about — about work, about institutions, about how the world fits together — no longer cohere in the way they once did, and the gap between what is true and what feels manageable has widened to the point of strain.

I have come to think this is not, in the main, a personal failing. It is not, in the main, even a generational complaint. It is a structural issue that has been building for a very long time, and we are now living in the steady-state version of it.

The most useful framework I have found for thinking about it pairs two bodies of work that are not normally read together: the long evolutionary account of how human cognition has fallen behind its own environment, and Leon Festinger’s 1957 theory of cognitive dissonance. Each describes a piece of the puzzle. Together, they describe something larger than either accounts for on its own.

The Three Layers of the Adaptation Gap

The brain in your skull is, in evolutionary terms, almost unchanged from the one that walked the African savannah two hundred thousand years ago. It was calibrated for small-band hunter-gatherer life: groups of around 150 people, seasonal rhythms, immediate-return food economies, threats that came in the form of fur and teeth. Once agriculture produced settlements of thousands, dense disease pools, abstract property and written contracts — roughly twelve thousand years ago — the environment had already outrun the genome’s capacity to retune itself. E.O. Wilson’s line on this is hard to improve on: paleolithic emotions, medieval institutions, godlike technology. That divergence is about four hundred generations old. It has been widening ever since.

For most of the intervening period, culture and institutions did the work that biology could not. Language, writing, law, schooling, markets — these are scaffolding. They allowed minds built for one kind of world to function in another. The arrangement worked, more or less, as long as the rate of change in the external world remained slow enough for the scaffolding to keep up.

It is the rate of change that is the second story. Herbert Simon’s work on bounded rationality made the clinical version of the argument: the information relevant to any serious decision had, by the mid-twentieth century, exceeded the capacity of any single mind to process it. Alvin Toffler’s Future Shock, published in 1970, made the popular version: the world an adult retired into would no longer resemble the one they had entered as a young person. Generational transmission of skills, norms, and meaning began to fail. The scaffolding had not collapsed, but it was creaking.

That, I think, is the third layer of the gap, and the one we are still inside. Biology fell behind in the Neolithic. Individual cognition fell behind during the Industrial Revolution. Culture itself — the institutions and shared meanings that had previously absorbed the shock — began falling behind somewhere around the middle of the twentieth century, when novel categories of decision (nuclear policy, global supply chains, ecological tradeoffs, financial derivatives, networked computing, and now machine intelligence) started appearing faster than any society could metabolise them into stable norms.

We have been running on borrowed scaffolding ever since.

What the Gap Feels Like from the Inside

The historical account is correct as far as it goes, but it is missing something important: it does not say what living inside the gap actually feels like. This is where Festinger becomes useful.

Cognitive dissonance, in the original formulation, is the discomfort that arises when new information collides with existing beliefs. The discomfort is not metaphor. It is a real psychological state the mind seeks to resolve, and Festinger described four standard resolutions. One can change the belief to fit the new information. One can add new beliefs that buttress the old one against the new information. One can diminish the importance of the conflicting information. Or one can dismiss the source of that information entirely.

Only the first of these resolutions involves actually updating one’s worldview. The other three are forms of defence. And — this is the crucial part — defence is almost always cheaper than revision. Changing a belief that is woven into identity, profession, relationships, and meaning is expensive. Discrediting the source of the inconvenient information is, by comparison, almost free.

Festinger’s original studies treated dissonance as an episodic event. A subject would be exposed to a piece of information that contradicted a held belief; the mechanism would activate; the resolution would follow; equilibrium would return. The framework assumes a baseline of cognitive coherence, with dissonance as the disturbance.

What happens when the disturbance becomes the baseline?

When Dissonance Becomes the Operating Condition

This is, I think, the actual phase shift we are living through. The adaptation gap has now widened to the point that the information environment is producing dissonance faster than any individual mind can resolve it. The mechanism Festinger described — designed to manage occasional conflicts between belief and reality — is now under continuous load.

When dissonance is continuous, the lowest-energy resolutions become the dominant ones. We do not, at scale, update our worldviews. We add buttressing beliefs. We discount the importance of conflicting evidence. We discredit the sources that produce it. And we do these things not because we are stupid or dishonest, but because the alternative — continuous belief revision at the pace the environment now demands — exceeds the cognitive budget of any normal life.

Three terrains where this is now visible at civilisational scale, I think, are worth naming.

The first is politics. What is read as polarisation is in part a sorting of people by which defensive resolution they have adopted. The information environment is producing more dissonance than can be metabolised, so subgroups settle on which sources to trust and which to dismiss, which evidence is important and which is noise. The arguments are not really arguments. They are negotiations over which defensive resolution gets to be authoritative.

The second is institutions. A large institution cannot acknowledge that its founding assumptions have been outrun by reality without dissolving the basis of its own authority. So it does, at the institutional level, what individuals do at the psychological level — it produces ever more elaborate justifications for its continued relevance, dismisses critics as bad-faith actors, and treats inconvenient evidence as a matter of communications rather than substance. Festinger would recognise the pattern immediately.

The third is the reaction to AI. Generative AI is, among other things, a very efficient dissonance machine. It surfaces tensions in our beliefs about creativity, expertise, authorship, and the value of human labour that most people would prefer not to have surfaced. The responses one observes — dismissal, minimisation, ridicule, the production of buttressing beliefs about what AI really is or is not — are nearly textbook examples of the cheaper resolutions. The harder resolution, which is to actually revise one’s working assumptions about what the next decade of work will look like, is being deferred at a scale that will eventually compound.

The Asymmetry That Makes It Worse

There is a quietly devastating implication in all of this. Festinger’s framework includes an asymmetry that the historical account, on its own, does not capture: belief distortion is almost always cheaper than structural change. The deeper the dissonance, and the more central the threatened belief is to identity or institutional survival, the larger the gap becomes between the two options — and the more strongly the defensive resolution is selected.

The mechanism is therefore self-reinforcing in exactly the direction that makes it worse. As the adaptation gap widens, the cost of honest belief revision rises faster than the cost of defensive distortion. Which means the proportion of cognitive activity devoted to defence, rather than revision, tends to increase over time. Both in individuals and in institutions.

This is, I think, why the current period feels strange in a way the previous ones did not. Earlier generations dealt with technological and social change at a pace that allowed at least some genuine adaptation. We are dealing with it at a pace that has tipped the cost-benefit calculation decisively toward defence.

Leave a comment

The Skill That Is Actually Required

If the diagnosis is right, then the most valuable cognitive capacity in this environment is not, as is often assumed, intelligence or analytical horsepower. There is no shortage of either, and both are perfectly compatible with sophisticated defensive resolution at scale.

The thing that is actually scarce — the thing that is, I would argue, the most genuinely countercultural cognitive skill of the present moment — is tolerance for unresolved dissonance. The capacity to hold a contradiction without immediately collapsing it into a defensive position. To say: I do not yet know how to integrate this new information with what I previously believed, and I am going to sit inside that discomfort for longer than is comfortable, because the alternative is to dissolve the contradiction prematurely and call the dissolution wisdom.

This is harder than it sounds. The mechanism Festinger described is not optional equipment. It activates whether we want it to or not. The skill is not to suppress it, but to notice it operating — to develop the metacognitive habit of asking, when a piece of information feels easy to dismiss: am I dismissing this because it is wrong, or because dismissing it is cheaper than revising the belief it threatens?

The people I have come to most respect, in business and outside it, are not the ones with the strongest convictions. They are the ones who can sit inside a live contradiction without resolving it on the wrong side just to feel better.

The world has outrun us. The honest response is not to pretend it hasn’t. It is to learn to think well from inside the gap.

Introduction

There is a scene in the novel Dune that has stayed with me — not because of spectacle, but because of its inversion of instinct.

The story is set two hundred centuries from now. And yet, in this distant future, combatants carry personal shields that deflect any fast-moving attack. The only way through is counterintuitive: move slowly. Deliberately. With restraint.

The slow blade penetrates the shield.

I keep coming back to this when I think about operating in New Zealand. And, by contrast, Japan — a culture whose own roots run back nearly two thousand years in the other direction.

What I Got Wrong at First

New Zealand looks, on the surface, like an easy place to operate. The society is informal. Hierarchies are flat. You can get in front of decision-makers with surprising ease. Meetings are relaxed. Nobody leads with their title. For someone arriving from the United States or parts of Europe, it can feel almost disarmingly open.

I was one of those people who found it refreshing — and then found myself stalled.

Deals I thought were progressing weren’t. Decisions I believed were close kept drifting. Feedback that felt positive turned out, in hindsight, to have been something softer — more conditional, sometimes illusory. The friction was never overt. It was embedded in the texture of the conversations themselves, and I kept missing it.

What I came to understand, slowly, is that New Zealand has a deeply ingrained aversion to direct confrontation. Not as a personal quirk, but as something structural — and something with deep roots.

Much of that root system is Māori. The concept of mana — broadly, the prestige, authority, and dignity a person carries — is not simply a measure of status. It is relational. Mana is held in connection with others, and it can be given, received, or diminished depending on how you conduct yourself. This shapes everything. In a society where mana matters, causing someone to lose face isn’t just awkward — it is a genuinely harmful act, one that ripples outward through networks and is long remembered.

Alongside mana sits manaakitanga — the ethic of hospitality, generosity, and care for others. To practice manaakitanga is to actively protect and uplift the mana of those around you. It is not weakness or deference. It is a discipline. And it goes a long way toward explaining why direct confrontation feels so costly here: it is not merely uncomfortable, it is, in a meaningful sense, wrong.

These values didn’t remain contained within te ao Māori. They permeated the broader culture. Pākehā New Zealand absorbed them — often without naming them — and they became part of the social operating system that everyone, consciously or not, runs on.

In a small, interconnected society where reputations travel quickly and endure, this makes particular sense. People work across overlapping networks — professional, social, often familial. The cost of open conflict is high because the distance between parties is short and persistent. So disagreement tends to go underground. Criticism gets softened or deferred. “Yes” can mean “I hear you” rather than “I agree.” Silence carries more weight than it appears. Decisions drift rather than get explicitly rejected.

The harder you push, the less traction you get.

The Playbook That Doesn’t Work Here

The approach I’d brought with me — clarity, urgency, driving toward closure — backfired in ways I didn’t immediately recognise.

What I experienced as directness was received as pressure. What I intended as momentum was perceived as aggression. And aggression, in this environment, doesn’t produce pushback. It produces withdrawal.

Conversations don’t escalate — they dissipate. You walk out of a meeting thinking you’ve moved something forward, and discover later that the opposite has happened. Not through any deliberate opposition, but because the shield absorbed your momentum and you never felt it.

In hindsight, what I was doing was threatening mana — not intentionally, not dramatically, but enough. Enough that people stepped back rather than engaged. Enough that the relationship, which is the thing that actually carries decisions here, quietly contracted.

What Actually Works

What I’ve found — and it took longer than I’d like to admit — is that influence here has to be built rather than applied. It moves through whanaungatanga — the web of relationships, obligations, and genuine connection that binds people together. You cannot shortcut it. You can only participate in it.

The decisions that matter are rarely made in the room. They’re formed through a series of smaller, often informal conversations that happen well before any formal moment. Which means the formal moment, by the time it arrives, is often already shaped. The people who are effective here seem to understand this instinctively — they invest the time upfront, the one-on-one conversations, the early sharing of half-formed ideas, the social interactions that aren’t obviously about anything in particular. It looks, from the outside, like relationship-building. It is, but it’s also strategy.

There’s a parallel here to the Māori hui — the gathering where community decisions are worked through collectively. A hui is not a debate. It is a process of arriving at shared understanding, and the real work happens in the conversations before and around it as much as within it. Formal meetings in New Zealand often function the same way, whether or not anyone would use that word to describe them.

There’s also a literacy to disagreement that took me a while to develop. Disagreement doesn’t disappear in a culture that avoids confrontation — it just expresses itself differently. A qualified endorsement can be a polite deferral. A question can be a disguised objection. A pause can mean more than the words on either side of it. And when you need to challenge something, the effective move is almost always to frame it as curiosity rather than critique — to offer an alternative rather than contradict, to understate rather than emphasise. The goal isn’t to win the exchange. It’s to shift the trajectory without triggering the shield.

What I’ve also learned, at some cost, is that the protection of mana is not optional here — it’s the whole game. Causing someone to lose face, even inadvertently, even in a minor way, can quietly end things. So the discipline of providing off-ramps, of never cornering someone into a binary choice, of separating the issue from the person and doing it visibly — that’s not courtesy. It’s operational. You are, in every interaction, either building mana or eroding it. Your own, and theirs.

Japan, By Contrast

Having worked and lived in Japan, I find the contrast genuinely fascinating — and even more so given the depth of what you’re dealing with. Japanese culture, in its contemporary form, draws on nearly two millennia of continuous tradition. The systems for managing harmony and avoiding confrontation didn’t emerge recently. They were refined over centuries.

Japan also avoids confrontation, but through a far more formalised system. Where New Zealand relies on social intuition — much of it absorbed from Māori frameworks of relationship and reciprocity — Japan relies on process. And that process has a name. Nemawashi, literally “going around the roots,” describes the practice of consulting stakeholders individually before any formal decision is made. Concerns are surfaced privately. Adjustments are made quietly. By the time a proposal reaches a meeting, it’s largely pre-approved.

The meeting itself isn’t a forum for debate. It’s a ritual of confirmation.

What this produces is a different kind of opacity. In New Zealand, resistance tends to be diffuse and unspoken — you have to infer it from context, from the quality of the relationship, from the mana in the room. In Japan, resistance gets surfaced, but only in private and early. Which means that by the time you’re sitting in the room, the question of whether things will go your way has usually already been answered. If you’ve reached the meeting without having done the groundwork, you’ve already failed — you just don’t know it yet.

In New Zealand, the timing of that realisation tends to come even later.

Japan also has a more explicitly articulated framework for the gap between what’s said and what’s meant — tatemae, the public position, and honne, the true feeling. Everyone understands this distinction and navigates it consciously. New Zealand has no formal equivalent, even though the gap very much exists. What it has instead is mana — the felt sense of what can and cannot be said without cost — a regulatory system that operates through relationship rather than convention. It’s just less acknowledged, which makes it harder to read. Silence in Japan has fairly defined meanings. Silence in New Zealand is more ambiguous — and I’ve misread it in both directions.

The practical difference I’ve noticed: Japan’s system is slower upfront but faster to execute. Once consensus exists, action follows with real precision. New Zealand can be faster to initiate — things feel like they’re moving — but slower to conclude. Alignment is looser, and what looks like agreement can remain fluid longer than you’d expect.

Leave a comment

The Actual Adjustment

What I’ve had to recalibrate isn’t just style — it’s the underlying theory of how influence works.

In Japan, the discipline is systematic. Every stakeholder needs to be brought along before the formal moment, and the process itself provides the map. In New Zealand, the discipline is relational — grounded in something closer to manaakitanga than to process. Trust and genuine care for the other person need to be established before you can press for anything, and there’s no process to tell you when you’re there. You have to feel it.

In both cases, the instinct to move faster is the thing that will undo you.

I used to interpret confrontation aversion as indecision. I don’t anymore.

Both New Zealand and Japan have evolved systems that prioritise cohesion over conflict, and they’ve done it for good reasons — these are systems designed for durability, not for rapid, adversarial resolution. In New Zealand’s case, some of that design is ancient. Mana, manaakitanga, whanaungatanga — these are not quaint cultural artefacts. They are a sophisticated operating system for managing human relationships in conditions where everyone is known to everyone else and the long run always matters more than the short.

The cost is real, though. Disagreement can stay hidden too long. Necessary conflict can be deferred past the point where it’s useful. And for someone arriving with a different operating model, the learning curve is quiet and mostly invisible — which makes it harder, not easier, to navigate.

What I’ve come to believe is that in these environments, influence doesn’t belong to the loudest voice or the fastest mover.

It belongs to the person who’s understood, early enough, that the shield is always there — and who has the patience to move slowly enough that, eventually, it gives way.

The Imperfection Economy

Something quietly strange is happening. At precisely the moment when a machine can produce polished, coherent, nearly flawless output in seconds — text, imagery, music, code — culture is moving in the opposite direction. Sourdough. Ceramics. Vinyl. Visible stitching. Raw materials. Imperfect finishes.

This is not a coincidence, and it is not nostalgia. It is causation.

The more AI strives for perfection, the more we are drawn to evidence that imperfection once existed. We are, whether we have named it yet or not, entering an imperfection economy.

The Machine That Cannot Fail

Generative AI optimises. Its outputs are smooth because smoothness is statistically efficient — it predicts the most likely next word, image, or phrase based on vast datasets of prior human expression. It does not hesitate. It does not cross things out. It does not sit with a half-formed idea for three days before finding the right sentence. It arrives at an answer, and the answer is coherent, and often impressive, and curiously weightless; when AI produces an output, it no longer surprises us—and, more than that, we assign it little significance. We register it, but we don’t dwell on it. When a human creates something, it carries emotional and intellectual weight; it invites a response, even engagement or resonance that AI outputs rarely command.

This is not a flaw in the system. It is the system working exactly as designed. The machine is optimising for the mean — not because it lacks intelligence, but because its intelligence is fundamentally statistical. It converges. It smooths. It finishes.

What it cannot do is fail in ways that matter.

And this, it turns out, is the thing that matters most.

Wabi-Sabi and the Cost of the Signal

There is a Japanese concept that the present moment is forcing back into focus. Wabi-sabi is often reduced in Western interpretation to an aesthetic — rough textures, muted colours, asymmetry, a cultivated kind of restraint. But this misses the depth of the idea entirely.

Wabi-sabi is not a style. It is an ontology. It holds that beauty is inseparable from impermanence, incompleteness, and wear. A crack, a stain, an uneven surface — these are not defects to be corrected. They are evidence of existence. They mark that something has been shaped by time, by use, by the world itself.

Economists have a framework for the same insight: costly signalling. A signal is credible precisely because it is expensive to fake. The peacock’s tail, the university degree, the hand-addressed envelope — all communicate something because they require genuine investment. When I look at a hand-thrown ceramic bowl, I am not simply seeing an object. I am reading a record. The slight asymmetry, the faint pressure marks, the variation in glaze — these are traces of time spent, decisions made, effort applied. They tell me that a human being was involved, that it could have gone wrong, and that it did not arrive without cost.

Synthetic perfection collapses the cost of the signal. And when the signal is free, it tells you nothing.

The practice of kintsugi — repairing broken pottery with lacquer and gold — makes this point with unusual force. The break is not hidden. It is illuminated. The object becomes more itself through honestly acknowledged damage. What has been repaired carries a depth that the untouched object never did.

AI can generate a flawless image of a cracked tea bowl. It can render the gold seam with photographic precision. But it cannot generate the history of the crack — the moment of breakage, the decision to repair, the patience required to restore. That history is not incidental to the object. For wabi-sabi, it is the object.

This is, I think, what we are groping toward when we bake bread by hand or buy vinyl records or seek out makers’ markets. We are not rejecting technology. We are trying to purchase something that technology, by its very nature, cannot produce: the record of a genuine process. As AI saturates media, communication, and creative production, people are recalibrating what they are actually paying for when they pay for anything. Increasingly, the answer is: signs of human involvement. Evidence that something was genuinely at stake. When everything is available, what matters is what is scarce.

What AI Is Revealing About Art

There are broadly three defensible positions on what AI is doing to art. The first treats AI as a tool — extending the artist’s reach, removing drudgery, expanding possibilities. The second treats it as a collaborator, capable of generating unexpected combinations that provoke new directions. Both acknowledge genuine utility.

The third position is more unsettling. It holds that AI can produce artefacts that fulfil many of the functions of art while severing the link to an authoring self. The output exists, circulates, and may be admired. But the human journey behind it is absent or minimal. No risk. No irreversibility. No cost in the making. Nothing had to be ventured.

Wabi-sabi clarifies why this matters. Art is bound to process — to time, effort, and the specific constraints of being human. A brushstroke matters partly because it cannot be undone infinitely. A live performance matters because it unfolds in real time, with a body that could stumble. AI can mimic style without having had a life. It can generate emotional signals without having experienced emotion. It can produce something that looks like the output of a journey without any journey having been taken.

That absence is becoming perceptible. Slowly, and then quickly.

The Emerging Synthesis

What I find genuinely interesting is the response some artists are already developing — not a retreat from AI, but a re-contextualisation of it.

The most thoughtful contemporary work does not reject AI-generated output. It introduces imperfection back into it. Distortion. Degradation. Physical intervention. Artists are printing, altering, layering, and breaking — using the machine’s smoothness as a baseline against which human presence becomes legible. The AI produces the substrate. The human leaves a mark.

This move is more sophisticated than either AI enthusiasm or AI resistance. It treats the machine’s optimising tendency not as a threat but as a foil — a mirror against which the specific, costly, irreversible character of human making becomes newly visible. The relationship is dialectical: the AI’s perfection is not the problem to be solved. It is the condition that makes the human gesture meaningful again.

This is already reshaping how authorship is understood. A painter who incorporates AI-generated imagery into a physically worked canvas is not diluting their practice — they are making explicit what was always true: that creative work is a negotiation between tool and intention, between constraint and possibility. The interesting question is no longer whether AI was involved. It is what the human chose to do with the encounter, and at what cost.

AI is not the enemy of craft. It is the condition that makes craft newly visible. And in its wake, the artists who will matter are not those who refuse the machine, nor those who are consumed by it, but those who learn to leave a mark on it that could not have been left by anyone else.

Leave a comment

What This Means

Every time a technology collapses the cost of something previously expensive, it forces a revaluation of what was actually being purchased. The printing press collapsed the cost of copying text; it did not collapse the value of ideas. Digital photography collapsed the cost of images; it did not collapse the value of vision. Generative AI is collapsing the cost of coherent output. It will not collapse the value of genuine human expression. It will, however, make that expression far harder to take for granted.

Culturally, craft is going to continue rising in status. The more synthetic output we encounter, the more deliberately we will seek signs of the human. Economically, the premium on the handmade will grow — not as a luxury indulgence, but as a rational response to abundance.

Aesthetically, the central question is shifting. It is no longer sufficient to ask whether something is well-made. The more pressing question is whether a human being genuinely had a stake in its making. Was there effort? Was there risk? Was there a real possibility of failure?

These are not sentimental concerns. They are signals of value in a world where surface perfection has become trivial.

The imperfection economy is not a reaction against technology. It is technology revealing, with unusual clarity, what it cannot replace — and in doing so, creating the conditions for a more deliberate and more honest kind of making.

Venture capital is often described as “fuel” for innovation. That metaphor misses the point. Capital does not create motion; it accelerates what is already in motion. The real question is not whether venture capital funds ideas, but how it shapes them—and in whose interests it operates.

Venture Capital as an Accelerator

At its best, venture capital does not originate innovation; it identifies it early and compresses the time required to test and scale it. Founders bring the insight—the non-consensus view of the future. Venture capital brings the means to test that view against reality at speed.

This compression of time is the defining function. It allows companies to move from concept to market validation, and from validation to scale, in a fraction of the time it would otherwise take. Alongside capital comes pattern recognition, access to networks, and a form of disciplined pressure.

That pressure is frequently misunderstood. It is not simply about growth at all costs. It is about forcing clarity—about product, market, team, and strategy. Markets are unforgiving; venture capital, properly deployed, is a mechanism for discovering quickly whether an idea deserves to exist at scale.

But acceleration is neutral. It amplifies both strength and weakness. Good ideas become great companies faster; flawed assumptions become expensive failures faster.

Stage Matters More Than Most Founders Realise

“Venture capital” is often spoken of as a single category. In practice, it is a sequence of distinct roles that evolve as a company matures.

At pre-seed, investors are operating in near-total uncertainty. There is often no product, limited data, and sometimes an incomplete team. The investment is fundamentally a bet on the founder and the insight. At this stage, the non-financial contribution of the investor is decisive. The best pre-seed investors help shape the problem definition, refine the initial thesis, and guide founders through the earliest structural decisions—choices that are difficult to reverse later.

At seed stage, the emphasis shifts to validation. There should be early signals of product-market fit, even if incomplete. The work becomes one of translation: turning insight into a repeatable business model. Hiring the first key leaders, establishing a credible go-to-market approach, and preparing for institutional capital all come into focus. Investors at this stage are still deeply involved in shaping the company, but with increasing attention to execution.

By Series A, the company is expected to demonstrate momentum. Metrics begin to matter in a structured way—growth, retention, unit economics. The investor’s role evolves accordingly. Governance becomes more formal, boards more active, and capital is deployed to scale what is already working rather than to discover what might work.

Beyond this point, venture capital increasingly resembles growth investing. The focus shifts to expansion, efficiency, and eventual liquidity.

The implication is straightforward: the earlier the stage, the more the value of the investor lies in what they contribute beyond capital.

The Non-Financial Edge

Capital is becoming commoditised. The differentiator is everything else.

The most effective venture investors bring a combination of judgement, experience, and access that founders cannot easily replicate:

• Pattern recognition—the ability to distinguish signal from noise in ambiguous environments.

• Founder development—helping technically strong individuals become effective leaders of organisations.

• Network access—opening doors to customers, partners, talent, and follow-on capital in global markets.

• Strategic framing—positioning companies not as local successes, but as global contenders.

• Continuity through cycles—providing perspective and stability when companies inevitably encounter setbacks.

At early stages, these contributions are not additive; they are often determinative.

At Quidnet Ventures, this is where we have chosen to operate. We bring global operating experience from the United States, Europe, and Asia, and apply it directly to New Zealand-founded companies that need to think globally from the start. We work closely with founders in a one-on-one coaching model, helping them evolve from domain experts into company builders.

We provide access to networks that extend beyond New Zealand—into the markets where these companies must ultimately compete. And we take a long view. Deep-tech ventures, in particular, do not conform to compressed timelines. They require patience, conviction, and sustained engagement.

This is not ancillary to the investment. It is the investment.

Founder-Friendly vs Investor-Friendly: A False Choice

The venture industry often frames itself as either founder-friendly or investor-friendly. It is a convenient narrative, and largely a misleading one.

Being genuinely founder-aligned is not about offering generous terms or avoiding difficult conversations. It is about engaging deeply with the company’s trajectory—supporting the founder, but also challenging assumptions early and directly.

Equally, being investor-disciplined is not about maximising control or protection. Overly engineered terms can undermine outcomes by constraining the very people responsible for creating value.

The best investors understand that long-term returns are built through alignment, not negotiation. A motivated, well-supported founder is far more valuable than a perfectly structured term sheet.

The Core Tension: Serving Two Masters

Every venture capitalist operates within a structural tension. They serve two masters:

• The founders of their portfolio companies, who require trust, support, and long-term commitment.

• The limited partners (LPs), whose capital they manage and to whom they owe fiduciary responsibility.

These obligations are not always aligned. A founder may wish to pursue a longer-term vision, while market conditions favour an earlier exit. A company may require additional capital despite uncertain prospects. Leadership changes may become necessary even when personally difficult.

These situations are not exceptional; they are routine. The simplistic notion of being either founder-first or investor-first breaks down quickly in practice.

The craft of venture capital lies in navigating these tensions with clarity and integrity. It requires making decisions that balance relationship and responsibility, ambition and discipline, long-term potential and near-term reality.

A More Mature View of Venture Capital

The venture ecosystem is evolving. Founders are more sophisticated. Capital is more abundant. The differentiation between firms is shifting away from access to money and toward quality of engagement.

For founders, the implication is clear: choosing an investor is not simply a financing decision. It is a strategic partnership that will shape the company over years.

For LPs, the message is equally direct: returns are not generated solely by deal selection. They are generated by what happens after the investment—by how effectively venture capitalists work with founders to build enduring companies.

The firms that will succeed are those that recognise venture capital for what it is: not just a source of funding, but a practice. One that combines capital with judgement, engagement, and a willingness to operate within persistent tension.

Leave a comment

Final Thought

Venture capital does not create the future. Founders do.

But venture capital determines which futures are accelerated, how they are shaped, and whether they are given the resources—and the guidance—to endure. That responsibility is not captured in a term sheet. It is exercised over time.

And it is, ultimately, what separates capital from craft.

This is the final part of a four-week series on growth, progress, and how we got them confused. Part I established that GDP doesn’t measure progress. Part II showed why better metrics haven’t changed anything—measurement is about power, not data. Part III argued that capitalism isn’t the villain; it’s an operating system that responds to the rules we set.

Now we come to the hardest question: what should New Zealand actually do, and who should do it? This week’s essay is longer because it’s where theory meets reality. It addresses housing, infrastructure, climate, distribution—and most importantly, political accountability. The core argument is uncomfortable but unavoidable: no politician deserves your vote.

Change won’t come from waiting for leadership. It comes from citizens who demand specific deliverables, track whether they happen, and vote out governments that fail to deliver. In a small democracy like New Zealand, that kind of accountability is actually possible. But only if enough people commit to it. This is where the series stops being analysis and becomes a challenge.

New Zealand is a useful place to ask.

Introduction

Small democracies operate under different conditions than larger ones. Feedback loops are shorter, policy consequences are easier to observe, and institutional failure is harder to obscure behind scale. In a country of five million people, decisions are not abstract—they are lived. New Zealand occupies an unusual position: wealthy, stable, and still retaining a degree of institutional trust that many larger democracies have already lost. It does not shape global trends, but it retains something more valuable—coherence. When it chooses to change direction, it can do so quickly. That makes it a test not of virtue, but of capacity.

In some respects, New Zealand has already moved ahead of many of its peers. By framing fiscal policy around wellbeing rather than GDP alone, it acknowledged that prosperity is multidimensional. That shift matters. But recognition is not reconfiguration. Declaring new priorities is relatively easy; governing toward them requires changing incentives. If wellbeing is to matter, it must function as a constraint on decision-making rather than simply a narrative about it. Policies that increase output while degrading long-term capability must become harder to justify, and trade-offs must be made explicit rather than deferred. Progress, in practice, requires limits as well as ambitions.

These tensions become clearest in the domains where incentives shape behaviour most directly. Housing illustrates the difference between growth and progress more starkly than any macroeconomic statistic. A system that treats housing primarily as an investment asset can generate wealth for some while eroding stability, mobility, and opportunity for others. When secure housing becomes unattainable for a large share of the population, the foundations of the economy begin to weaken. Treating housing as infrastructure rather than speculation would do more for long-term progress than almost any macroeconomic adjustment.

A similar dynamic applies to infrastructure more broadly. Transport, energy, water, and digital systems quietly determine the productive capacity of a society. When they are designed for resilience and access, they expand opportunity; when they are deferred or fragmented, they constrain it. The constraint is not technical capability but political will. New Zealand’s energy system offers a useful counterexample. Its high share of renewable electricity reflects decisions made decades ago—investments that now deliver both climate advantage and energy security. The lesson is straightforward: capability investments compound. The open question is whether today’s decisions will do the same.

Underlying all of this is a distributional reality that can no longer be deferred. For decades, economic growth substituted for distributional politics. As long as the economy expanded, questions about who benefited could be postponed. That bargain is now breaking down. Growth increasingly accrues to those who already hold capital, while risks are distributed more broadly. The result is not just inequality of income, but inequality of security, opportunity, and voice. Progress without inclusion is not progress; it is instability postponed. Addressing this does not require abandoning markets, but it does require deciding what those markets are for—and aligning incentives accordingly.

At this point, the constraint is no longer conceptual or technical. It is political. Every change described above is achievable within existing institutions. The barrier is incentive. Politicians respond to pressure and consequences, and the persistence of contradictions between rhetoric and policy reflects a simple reality: we have tolerated them. If housing affordability determined electoral outcomes, policy would change. If emissions reductions determined electoral outcomes, climate policy would accelerate. The system is not unresponsive; it is responding to the signals it receives.

What This Means in Practice

If progress depends on the incentives we set, the question becomes immediate: where is the system producing the wrong outcome—and what rule is causing it?

Start with something you can see. In housing, it might be a tax setting that rewards speculation over building. In infrastructure, it might be a funding model that delays projects everyone knows are needed. In climate, it might be the absence—or weakness—of a real price on emissions. The point is to be specific. Vague frustration changes nothing. Specific rules can be changed.

Next, ask a harder question: who benefits from the current setup—and who pays for it? Most systems persist not because they work well, but because those who benefit are organised and those who don’t are not.

From there, define one concrete change. Not “fix housing” or “reform capitalism,” but something you could recognise if it happened. Remove a distortion. Change a rule. Tie an outcome to accountability.

And then do the part most people skip: apply pressure.

Email a politician. Ask a direct question. Ask it again if the answer is vague. Support a group already working on the issue. In a country the size of New Zealand, it does not take many people—acting consistently—to shift a national conversation.

If you are not participating in that process, you are not outside the system. You are simply leaving the choice to others.

Leave a comment

The Choice Ahead

At the beginning of this series, we asked why modern societies treat growth as if it were progress. The answer is that growth was easier to measure.

But ease is not truth.

Progress is not an automatic outcome of economic activity. It is the result of choices—about what we value, what we measure, and what we are willing to insist upon. Capitalism will optimise whatever we ask it to optimise. The question is whether we are prepared to specify what that should be—and to hold the system accountable for delivering it.

The configuration of the system is not fixed. It is being chosen, continuously, through policy, incentives, and participation. That choice can be made deliberately—or it can be made by default.

That is the unfinished work of progress.

This week’s essay argues that the debate about capitalism has been asking the wrong question—and that once you understand what the right question is, the path forward becomes surprisingly clear.

The Price Tag You Were Never Shown

In Auckland, a teacher in her mid-thirties has done everything right. She’s educated, employed, contributing. She has been saving for a first home for six years. She is further away from buying one than when she started.

This is not a market failure. The market is functioning exactly as designed. It’s just that nobody asked her—or you—what the market should be designed to do.

That distinction is the entire argument of this essay.

An Operating System, Not a Moral Agent

By the time the conversation turns to capitalism, the temperature usually rises. Growth has disappointed. Measurement has failed to capture what matters. Inequality feels entrenched, the climate unstable, and younger generations increasingly alienated from a system they have inherited but did not design.

At that point, capitalism is often treated not as a mechanism, but as an autonomous force—an engine producing outcomes that feel unchosen and beyond control.

This reaction is understandable. It is also mistaken.

Capitalism is not a moral agent. It does not have intentions, values, or goals of its own. It is an economic operating system: a set of rules governing ownership, exchange, risk, reward, and coordination. Like any operating system, it produces outcomes consistent with how it is configured. When those outcomes are destructive, the cause lies not in the existence of the system, but in the parameters we have chosen to set—or failed to set—around it.

Treating capitalism as an immutable villain is not just analytically wrong. It is politically disabling, because it makes the system seem unchangeable when in fact it is changed all the time—just usually by people other than you.

How We Got Here

Over the past half-century, many advanced economies converged on a particular configuration of capitalism: one that elevates growth above all other objectives, tolerates extreme concentration of wealth, externalises environmental costs, and treats distribution as a secondary concern.

This configuration did not emerge naturally. It was designed—incrementally and often implicitly—through tax policy, labour law, financial regulation, trade agreements, corporate governance norms, and political choices about what should be constrained and what should be left to “the market.” The system has behaved exactly as instructed.

Growth became the master signal. Firms that expanded were rewarded; those that stagnated were punished. Executives were compensated for revenue growth, market share, and share-price appreciation. Governments were judged on headline economic performance. Environmental degradation, social fragmentation, and long-term risk were discounted because they didn’t register cleanly in the metrics that mattered.

From the system’s point of view, this was not failure. It was optimisation.

The Degrowth Trap

This helps explain why debates about growth and degrowth so often become circular. Advocates of degrowth are responding to real constraints: ecological limits, planetary boundaries, and the social exhaustion that comes from perpetual acceleration. Defenders of growth point, not unreasonably, to the role expansion has played in reducing poverty, funding public services, and sustaining political stability.

Both sides are partly right—and both are incomplete—because they treat growth as a binary choice rather than a conditional instrument.

Growth is neither inherently virtuous nor inherently destructive. Its value depends on what is growing, where, for whom, and at what cost. What is missing from much of the debate is a serious confrontation with distribution.

For decades, growth functioned as a substitute for distributional politics. As long as the economic pie expanded, questions about how it was divided could be deferred. That bargain has broken down. In many societies, growth now accrues disproportionately to those who already hold capital, while risks are socialised and gains privatised. The result is not just inequality of income, but inequality of security, opportunity, and voice.

Progress without inclusion is not progress. It is instability postponed.

What Reconfiguration Looks Like

This is where the operating-system metaphor becomes useful. Capitalism does not dictate outcomes; it amplifies incentives. If the rules reward extraction, it will extract. If they reward long-term stewardship, it will adapt. If environmental costs are priced, behaviour changes. If monopoly power is constrained, innovation re-emerges. If labour has bargaining power, productivity gains are shared.

None of this requires abandoning markets. It requires governing them.

The challenge is not to dismantle capitalism but to re-specify its objectives. If progress is defined narrowly as aggregate output, growth will dominate and distribution will remain secondary. If progress is defined as human capability, ecological stability, social trust, and intergenerational continuity, then growth becomes conditional—encouraged where it supports those ends, constrained where it undermines them.

This shift has practical implications. It suggests moving away from universal growth targets toward sector-specific strategies. It implies treating infrastructure, housing, healthcare, and energy differently from speculative finance. It requires measuring success not only by how much value is created, but by how resilient, inclusive, and durable that value proves to be.

Some forms of growth will need to slow or stop. Others will need to accelerate. Consumption patterns will change. Distributional conflicts will surface. Pretending otherwise is how trust erodes. But abandoning growth as a moral objective does not mean abandoning ambition. It frees ambition from a single dimension and allows societies to aim for better, not just more.

The Deepest Irony

Capitalism is often criticised for being too powerful, when in fact it is under-governed. We have been reluctant to specify what the system is for, and then surprised when it delivers outcomes misaligned with our values.

Growth is a tool. Measurement is a means. Capitalism is infrastructure. None of them defines progress on its own. Progress emerges from choices made above the system—about purpose, constraint, and distribution—and from the willingness to revisit those choices as conditions change.

The configuration of the system is a choice. Right now, it is being made without you. That can change.

What This Means—and What You Can Do

The Core Problem

We’ve set up an economic system to optimise for growth at any cost, then acted surprised when it delivers exactly that—growth, regardless of whether it improves lives, depletes resources, or concentrates wealth. The problem isn’t markets or capitalism as concepts. It’s that we’ve been unwilling to set rules that align market incentives with actual human wellbeing.

Why This Matters

Once you understand capitalism as an operating system rather than a force of nature, you realise it can be reconfigured. Different rules produce different outcomes. The question shifts from “capitalism: yes or no?” to “what specific rules will produce the outcomes we actually want?” That’s a question we can actually answer—and answer we must, because someone is answering it right now.

Proof That This Works

New Zealand’s Accident Compensation Corporation (ACC) is a functioning example of capitalism reconfigured. By tying levies to injury rates, it turned safety into a market incentive. Workplace injuries fell significantly over decades. The system didn’t change—the rules did. The Emissions Trading Scheme, building energy codes, and the Reserve Bank’s housing risk weights are further examples: each is a policy decision that changed what the market optimises for. None required abolishing anything.

Leave a comment

What You Can Do This Week

Step 1: Name the Misaligned Incentive

The next time someone says “capitalism is broken,” you’ll have something more useful to say—something specific enough to actually embarrass a politician with.

Pick one example from your work, community, or observation:

• Does your organisation reward short-term gains over long-term value?

• Does your tax code give advantages to property speculation over productive investment?

• Does regulation prevent good outcomes while allowing harmful ones?

• Do executive pay structures reward revenue growth regardless of social or environmental cost?

Write it down specifically. Not “capitalism is bad” but “our tax code gives a 30% advantage to property speculation over business investment.”

Step 2: Map Who Benefits and Who Loses

• Who wins from the existing policy, rule, or incentive?

• Who loses?

• What coalition of interests would be needed to change it?

• What’s the political obstacle—ideological, or is it that beneficiaries are organised and losers aren’t?

Step 3: Identify the Specific Fix

Not: “Fix capitalism” or “Make housing affordable.”

Yes:

• Remove mortgage interest deductibility for investment properties

• Require full lifecycle cost reporting for infrastructure projects

• Price agricultural emissions at the same rate as industrial emissions

• Allow missing middle housing (3–4 storeys) in all residential zones

• Tie executive compensation to 5-year outcomes, not quarterly earnings

The more specific, the better. You should be able to tell whether it happened or not.

The One Action That Actually Matters

Email a politician. Literally. Today.

Use this template:

“I’m a constituent and I want you to commit to [specific change] because [specific reason]. Will you commit to introducing or supporting [specific legislation or regulation] by [specific date]? Yes or no?”

If they respond with vague promises, reply with the same question. Keep asking until you get a yes, a no, or a silence that tells you everything you need to know.

Going Deeper

• Read your local council’s planning rules—see what they incentivise and prohibit

• Look at political party policy platforms—not rhetoric, actual commitments

• Save candidates’ campaign promises. You will need them to hold them accountable

• Find one organisation working on the misaligned incentive you identified and follow their work

Examples of Well-Aligned Incentives to Study

• ACC (incentivises injury prevention through experience-based levies)

• ETS (prices carbon emissions to incentivise reduction)

• Building code energy requirements (make efficiency mandatory, not optional)

Coming in Part IV

The final part brings it all home. We’ve established what’s wrong (confusing growth with progress), why better metrics haven’t fixed it (measurement is about power), and where agency actually sits (in the rules we set for capitalism). Part IV addresses what we should specifically do—and more importantly, who should do it.

Spoiler: it’s not politicians. It’s us. Citizens. Through sustained engagement and political accountability. No politician deserves your vote. Make them earn it.

Part IV is the conclusion—but also the beginning. Check back next Wednesday for the action plan.

Editor’s note: This is Part II of a four-part series on growth, progress, and governance.

Part I examined why GDP does not measure progress.

This essay explores why better metrics alone rarely change outcomes.

Next week’s essay asks what role capitalism actually plays in shaping economic incentives.

Last week, we established that GDP and progress aren’t the same thing—that we’ve been optimising for a measure that doesn’t capture what we actually care about. The obvious next question: if not GDP, then what?

This week’s essay argues that better metrics already exist—and that they haven’t changed anything. Understanding why is the most important question in public life that almost nobody is asking.

Your Council Spent Millions Last Year. Do You Feel It?

Your local council spent tens of millions of dollars last year. Central government spent tens of billions. They published annual reports. They held press conferences. They declared progress.

Did your commute get shorter? Did the emergency housing waitlist get smaller? Do you know a young person who can now afford to buy a home?

If you can’t answer those questions—and you probably can’t, because the reports don’t make it easy—then you have just encountered the central problem of modern governance. It isn’t that governments don’t measure things. It’s that they measure the things that are easy to defend, not the things that matter to you.

That choice is not accidental. And it is not inevitable.

The Measurement Problem Is a Power Problem

Once the distinction between growth and progress is acknowledged, an obvious question follows: if GDP is not the right proxy, what should replace it?

For decades, this question has attracted earnest answers. Indicators have multiplied. Dashboards have become more elaborate. Composite indices now promise to capture wellbeing, sustainability, inclusion, resilience, and opportunity in a single frame. The problem is no longer a lack of alternatives. It is that none of them has displaced GDP as the organising principle of public decision-making.

This is not because better measures do not exist. It is because measurement is never neutral.

How GDP Became Untouchable

GDP succeeded not simply because it was measurable, but because it was legible, comparable, and politically useful. It reduced a complex society to a single figure that could be tracked quarterly, compared internationally, and tied directly to fiscal and monetary levers. It made governments governable in a statistical sense. Any effort to move beyond it must contend with that inheritance.

There is a familiar adage: we manage what we measure. It is usually offered as managerial common sense. In practice, the causality often runs the other way. We measure what we can manage—and then allow those measures to define what management itself means.

GDP and growth rose to dominance not because societies first agreed that economic expansion was the essence of progress, but because expansion was tractable. It could be counted, standardised, audited, and communicated. A single number could anchor a press conference, justify a budget, or frame an election. It could be embedded in rules, targets, and incentives. It travelled easily through institutions.

This logic took hold in the middle of the twentieth century, as states emerged from the Second World War facing the immense task of reconstruction and coordination. Governments needed ways to see their economies as systems rather than collections of firms and households. National income accounting provided that vision. Statistics became a form of statecraft.

What could be quantified could be controlled; what could be controlled could be improved. In that context, GDP was not merely an economic indicator. It was an instrument of governance. Over time, proof of growth hardened into purpose.

The SDGs: Moral Clarity, Operational Paralysis

Against this backdrop, the United Nations Sustainable Development Goals represent a deliberate attempt to widen the frame. Adopted in 2015, the SDGs explicitly reject the idea that progress can be reduced to economic output alone. They articulate a multidimensional vision of human advancement, spanning poverty, health, education, gender equity, environmental protection, and institutional trust.

Their strength lies in moral clarity. They make explicit what GDP leaves implicit: that development is about human lives, not just economic throughput.

Their weakness is that they describe what matters without determining what governs. Seventeen goals, dozens of targets, and hundreds of indicators diffuse responsibility rather than concentrate it. Governments can endorse the framework rhetorically while continuing to optimise for growth in practice.

The result is performative measurement: reports are published, dashboards updated, progress declared—while underlying incentives remain entirely unchanged. This is not a technical failure. It is a political choice dressed up as administration. Governments learned that the easiest way to respond to demands for better metrics was to produce more metrics—not to change what they were actually rewarded for delivering.

The Wellbeing Budget: When the Language Changes but the Logic Doesn’t

New Zealand’s attempt to embed wellbeing directly into fiscal decision-making went further than most. The Wellbeing Budget was a serious conceptual intervention—an explicit effort to loosen GDP’s grip on the state by reframing policy objectives around mental health, child poverty, environmental stewardship, and social cohesion.

You may remember when it was announced. The coverage was extensive. The ambition was real.

You were told the budget had changed. The GDP growth forecast was still the first number on every front page.

Budget processes remained shaped by short political cycles, departmental silos, and the enduring authority of traditional economic indicators. Wellbeing metrics informed narrative and justification but rarely displaced growth as the binding constraint. The language of policy shifted faster than its logic. Progress was reported. The underlying system kept optimising for what it had always optimised for.

This pattern is instructive. Better metrics do not automatically produce better decisions because metrics do not operate in isolation. They sit inside systems of accountability, power, and incentive. What is measured matters—but who is rewarded, punished, or re-elected on the basis of that measurement matters more.

The Deeper Problem

No single metric can tell us whether we are progressing. Measurement can inform judgement, but it cannot replace it. Until we clarify what we actually want from our institutions—and hold those institutions to account for delivering it—measurement will remain a tool for managing appearances rather than producing outcomes.

The dashboard doesn’t lie. It just only shows you what someone decided to count.

The question is whether you’re going to let someone else make that decision.

What This Means—and What You Can Do

The Core Problem

We have better metrics than GDP. They exist. They’re well-designed. But they haven’t changed how decisions actually get made, because measurement alone doesn’t shift power or incentives. The Wellbeing Budget changed the story politicians told. It didn’t change what they optimised for.

Why This Matters to You Specifically

Your council chooses what to measure. That choice determines what problems they are even capable of seeing. Right now, you are not part of that choice. The action steps below are about changing that—not someday, but this week.

Leave a comment

What You Can Do This Week

Step 1: Find Out What Your Council Actually Measures

Find your council’s annual report or Long-Term Plan (available on their website). Look for:

• What do they actually measure and publicly report?

• Do these metrics reflect what matters to your community?

• What is glaringly missing? Housing affordability? Commute times? Mental health service access? Environmental indicators?

If you can’t find clear, comparable numbers on the things that affect your daily life, that absence is itself the answer.

Step 2: Submit Feedback on What Should Be Measured

Most councils have consultation periods for Long-Term Plans and Annual Plans. Find the next one and submit. It can be a short email—it doesn’t need to be formal:

• “Why don’t you publicly report on housing affordability relative to median income?”

• “How do you measure whether our community is actually more resilient or healthier than last year?”

• “What would it take to make [your priority] a reported metric with targets?”

Submissions are read. Councils are legally required to consider them. Most people don’t submit. Yours will stand out.

Step 3: Demand Specifics From Politicians

Next time any politician—local or national—claims success, ask:

• “What specific metric improved, and by how much?”

• “Who specifically benefited, and who didn’t?”

• “What trade-offs were made to achieve this?”

Don’t accept:

• “The economy is strong”

• “We’re making progress”

• “We’re committed to wellbeing”

Do accept:

• “Child poverty decreased from X% to Y%”

• “Emergency housing waitlist reduced by Z people”

• “Emissions fell by W% from the 2019 baseline”

The Six-Month Test

Download your council’s current performance report today. Set a calendar reminder for six months from now to pull it up again.

If nothing has changed—not the numbers, not even the questions they’re asking—you’ll have learned something important about who your council is actually accountable to.

Then do something about it.

Going Deeper

• Read your council’s Long-Term Plan consultation document (released every 3 years)—see what they commit to measuring and why

• Look up New Zealand’s Wellbeing Budget reports on the Treasury website; note which metrics are tracked versus which actually drive decisions

• Find out when your council next opens public submissions and put the date in your calendar now

Coming in Part III

We’ve established that growth isn’t progress, and that better metrics don’t automatically change outcomes. So what’s actually driving the system? Part III examines capitalism’s role—not as a villain, but as an operating system that responds to the rules we set. The question isn’t whether to keep or scrap capitalism. It’s whether we’re willing to specify what it’s for.

Part III drops next Wednesday. If you found this useful, share it with someone working in local government or policy—they’re the ones who set these metrics.

Editor’s note: This essay is Part I of a four-part series on growth, progress, and economic governance.

Part II examines why better economic metrics rarely change policy outcomes.

Part III explores capitalism as an operating system shaped by incentives.

Part IV asks what citizens and governments can actually do to govern for progress.

This is the first in a…

Read more

TL;DR:

If you’ve ever thought “my vote doesn’t matter” or “it is what it is” — you’re not wrong, but you’re blaming yourself for a design flaw. The system we call democracy is wired so that your voice fades to a whisper long before it reaches the decisions that shape your life. That’s not inevitable. It’s an architecture, and architectures can be rebuil…

Read more

In the first essay in this series, I argued that sustained reading was not merely a cultural habit, but an enabling technology of modern civilisation. It trained habits of thought—patience, argument-following, internal judgment—that democratic self-government quietly depends on. In the second, I suggested that the real existential risk posed by artifici…

Read more

Part II Introduction

Much of the public conversation about artificial intelligence focuses on dramatic risks: runaway systems, misaligned goals, or machines that decide humanity is no longer necessary. These scenarios dominate debates about AI safety and existential threat.
The first part looked backwards, arguing that sustained reading functioned as an …

Read more

Series Introduction & Author’s Note

This is the first essay in a three-part series on artificial intelligence, cognition, and democracy—not focused on what machines might do to us, but on what we may quietly stop doing ourselves.

I write this as someone who uses AI tools extensively and values them highly. I rely on them to draft, organise, and interrogat…

Read more

Carney is right: nostalgia is not a strategy—and pretending otherwise is now dangerous.

We are in the period of chaos that precedes the emergence of a new paradigm, as Thomas Kuhn explained in his seminal work, “The Structure of Scientific Revolution”

At Davos this week, Canadian Prime Minister Mark Carney did something most middle-power leaders still av…

Read more

For a couple of years now, I’ve been using ChatGPT and other large language model–based systems almost daily. Not experimentally, and not casually, but as working tools. I use them to draft and refine documents, to collect and organise information, and to build small but genuinely useful applications for myself—things that compress time, reduce friction…

Read more

Every December I take stock. Not in the performative way of annual reviews, but in a harder register: What did I try to build? What moved? What resisted? And what future am I implicitly betting on?

This reflection carries particular weight for me because Aotearoa is not where I was born. It is where I chose to belong.

I came to New Zealand late in life, a…

Read more

My most recent post concerned the role of economics. In the lead-up to the long holiday here in Aotearoa, I have been adding to my reading list, which has led me to share more thoughts in this post.

We need a democracy in which corporations are not treated as people and people are not treated as numbers.

Democracy and capitalism are inseparable. One cann…

Read more

This is a conversation worth having—before we lose faith in the whole thing.

We hear about “the economy” every day. Headlines talk about GDP growth, inflation, interest rates, unemployment numbers, and bond markets. The numbers rise and fall like a scoreboard. But for many people, they don’t match the lives we’re actually living.

Groceries keep getting mo…

Read more