The Game We Must Win

The pieces keep falling. Climate change, wealth inequality, technological disruption, social isolation - each one drops from above, demanding attention, requiring perfect placement. Like a game of Tetris gone wrong, nothing quite fits together, and the board keeps filling up. Our margin for error shrinks with every misplaced piece.

Now artificial intelligence has entered the game. It doubles the speed. But it also gives us something we have never had before: the ability to clear lines we could never reach.

The Wrong Score

We built our economy to reward the wrong things. We tax work and barely tax pollution. We measure national success by GDP - a number that rises whether we build a hospital or clean up after a disaster. Simon Kuznets, who created the metric, warned Congress in 1934 not to confuse it with wellbeing. We have been ignoring him for ninety years.

The misalignment runs deep. Consider food. We subsidise industrial farming practices that produce cheap, ultra-processed food engineered to be addictive. This food makes millions of people chronically ill. Then we spend billions on pharmaceuticals - Wegovy, Ozempic - to treat the obesity and metabolic disease that our food system created. We are paying to cause the problem and paying again to manage the symptoms, and somehow we call both sides of this equation economic growth.

Nobody designed this on purpose. Nobody wants it. And yet we collectively sustain it, because no single actor - no farmer, no supermarket, no health service, no government department - has the incentive or the authority to fix the whole chain. This is what Jerry Harvey called the Abilene paradox: a group marching somewhere that none of its members actually want to go, because each one assumes the others must want to be there.

Our economy is full of Abilene paradoxes. We know our incentives are broken. We know we are scoring points for creating gaps rather than clearing lines. But the system is too complex, too interdependent, too expensive to redesign. Or at least it was.

What AI Actually Changes

Most people talk about AI in predictable terms: it will make existing work faster, cheaper, more efficient. This is true, but it misses the point. If all we do is optimise what already exists, we will run straight into Jevons paradox - efficiency gains consumed by increased demand, the broken game played faster, the board filling up more quickly.

The interesting change is different, and it comes in two forms.

First, AI drops the cost of solving problems by orders of magnitude. Some problems always had willing buyers but impossible economics. Personalised education, for example. Parents and employers would pay for it, but delivering genuinely individualised learning at scale was prohibitively expensive. AI changes the unit economics, and a real business model appears. These opportunities are relatively straightforward.

Second - and this is where it gets important - AI transforms our ability to coordinate. The hardest problems we face are not hard because we lack technology or money. They are hard because they require orchestrating thousands of moving parts across organisations, sectors, and timescales that do not naturally align. Infrastructure projects run late and over budget not because we cannot pour concrete, but because procurement, planning, regulation, design, and construction are agonisingly difficult to coordinate. AI does not yet reduce the cost of laying bricks, but it dramatically reduces the cost of orchestrating the build - streamlining procurement, modelling resilience, optimising design, and compressing timescales that used to stretch for decades.

This matters because our biggest collective action traps - the Abilene paradoxes of food systems, energy policy, healthcare, housing - persist largely because coordination is so expensive that reform feels impossible. When the cost of coordination falls, problems that everyone wanted to fix but nobody could fix alone suddenly become tractable.

Elinor Ostrom won a Nobel Prize for proving that communities can manage shared resources without top-down control or privatisation. They need clear rules, local adaptation, and ways to hold each other accountable. Her work was brilliant but hard to implement at scale, because the monitoring and coordination it required was too costly. AI makes it scalable. Not by replacing human governance, but by making the information flows and feedback loops cheap enough to actually work.

Clear the Board

This is not about playing the current game better. It is about clearing the board to play a different game entirely.

Business strategists call this Horizon Three thinking: instead of optimising your current operations or extending into adjacent markets, you ask what becomes possible now that was not possible before. You start there.

Business as usual will be disrupted. Every process that can be automated will be. Every middleman whose value depends on information asymmetry will be squeezed out. If your strategy is to keep doing what you did last year with AI bolted on, you are standing where the ball was.

The builders who matter now are the ones turning their attention to problems that used to be too expensive to coordinate. Founders building tools for communities, not just consumers. Companies that measure success by problems solved, not just revenue captured. Civic technologists creating infrastructure for collective decision-making. Reformers who can finally grasp the nettle on systems - food, energy, housing, health - where everyone knows the incentives are wrong but nobody could afford to redesign them.

The cost of coordination just changed. The problems we have been marching past, pretending someone else would fix, are now within reach.

Your Move

The pieces will keep falling, faster than before. You cannot slow the game down. But you can stop scoring points for creating gaps and start clearing lines that matter.

The board is filling up. For the first time, we have the tools to clear it.