More articles now on substack: https://andersindset.substack.com/

Seven hundred global executives and the game of the year at the Puskás Aréna. Humbled by the many reflections after the Saturday summit in Budapest, I sat down yesterday and wrote out the concepts of Possibilism and Anticipatory Leadership expanded on Saturday’s keynote. See this as a Leadership Piece on how to hold a course through it. Please let me know what you think.

Getting the forecast right used to be the most valuable thing a leader could do. The era that rewarded it is ending. What replaces it is something we have undertrained, underbuilt, and quietly punished out of our institutions: perception.

This is the case for a different posture, one I relate to as Possibilism, and for what it asks of leadership now.

Why Possibilism

Most of what passes for serious conversation about the future today collapses into one of two failure modes.

On one side, naïve optimism. The assumption that technology will sort itself out. That markets will correct. That progress is linear and benign. On the other, dystopian negativism. The conviction that we are sleepwalking into collapse. That AI ends us. That institutions are finished. Both are evasions, they let leaders off the hook by treating the future as something that happens to us rather than something we make.

Possibilism is the third stance. Its premise is that nothing about the next decade is settled in advance: the future is the residue of choices not yet made, and the highest-leverage move a leader has is to widen the field of what those choices can be, to author the future instead of waiting for it.

That is the whole argument in a sentence. To me “Future” is a verb, to future, it is something we create, do, or even write. The rest of this piece is how.

The future is not a forecast. It is not a fate. It is authored.

From Prediction to Perception

For the last thirty years, roughly the era of modern strategic management, the most valuable skill in finance and business was prediction. Prediction worked because the world it modelled was relatively stable. Inflation behaved. Globalisation deepened in one direction. Interest rates moved within bands. Consumer behaviour was patterned. Even disruptions like 2008 and COVID were the kind of shocks that pattern recognition could eventually digest. In that world, prediction had real value. Better forecasts meant better hedging, better capital allocation, better trading. An entire stack of finance, derivatives, structured products, much of asset management, was built on a single assumption: that the future is a probability distribution over outcomes that resemble the past.

That assumption is now contested. And the deeper change beneath it is one I think we are only beginning to take seriously: today, management is technology. The functions that used to require trained human managers, coordination, monitoring, performance review, much routine decision-making, are being absorbed by software and AI. Globalisation is regionalising. Geopolitical pressure is splintering financial infrastructure. The marginal cost of analysis is collapsing toward zero. The future no longer looks like a smooth extrapolation of the past. It looks like a series of regime shifts, each one rewriting the rules of the previous one. In short, Management is dead. As a human skill. And, in this world, the strategic edge is not prediction. It is perception.

The distinction matters, so it is worth slowing down.

Prediction tells you what is likely to happen next, conditional on current patterns continuing. Perception tells you that the pattern itself is changing, before the change shows up in the data. Prediction is what you do with a model. Perception is what you do before you have one.

This is also, I think, the right way to read the companies that have actually defined the last two decades. Apple’s iPhone. Amazon’s AWS. Tesla’s electric car. SpaceX’s reusable rocket. OpenAI’s GPT. None of these were founded on a forecast of what the market wanted. They were founded on a perception of what the world was already becoming, before the trend lines confirmed it. The dominant companies of any given era are usually those that perceived a structural shift earlier and committed to it harder. They weren’t riding the curve so much as drawing it.

A predictor in 2007 would have given you confidence intervals on housing prices. A perceiver would have noticed that the loans being written no longer resembled the loans being modelled, and that the assumptions inside the models had quietly hollowed out.

The institutions that win the next decade will not be those with the best forecasting machinery. They will be the ones whose senior teams have invested in perception. In trained noticing. In unhurried thinking. In the capacity to see regime change before it announces itself.

Prediction is what you do with a model. Perception is what you do before you have one.

Three Forces Reshaping the World

Perception, in 2026, has to be calibrated against three structural forces that are working at once.

Force One: Regulatory Regionalisation. Money is where you see this first and most starkly, but the same pull runs through supply chains, data, and technical standards. For most of the last thirty years, financial infrastructure trended toward one global system. SWIFT for messaging, the dollar for settlement, Visa and Mastercard for cards. The exceptions, gold standards, exchange controls, regional payment networks, looked like residue from an older era.

That trend has reversed. We now live in a world of regional rails connected by interoperability layers. India built UPI, processes more transactions per month than the rest of the world combined, and is exporting the architecture to Bahrain, France, Singapore, Mauritius. The European Union is rolling out instant payments and weighing a digital euro. China has CIPS, its own SWIFT alternative, and the e-CNY. The United States is formalising stablecoins, effectively building a private dollar rail alongside the public one. Russia, Iran, and others have spent ten years building sanctions-resistant infrastructure.

This is not a temporary fragmentation that will reverse when geopolitics calms. It is a structural rearrangement. The world of one global rail is becoming a world of many, with interoperability as the contested layer. Cross-border is no longer a back-office line item. It is a geopolitical posture.

Force Two: Trust as Geopolitics. Trust has always been the implicit backdrop of every institution, and of financial infrastructure most of all. We didn’t have to think about it, because we shared a deep enough consensus that it could be assumed. That assumption has broken.

We saw it in 2022. When Russia was disconnected from SWIFT, what was revealed was not just a technical capability, it was that financial infrastructure had become a geopolitical weapon. Every country that watched that happen, friends and rivals alike, took the same lesson: do not depend on a system you do not control. The result is a steady migration toward financial sovereignty: data localisation laws, sovereign cloud, the rebuilding of basic financial plumbing inside national borders.

Banks have always been keepers of trust. That role, which felt almost invisible for decades because it was so deeply embedded, has become one of the most exposed and contested positions in the global economy. For institutions whose business model is, at its core, trust at scale, this is the headline story of the decade. Trust is no longer a free input. It is a contested asset.

Force Three: Technology as Substrate. AI is not a feature. It is not a project. It is becoming the substrate of decisioning, fraud detection, underwriting, compliance, customer interface, and increasingly product itself. And it is only one part of a deeper technological maelstrom, compute, energy, and quantum compounding against each other in the same window of years.

Institutions that treat AI as a programme to manage will be outpaced by those that treat it as the operating system to build on. The question is no longer whether to adopt AI, or even where to adopt it. The question is what to organise around it. If every decisioning function in the firm is going to be AI-mediated, the organisational design, the hierarchy, the roles, the metrics, has to change too. The next decade of competitive advantage, in finance and well beyond it, will be defined less by who has the best models, and more by who has restructured the firm around them.

Trust is no longer a free input. It is a contested asset.

Three Curves We Are All On

These three forces don’t operate in isolation. They are accelerated by a deeper structural fact, the curves we have been on for decades, and that almost nobody intuitively feels.

Curve One: Computation. Roughly eighty years of price-performance in computation, plotted on a logarithmic axis, produces an almost perfectly straight line that climbs steeply across three different paradigms, vacuum tubes, transistors, integrated circuits. The number that matters: every dollar buys roughly a trillion times more compute today than in 1945. A trillion. Not double. Not ten times. A trillion. And the line is still climbing.

Human cognition is linear. Our intuition about the future extrapolates from recent experience. When something compounds exponentially for long enough, two things follow. First, we systematically underestimate what is possible at the leading edge. Second, we are repeatedly surprised by capabilities that arrive years before our planning cycle expected them. AI capabilities in 2026 are running roughly a decade ahead of what the consensus forecast in 2018 would have allowed. That gap is the standard error of a linear forecaster looking at an exponential process.

Curve Two: Energy. The renewable electricity chart is the same shape as the computation chart, on a different timeline. In 1965, the line is invisible. In 1995, still essentially flat. By 2010, it begins to climb. By 2020 it is at three thousand terawatt-hours and accelerating. In a single human career, wind and solar moved from rounding error to industrial scale.

Solar has its own version of Moore’s Law, Swanson’s Law, where the cost of a photovoltaic module falls roughly twenty percent for every doubling of cumulative production. That has held for forty years. The result: solar is now the cheapest electricity in human history in most of the world, under coal, under gas, under even the nuclear we never built.

And it changes more than the price of a kilowatt-hour. The intuition that says distributed systems are fragile is being inverted. Decentralisation is stabilisation. Every household with solar and a battery is a node in a network that, at scale, becomes more resilient than the centralised grid it replaces. Houses charged from cars. Cars charged from roofs. Industrial loads optimised in real time against the grid’s actual state. Germany already pays significant redispatch costs because the legacy grid wasn’t built for this. Those costs are not waste. They are the visible price of a transition the underlying physics has already won.

Energy is the substrate of everything material in the economy. When the substrate transforms, capital allocation has to follow. Anything that runs on electricity, which is increasingly everything we want more of, including AI, inherits the curve.

Curve Three: Forecasts vs Reality. The third chart is, in some ways, the most revealing. Every line that fans out on the lower part of the picture is an official five-year forecast for global solar installations. The line that climbs above all of them is reality. Actual installations have beaten the consensus five-year forecast by roughly three times, on average, year after year, for two decades.

These are not amateur predictions. The IEA produced many of them. These are the most carefully researched, most institutionally credible energy forecasts in the world. And they have been systematically wrong, in the same direction, for two decades.

Why? Because forecasters are trained to be conservative, and conservatism is the right posture for short-term, linear systems. For an exponential system, conservatism becomes a systematic bias. The forecast assumes the curve will moderate. The curve doesn’t moderate. The forecast undershoots. Repeat.

This is not just a story about solar. It is the story of every exponential adoption curve. Mobile phones in the 1990s. Internet in the early 2000s. Smartphones after 2008. AI capability since 2020. Every one of them was undershot by sober institutional forecasts at the time. Every one of them rewarded the actors who built strategy for the curve, not for the dot.

If you are running any sizeable organisation today, you are almost certainly making capital allocation decisions, hiring plans and product roadmaps against forecasts that are conservative in this same systematic way. The default planning posture undershoots the future. Anticipatory leadership starts with that recognition. You don’t solve it with better forecasts. You solve it with a different posture toward the curve.

Build for the curve, not the dot.

The Diagnosis: Knowledge, Intelligence, Understanding

Underneath all of this sits a deeper diagnosis. It is, for me, the philosophical heart of the whole argument, and the moment I have seen most senior leaders feel something shift, because it names a discomfort they have been carrying without language for it.

Knowledge is facts, data, information. Wikipedia. The papers. Everything an LLM can quote. Knowledge is now effectively infinite and effectively free. Anyone with a phone has access to more knowledge in five seconds than the entire Library of Alexandria contained.

Intelligence is the capacity to process knowledge. To synthesise it, to apply it, to act on it. Intelligence used to be relatively scarce, it lived in trained human minds and required years of education to build. That is changing fast. AI systems are now processing, synthesising and applying knowledge at scales no human can match. Intelligence is becoming abundant.

Understanding is something else. Understanding is the capacity to know what to do with knowledge and intelligence, why it matters, what it implies, what to act on, what to leave alone. Understanding requires judgement. It requires context. It requires the kind of long-form attention that the rest of the information environment now actively erodes.

I want to be honest about something here. AI systems are getting better at what looks like understanding, too, connecting context, raising relevant judgement, surfacing what matters. The frontier is moving. But for now, and for the foreseeable horizon of senior decision-making, understanding remains the binding constraint. It is what humans are still uniquely positioned to provide.

The shift looks like this. For most of human history, the binding constraint on good decisions was knowledge. If you knew more than the next person, you could decide better. So we built institutions to gather knowledge, libraries, universities, research departments, intelligence agencies, consultancies. We spent thirty years building what we called the Knowledge Society. Information was digitised. Search became free. The argument was that knowledge would set people free. It has done many things. It has not, on its own, made us wiser.

Then, for the last forty years, the binding constraint shifted to intelligence. Knowledge was abundant, but the capacity to do something with it was scarce. We built institutions for intelligence too, analyst desks, strategy teams, the entire architecture of modern professional services.

We are now entering an era where neither knowledge nor intelligence is the binding constraint. Both are abundant, and both are getting cheaper by the month. The new binding constraint is understanding. The capacity to know what is worth doing in the first place.

This is uncomfortable for senior leaders specifically. Because the experience that built their career, the pattern recognition, the gut feel, the seasoned intelligence, was built for the previous regime. In a world where the patterns themselves are changing, experience can quietly become a liability. The defence isn’t new tooling but a new posture — investing in understanding while everyone else is still optimising for intelligence.

Knowledge is abundant. Intelligence is accelerating. Understanding is scarce.

What Are We Becoming?

Which raises an uncomfortable question. If the machines are taking knowledge and intelligence, what is left for us? The image I used on stage at this point in the talk is deliberately ambiguous.

A human profile composed of solar panels, wind turbines, fragments. Birds dissolving outward. Is the figure being built, or being undone? That ambiguity is the point. The technologies that promise to free us are reshaping what it means to be human in the first place. They are not external tools we pick up and put down. They are environment. They are increasingly the substrate of how we think, communicate, choose, work, and relate.

We are building systems that out-perceive us in specific domains, fraud patterns in milliseconds, market microstructure, image recognition. We are building systems that out-process us at scale, every category of human cognitive labour is being recoded. And we are increasingly building systems that out-create us in narrow ways, generating, drafting, designing, composing.

Which leaves a question that finance and technology cannot answer alone. What is the human contribution in this picture? It is not raw knowledge. It is not raw intelligence. So what is it?

My answer, after some years of working on this, is twofold. The human contribution is judgement about what kind of future is worth striving for. And the agency to build it. What do we want to build, and how? That is the question machines cannot answer for us, and the question Possibilism puts in front of every leader.

The technologies that promise to free us are reshaping what it means to be human.

Anticipatory Leadership: Three Anchors

That is also where Possibilism stops being a posture and starts being a practice. The translation from stance to work is what I have come to call Anticipatory Leadership. This is my little imprint of Possibilism, my attempt to name the capabilities a leader actually needs in this moment, and the ones I keep watching the best of them quietly develop.

It rests on three anchors. They hold up in volatility precisely because they do not depend on prediction.

Anchor One: Perception. Perception is the trained capacity to see what others do not yet see. It isn’t psychic, and it isn’t prediction. It is what a skilled fraud investigator does when she notices a cluster of anomalies in transaction patterns three weeks before the pattern becomes a typology. It is what a good central banker does when she catches something off in money market spreads before it surfaces in the headline indicators. It is what a seasoned operator does when a supplier’s tone shifts weeks before the delivery problem reaches the numbers. It is the ability to weight small signals correctly when the world is noisy.

Perception requires three conditions that most organisations no longer create. Time, unhurried attention to weak signals. Exposure, senior people who actually look at primary data, customer complaints, fraud disputes, regulatory chatter, not just dashboards. Permission, people paid to notice, not just to deliver. In most institutions today, the cost of noticing is high. It slows things down. It makes meetings harder. It raises uncomfortable questions. So perception atrophies in the very organisations that need it most.

Who in your organisation is paid to notice rather than to deliver? If the honest answer is no one, that absence is already your strategy.

Anchor Two: Meaning-Making. Meaning-making is the capacity to turn noise into narrative. In volatility, the most strategic act a senior leader can perform is offering coherence. People, and capital, do not move because they have data. They move because they have meaning. The same is true for boards, investors, regulators, and the markets you depend on.

Meaning-making is not spin. It is the disciplined work of finding the structure underneath the events. Why is this happening, what does it imply, what does it ask of us. The best CEOs of the last decade have been those who could metabolise complexity into a story their organisation could act on. Not a simplification, a compression. A story sharp enough to guide decisions, and honest enough to update when the world updates.

Somewhere in your current strategy is a sentence that still assumes the last cycle. That sentence is the hole capital will eventually find.

Anchor Three: Directionality. Directionality is conviction without certainty. It is the discipline of committing to a horizon you cannot prove, then acting on it, then updating as evidence comes in, then acting again. It is venture-style behaviour applied to the operating firm. Portfolio thinking. Willingness to be early. Tolerance for being wrong in specific bets while right about the direction.

The opposite of directionality is not caution. It is the institutional habit of presenting strategy as a plan to defend, rather than a hypothesis to evolve. Boards reward the appearance of certainty. Defenders of plans are protected. Updaters of hypotheses look indecisive. That cultural pattern is now actively dangerous, because the speed of regime change means strategies need to evolve faster than annual planning cycles allow.

At your next board, present the strategy as a hypothesis to evolve, not a plan to defend, and watch who flinches. The flinch is the culture you are fighting.

Perception sees what is forming. Meaning-making turns it into a story. Directionality moves on it. Take any one out and the others collapse. Perception without meaning-making is paralysing data. Meaning-making without directionality is great narrative with no movement. Directionality without perception is conviction in the wrong direction. The three together are anticipatory leadership.

People, and capital, do not move because they have data. They move because they have meaning.

The Equation: Engines × Culture

Anticipatory leadership is the inside of the firm. What sits around it, what makes it possible at the scale of an economy, is a deeper model of progress itself. The model is simple and, I think, useful.

Progress, in any economy, is the product of two things. Engines and culture. Engines are the resources, the raw material. Culture is the conditions, the environment that lets the resources actually do something.

Engines are talent plus capital. By capital I specifically mean free cash flow, the kind of money that can move at the speed of frontier opportunity, not the kind that sits in pension pools optimising for liability matching. The people who can build, and the money that lets them.

Culture is trust plus friction. The trust that allows commitments to hold across time and distance, and the productive friction that keeps thinking honest.

Talent without capital builds toys. Capital without talent builds nothing. So far, so obvious. That is the first half of the equation, and it is where most economic conversation stops.

But engines on their own do not move. Look at any economy with enormous talent and enormous capital that nonetheless stagnates. The missing variable is always cultural. Specifically, two cultural variables.

Trust first. Trust is what lets capital flow toward people whose work product can’t be verified in advance. It is what lets institutions enter long-term commitments. It is what lets one generation invest in the next. Where trust is high, transaction costs collapse and time horizons extend. Where it is low, everything has to be enforced contractually, and the dead weight kills the system.

Friction second. And here I want to be careful, because friction has a bad reputation in business culture. Most of what people call “friction”, bureaucratic, procedural, political friction, is parasitic and worth optimising away. But there is a different kind of friction. Productive friction. The friction of honest disagreement. The friction of someone willing to push back on the consensus. The friction of ideas tested against other ideas in good faith. That friction is not the enemy of progress. It is its immune system. Without it, organisations and economies tell themselves the same comfortable stories until reality breaks them.

That is why the equation multiplies rather than adds. Engines without trust go nowhere. Trust without friction stagnates. Friction without engines just argues. You need all four.

Friction is not the enemy of progress. It is its immune system.

Europe: The Stress-Test

There is one case study, right now, where the model is being stress-tested in plain sight. Europe is in an unusual position in 2026. It has the cleanest, cheapest electricity in a generation. Negative wholesale prices on sunny windy afternoons. Industrial costs that should be the envy of the world. By the logic of the energy curve above, this should translate into a wave of European industrial competitiveness. It is not happening. Or it is happening too slowly. The continent is failing to convert cheap clean power into industrial advantage at the pace the situation deserves.

Read in the language of the equation, the diagnosis is sharper. The engines are present. The talent is here, European universities produce more engineering and science graduates than the United States. The capital exists, European savings rates are high, pension pools are large. But the deployment apparatus is missing pieces. Capital is risk-averse. Regulatory regimes are complex. Procurement is fragmented across twenty-seven jurisdictions. The cultural posture toward big bets is cautious where the moment rewards conviction.

This is not a critique. It is a diagnosis. Europe has the inputs. The question is whether the culture, the trust and the productive friction, is going to evolve fast enough to use them. Europe has every reason to be the rule-maker of the next financial system, and is currently in danger of becoming the rule-taker inside someone else’s. There is no third option.

Between US capital depth, European trust architecture and Asian innovation velocity, the equation is not a problem to manage but a future to author.

Rule-maker of the next system, or rule-taker inside someone else’s. There is no third option.

The Future Is Built

All of it lands in one place. The future is not something that happens to leaders. It is something authored by them. The forecast is not the fate. The strongest, the fastest and the most certain do not win the next decade. The ones who see clearly, give meaning, and move with conviction into the unknown do.

That is anticipatory leadership, not a soft skill but the strategic infrastructure of the firm, and the posture this moment rewards.

That is Possibilism.

The future is built, not predicted.

If any of this is useful, the longer arguments live in The Quantum Economy and The Singularity Paradox. The work continues at tomorrowmensch.