Elon Musk is right – we need to plan for a world where AI has made work redundant

The newest technology requires an answer to the most ancient question there is, writes Andrew Griffin: how do we finding meaning in a world where work has become obsolete?

The world is stood at a crossroads between those two men, and it is difficult to know which way to turn (EPA)

We’re only five years away from the future. Almost a century ago, John Maynard Keynes said that by 2028 we would be living in a world of such abundance, powered by technology, that we would work for three hours a day – and mostly do that just to give ourselves something to do.

In his essay published in 1928, “Economic Possibilities for Our Grandchildren”, Keynes predicted that the rapid advance of technology would bring with it efficiency such that the working week would be transformed. Nobody would deny that the technology he predicted is here – but the working week hasn’t changed quite so much. That future, promised in five years, feels a very long way away indeed.

This week, Elon Musk appeared to take on the role of a modern-day Keynes when he told Rishi Sunak at the end of the UK’s AI summit that we weren’t just going to reduce the amount we’d work, but get rid of it entirely. When the prime minister asked Musk what he thought artificial intelligence would do to the labour market, he painted a picture that could be a utopia or a dystopia, depending on how you look at it.

AI is the "most disruptive force in history", he said; "we will have for the first time something that is smarter than the smartest human". While he noted that making predictions was hard, he said that "there will come a point where no job is needed". The only reason to work would be for "personal satisfaction", he claimed, since "the AI will be able to do everything".

"I don’t know if that makes people comfortable or uncomfortable," he said. "It’s both good and bad. One of the challenges in the future will be, how do we find meaning in life if you have a magic genie that can do everything you want?"

Sunak, for his part, seemed decidedly uncomfortable about that proposition. Working gives people meaning, he suggested, and he expressed his hope that the AI would improve the world of work rather than eliminate it altogether.

The world is stood at a crossroads between those two men, and it is difficult to know which way to turn. Part of the question is about technology: how much is inherent in humans, and will machines eventually be able to take over every part of our world? But a far deeper and more important question is not very technological at all: what are we here for, and what do we want to do with our lives?

It is a question we have got used to asking ourselves recently. The pandemic spurred all sorts of thought about the future of work and how people wanted to live, and some used it as an opportunity to change their lives in profound and lasting ways. But a new and deep version of that question is coming with artificial intelligence.

We might not have to answer it for a long time. The current pace of artificial intelligence, and the frenzy with which it is discussed, can make it easy to think that the robots are just waiting moments away, ready to take our jobs (and maybe our lives). But most experts suggest that is slightly overstated, and that at least many industries might be safer for longer. We should start thinking about it now, however, precisely because it is not here yet; we have an opportunity to shape how we embrace these technologies in a way we have not done before.

Much of the debate around AI tends to abstraction and science fiction: debates around it can often sound more like people suggesting stories for future Terminator films than policy discussions. It is important that we match that abstracted discussion with real concrete thinking about what we want the world – of work, of information, and much more besides – to actually look like.

But answering that might mean taking on even more philosophical discussions about purpose, meaning, and why we are actually here. They are questions that human intelligence has been grappling with for millennia – but artificial intelligence is about to give them a new and ever-more-urgent volume.

At the moment, those discussions are happening in a tone of panic and anxiety. Sunak is surely not alone in being pestered about the pessimistic view of automation, and how many jobs it is going to take. This is another reason that we should discuss what that automated future might look like; because there is an opportunity to make it a lot less terrifying.

This is surely the big unspoken part of the panic about machines and artificial intelligence systems taking our jobs. It’s not the robots we’re afraid of – it’s the humans. The assumption underpinning all of the panic about an AI dystopia is that none of the benefits of jobs being automated will go to the human workers who previously did them. This anxiety is everywhere, and Rishi Sunak pointed out during his discussion with Elon Musk that when he meets people out in the world they are not interested in grand questions of intelligence or the limits of computing, but in jobs.

The world would be a lot less of a worried place if people felt invested in this process – if they believed that they would benefit from automation. That can happen in a variety of ways. But they will all come to allowing people to share in the efficiency that is gained through automation.

The world has a mixed track record on that question. Technological change has always brought disruption to the labour market, but the effects have varied. Often, people who have been made redundant by machines in history have moved to new jobs, which usually bring less danger and less drudgery. To most people in history, today’s world of robots and computers would look like a utopia compared with the life-threatening and energy-sapping work they were used to. We shouldn’t lose sight of those gains just because they are now taken for granted.

But we haven’t always got the utopia those people of the past promised us, either. When in 1928 Keynes promised that world of working a few hours a day, it was less of a hope than a prediction. Even amid a financial crisis, he pointed to the vast array of technology that was being worked on – "electricity, petrol, steel, rubber, cotton, the chemical industries, automatic machinery and the methods of mass production" – in a way that is remarkably redolent of those promising the benefits of artificial intelligence today.

There is no one good answer for why we didn’t get that world of abundance and relaxation that he had promised. One important reason is that Keynes predicted that people would use their extra resources to spend more time at leisure. What has happened, however, is that we have spent it on more stuff: in large part, the economic advances of technology have just been spent on buying more technology, such as phones. But part of it is surely also because there has never really been much of a reckoning with this question: nobody asked the world what we should do with the efficiency gained through technology, and we ended up where we are now.

Even as he predicted that world of plenty and plenty of time, Keynes didn’t disagree with Rishi Sunak entirely. He said that it was impossible to "look forward to the age of leisure and of abundance without a dread", and that "we have been trained too long to strive and not to enjoy". People worry that they would have nothing to do – no special talents, no interests – without being rooted to the world by work. You only have to look at the rich to see that it can be difficult to make a life without work, he noted.

But people could do three hours a day of work to keep themselves contented, he noted; those last bits of work would mostly be about giving ourselves something to do. We’d been working primarily for purpose, rather than pay.

How do people find that purpose? What are people for? How do we find our Ikigai, as the Japanese refer to the sense of motivation that makes life worth living? We didn’t have a good response 100 years ago, when Keynes asked us; nor did we thousands of years ago, when Plato asked a similar version of it. But we’d better start trying to think of one now; the newest technology requires an answer to the most ancient question there is.