We’ve all seen the headlines and the job figures. We’ve heard the debate about how many jobs AI will destroy or create, how soon, and how quickly. But to engage the issue mainly on those terms is to miss the point.

The central question of the AI economy is not simply who loses work. It is who loses power.

That shift is already underway. Even before we know exactly how many jobs will disappear, be transformed, or be created, AI is beginning to reshape the balance of power in the economy. It is changing who has leverage, who sets the terms, who captures the gains, and who is forced to adapt. You do not need to lose your job for your bargaining position to weaken. You only need to live in an economy where intelligence, productivity, and decision-making are becoming more scalable, more concentrated, and easier to deploy in ways that leave ordinary people with less room to negotiate.

That point matters because the public debate is often too narrow. We tend to talk about AI as though the only real question is whether a person remains employed. But work is about more than having a job title. It is also about the terms on which you work, the security you have, the choices available to you, and the degree to which you can negotiate rather than simply absorb whatever changes come from above. A person may remain employed and still lose ground. They may face weaker bargaining power, less autonomy, less time to adapt, and less ability to refuse bad terms. That is often how technological shifts are first experienced, not as instant mass displacement, but as a gradual erosion of leverage.

That is why I think the debate about basic income has to start in a different place. Too often, basic income is framed as relief for people after the damage is done. I think that is backwards. In the age of AI, basic income should be understood as leverage before exclusion. Its value lies not only in helping people after they fall, but in making sure they have bargaining power before the terms of the new economy harden around them.

This is an important distinction. Relief is reactive. Leverage is pre-emptive. Relief steps in once people have already been pushed aside. Leverage helps ensure they are not forced into a weaker position in the first place. If the economy is entering a period of rapid technological change, then we should not be thinking only about how to rescue those who lose out. We should be asking what institutions give ordinary people a stronger footing throughout the transition itself.

We have seen this play out before. The Industrial Revolution did not simply deliver broadly shared prosperity on its own. It dramatically increased productive power, but it also shifted the balance of power within society. Without countervailing institutions, the gains flowed upward while workers absorbed the disruption. Long hours, unsafe conditions, community dislocation, child labour, and deep economic insecurity were not historical side effects. They were what industrial progress looked like before society built the institutions needed to make that progress more widely shareable.

The answer was not to stop industrialization. It was to build new social infrastructure around it.

Workers organized into unions. Governments introduced labour standards, public education, pensions, and social insurance. These were not side issues. They were the institutions that allowed ordinary people to claim a share in the rise. They gave workers leverage. They helped ensure that technological and economic progress did not simply enrich those at the top while everyone else was told to adjust.

That is the real lesson for the AI age.

Too often, new technologies are discussed as though their benefits will naturally spread through society. History suggests otherwise. Economies do not distribute gains automatically. They distribute according to bargaining power, ownership, and the rules we build around them. When new technologies increase productivity, the key political and social question is always the same: who gets to claim the value that is created? Who has the institutional power to share in it, and who is left exposed to the upheaval?

If AI is a major economic transformation, then the question is not only what it will do to jobs. The question is what institutions we need so that ordinary people also share in the upside. If AI increases wealth, efficiency, and productivity, that does not mean those gains will be widely shared by default. It means we have to decide whether they will be.

This is where basic income fits in.

A basic income is not a complete answer to every challenge posed by AI, nor is it a substitute for unions or other worker protections. But it can be part of the new social infrastructure this moment requires. It can provide stability amid upheaval. It can give people the flexibility to adapt, retrain, negotiate, care for others, take risks, and refuse bad terms. It gives people a real claim on the prosperity of a richer society. In that sense, it is not simply anti-poverty policy. It is also a way of distributing power.

That matters whether or not someone loses their job. The point is broader than displacement. In a more automated economy, people need more than promises that growth will eventually benefit everyone. They need institutions and guarantees that ensure they are not forced to absorb all the risk while others capture most of the upside. They need some measure of economic independence so they are not negotiating from desperation in the middle of a historic shift.

That economic independence has practical consequences. It means more room to make decisions that are currently too costly for many people to make. More room to leave a bad job. More room to hold out for better work. More room to retrain or change direction. More room to care for children, family members, or aging parents. More room to take entrepreneurial risks or participate in community life. In each case, the underlying point is the same: a stronger floor gives people more real freedom, and more real freedom changes the balance of power.

That is one reason I think the usual critique of basic income often misses the point. It is often treated narrowly as a transfer program, or as a safety net only for those who have fallen out of the labour market. But in a period of major economic change, universal economic security does more than alleviate hardship. It changes the terms on which people participate in the economy. It affects how much pressure they can absorb, how much choice they really have, and how easily they can be compelled to accept whatever terms are placed in front of them.

I increasingly think of basic income in exactly those terms. If unions were part of the social infrastructure workers needed in the industrial age, then basic income can be part of the social infrastructure people need in the AI age. Not because it replaces older forms of worker power, but because every major technological transition requires new institutions to keep power from becoming too concentrated and progress from becoming too exclusionary.

That point matters especially because AI is likely to widen existing imbalances unless we act deliberately. The firms and institutions that develop, own, and deploy these systems are likely to benefit enormously from scale. The gains from automation and artificial intelligence are accumulating quickly, and they are accumulating unevenly. If the public response is limited to watching the numbers and waiting for the disruption to become obvious, then we will have already accepted a model in which most people bear the uncertainty while a smaller group captures most of the gains. By then, it will be harder to rebalance.

This is not an argument against innovation. It is an argument about what it takes to make innovation legitimate and durable. If AI is going to reshape the economy, ordinary people should not be left waiting to see whether they still matter once the dust settles. They should have a stake from the beginning.

That stake should not depend entirely on whether someone happens to own the right assets, work in the right sector, or hold the right credential. It should be built into the social architecture of the economy itself. The economy does not stand apart from society. It is a set of laws, institutions, and choices that we make. If we want to live in a society where democracy is more than a formal right, then we need an economy that ensures people share in the prosperity they help create.

That is why I do not think the basic income debate should begin with job-loss forecasts. It should begin with power. AI will change the distribution of leverage in our society. We have seen what happens when that kind of transformation arrives faster than the institutions needed to balance it. We should learn from that history.

The Industrial Revolution needed new social infrastructure so workers could share in the gains. The AI revolution will need the same. Basic income should be part of that answer, not as relief after exclusion, but as leverage against it.

‍