Shane Legg, co-founder and Chief AGI Scientist at Google DeepMind, believes that the rapid rise of artificial intelligence (AI) could radically shrink,
or even eliminate, many work-from-home roles in the years ahead. This warning from Legg will be unsettling for millions of professionals who have comfortably switched into remote work. In a recent conversation with mathematician and broadcaster Professor Hannah Fry, Legg outlined how advances toward human-level AI are reshaping the very foundations of modern employment. His message was loud and clear that the jobs that exist entirely online are facing the greatest risk. Why Remote-First Jobs Are In The Firing Line Legg explained that roles relying solely on cognitive effort and computer-based work are especially exposed to automation. As AI systems become more capable, businesses may no longer need large, distributed teams to perform complex tasks. “Jobs that are purely cognitive and done remotely via a computer are particularly vulnerable,” Legg said in the podcast. Pointing out software development as a telling example, he said that with powerful AI tools assisting in coding, testing, and debugging, companies could dramatically reduce team sizes. What once required 100 engineers might soon be handled by just 20 highly skilled professionals working alongside advanced AI systems. The ripple effect could be severe for entry-level workers and remote positions that were once seen as gateways into tech careers, he added. Which Industries Will Feel The Impact First? According to Legg, the disruption will not hit all sectors equally. Fields that depend heavily on digital skills, such as language-based work, data analysis, mathematics, research, and complex problem-solving, are likely to face pressure first. AI has already surpassed human performance in areas like language processing and factual recall, and rapid progress is expected in reasoning, visual interpretation, and continuous learning. By contrast, hands-on professions that require physical presence and real-world interaction may be safer for longer. Trades like plumbing, construction, and other manual roles remain difficult to automate due to the unpredictability of physical environments and the limitations of robotics. A Broken Economic Model? Legg also warned that widespread AI adoption could undermine the traditional economic model where people trade mental labour for wages. As machines become cheaper and more efficient at cognitive work, many roles could simply disappear. He cautioned that ignoring these signals would be as shortsighted as dismissing early warnings about major global crises, urging policymakers and societies to start preparing now. A “Golden Age” — If We Get It Right Despite the sobering outlook, Legg is far from pessimistic about AI’s broader potential. He believes that the technology could unlock unprecedented productivity, scientific discovery, and prosperity, calling it a coming “golden age.” The real challenge, he stressed, will be ensuring that the wealth created by AI is shared fairly, so people can maintain stability, purpose, and dignity as the nature of work evolves. Legg stressed that these changes will unfold gradually, but once AI reaches professional-level performance across knowledge-based jobs, the pace of transformation could accelerate quickly.
