Every technological revolution seems to arrive with a promise. Life will become easier, work will become less demanding, goods will become cheaper and prosperity will eventually reach everyone. I guess we have been hearing some version of this story for centuries. Only the technology changes.
Now it is artificial intelligence, and the promises are becoming rather extraordinary. Sam Altman has spoken about a future of abundance, where intelligence becomes widely available and the cost of producing goods and services falls dramatically. Elon Musk has gone even further, imagining a world where work becomes optional and people enjoy a universal high income.
I find it rather ironic that some of the biggest beneficiaries of capitalism are now describing a future that sounds, in certain respects, remarkably close to socialist ideals. A world where people may no longer need to work to survive, where essential services become affordable and where material comfort is available to everyone. It sounds wonderful. But haven’t we heard something like this before?
The Industrial Revolution transformed production. Machines could manufacture goods on a scale that human labor alone could never achieve. Factories expanded, industries grew and wealth increased. Yet for many workers, progress initially meant long hours, dangerous conditions and little control over their working lives. The machinery was becoming more productive, but the people operating it were not automatically becoming more prosperous.
Living standards eventually improved, but not simply because better machines had been invented. Labor movements, legislation, public education and political reforms played their parts. Many of the benefits we associate with industrial progress had to be fought for and negotiated.
In 1930, John Maynard Keynes imagined that technological progress might eventually reduce the working week to around fifteen hours. He believed increasing productivity could free people from much of the economic necessity that had governed human life for centuries.
Almost a hundred years later, we have machines that manufacture, software that calculates, systems that coordinate businesses across continents and now AI that can write, analyze, design and assist with increasingly complex decisions. Yet most people are nowhere near a fifteen-hour working week.
Of course, we are materially better off in countless ways. Technology has made things possible that previous generations could scarcely imagine. But the freedom from work that Keynes anticipated hasn’t arrived for most of us.
If we have become so much more productive, why does the promise of working less keep moving further into the future?
Perhaps because productivity and prosperity are not quite the same thing. A company can produce more with fewer employees, reduce costs and increase profits. Whether those gains become higher wages, lower prices, shorter working hours or greater returns for shareholders depends on decisions that have little to do with the technology itself.
And this is where the present conversation about AI becomes particularly fascinating.
Altman has discussed how AI could create enormous wealth and proposed ways of distributing some of it more broadly. Musk has suggested that sufficiently advanced AI and robotics could eventually make conventional employment unnecessary. Both are describing possibilities in which people’s economic security becomes less dependent on paid work.
Isn’t that close to something socialist thinkers have argued about for generations? The possibility that advances in production could reduce economic necessity and give people greater freedom over their lives.
Of course, socialism is not simply about making everything cheaper. It also concerns ownership, control and the distribution of economic power. A society can provide affordable healthcare, education and public services while remaining capitalist. And a world where a handful of corporations own the systems producing almost everything would hardly represent collective ownership, however inexpensive their products became.
That is why I find the promises of AI abundance difficult to accept without asking a few more questions.
Who will own the technology producing this abundance? Who will decide how its benefits are distributed? And what happens to people whose livelihoods disappear before the promised prosperity arrives?
Consider a company that uses AI to do the work previously performed by a hundred employees. Its productivity increases, its costs fall and perhaps its profits rise. For the company, this is progress. But what does that progress mean for the employees who are no longer needed?
Perhaps some will find new opportunities. History certainly gives us examples of technology creating occupations that previously did not exist. But it also tells us that the transition can be painful, and the people who bear its costs are not necessarily the ones who enjoy its rewards.
There is another complication. Developing advanced AI requires enormous investment in computing infrastructure, energy, specialised expertise and capital. Although smaller companies and open models are expanding access, much of the most powerful infrastructure is concentrated among a relatively small number of corporations.
AI could give small businesses capabilities they could never previously afford. It could make education, healthcare and professional knowledge more accessible, particularly in countries where such services remain out of reach for millions. I genuinely hope it does.
But widespread access to a technology is not the same as widespread ownership of the wealth it creates.
And I sometimes think the promise of future abundance makes the present concentration of economic power easier to accept. If everyone is told that today’s disruption will eventually create extraordinary prosperity, perhaps the inequalities developing along the way begin to seem like a temporary inconvenience.
That doesn’t necessarily mean the people making these predictions are insincere. They may genuinely believe AI will improve life for billions. But believing technology can create abundance is different from explaining how that abundance will reach ordinary people.
History has given us plenty of reasons to believe in technological progress. It has also shown us that fairer wages, better working conditions, public services and economic protections do not simply appear because machines become more efficient. People, institutions and governments have to make choices about them.
And here we are again, listening to predictions that another technological breakthrough might finally free humanity from economic insecurity.
I don’t doubt that AI can create extraordinary wealth. What I question is the assumption that creating more wealth necessarily means sharing it more fairly.
If the people promising abundance are also the people who own the technology producing it, what exactly will make this revolution different from the ones before it?