He Coined the Term 'AI Safety' — Now He Says Your Job Has Five Years Left

Up to 99 per cent of jobs could be automated within the next five years, a leading artificial intelligence researcher has warned.

Dr Roman Yampolskiy, an associate professor of computer science at the University of Louisville, made the claim during an appearance on The Diary of a CEO podcast.

'We're looking at a world where we have levels of unemployment we've never seen before,' he said. 'Not talking about 10 per cent unemployment, which is scary, but 99 per cent.'

The figure has been met with alarm across the tech industry and beyond.

What Yampolskiy Actually Predicted

Yampolskiy's timeline rests on a specific trigger. He believes artificial general intelligence — systems capable of matching human cognitive ability across all domains — could arrive by 2027, according to Entrepreneur.

Once AGI exists, he argued, the economic logic of hiring humans collapses.

'It makes no sense to hire humans for most jobs if I can just get a $20 subscription or a free model to do what an employee does,' he said. 'First, anything on a computer will be automated. And next, I think humanoid robots are maybe five years behind.'

He did add a caveat. 'It doesn't mean it will be automated in practice. A lot of times technology exists, but it's not deployed,' Yampolskiy told KRON4. 'We may have a lot more time with jobs. But the capability to replace most humans in most occupations may come very quickly.'

That distinction matters. His warning is about capability, not a fixed deadline.

Automation Without Limits

According to Yampolskiy, this technological change is unlike any of the previous industrial revolutions. When machines replaced jobs with earlier machine developments, these advancements gave rise to new industries. For example, when steam engines replaced hard labour jobs, and computers replaced paperwork, these developments created new jobs.

This time, however, he feels as though there will be no more room for new jobs or industries to be created by replacing human labour with automation.

In his view, advanced AI systems will soon outperform humans in both physical and cognitive tasks in terms of speed, accuracy, and cost. These advanced systems will be able to complete tasks much faster and more cheaply than people can, making them the go-to source for all types of work.

His conclusion is candid and, if true, leads to many unsettling thoughts: there may soon be very few jobs that cannot be automated, with only a small fraction of roles remaining dependent on human empathy or craft.

A Future of Mass Unemployment

The effects of artificial general intelligence could become increasingly apparent in the coming decade if researchers' predictions hold true.

Unemployment could reach unprecedented levels, with vast portions of the workforce potentially displaced by advanced AI systems.

Imagine massive vacancies in businesses, routine tasks being automated, and AI increasingly performing functions once reserved for human workers, from writing to logistics.

It seems no industry will be spared by the emergence of AGI. Automated systems are currently producing first-draft articles, designing advertising campaigns, and creating musical compositions. As the abilities of those systems grow, the human element will eventually be seen as a luxury rather than a requirement.

This is not an abstract concept to many families; they are asking themselves, 'How am I going to pay my mortgage?'

The Few Roles That Might Endure

Still, Yampolskiy does not predict absolute extinction of work. A handful of niches may survive.

Human-made craft and authenticity: Some people will always pay for handmade goods. Artisans, tailors and specialist makers could serve a small premium market.

Counselling and lived experience: Therapists and carers may remain relevant because they understand what it feels like to be human. Empathy is harder to simulate than calculation.

AI oversight and regulation: Governments will need people to monitor and restrain powerful systems. Even if control proves temporary, supervision buys time.

AI interpreters and deployers: Specialists who explain complex systems to businesses and the public may act as translators between humans and machines.

Yet these roles would employ only a fraction of today's workforce.

Why This Prediction Carries Weight

Yampolskiy is not a fringe voice. He coined the term 'AI safety' in a 2010 paper and has published more than 100 research papers on the risks posed by advanced artificial intelligence. His 2025 book, Considerations on the AI Endgame: Ethics, Risks and Computational Frameworks, sets out his case in detail.

His claims also align with warnings from other senior figures. Anthropic chief executive Dario Amodei said at the World Economic Forum in Davos in January 2026 that AI could handle 'most, maybe all' of what software engineers do within six to 12 months. Geoffrey Hinton, widely known as the 'godfather of AI', stated in mid-2025 that artificial intelligence would replace 'everybody' in white-collar roles.

Yampolskiy's prediction is the starkest of the three. But the direction of travel is shared.

The Race Humanity May Not Win

Looking further ahead, Yampolskiy warns that a technological singularity may occur within the coming decades, when AI progress could accelerate beyond human comprehension. Research cycles could compress from years to days. Innovations would arrive too quickly for regulators or workers to adapt.

Even if the economic disruption produces abundance — cheaper goods, free labour, lower costs — Yampolskiy warned the social consequences could be severe.

'The hard problem is what do you do with all that free time? For a lot of people, their jobs are what gives them meaning in their life,' he said.

Governments, he added, have no programmes prepared for 99 per cent unemployment.

Even experts, he admits, already struggle to keep up. 'That may be the most sobering point. We are not just losing jobs. We may be losing control of the pace of change itself.'

For societies worldwide, the question is no longer whether AI will transform work. It is whether society can redesign itself before work disappears altogether.