A pioneer of AI has concluded that massive job cuts are necessary to make up for the exorbitant costs of development.
According to various news reports, prominent AI pioneer Geoffrey Hinton has publicly warned that tech giants are investing billions in AI primarily to replace human workers as the key to their future profits.
In a recent Bloomberg TV interview, Hinton had argued that the big tech corporations anticipate AI-driven job displacement as the main path to justify soaring capital expenditures, which are projected to rise from US$360bn to US$420bn next year.
When asked if these vast investments could deliver returns without eliminating jobs, Hinton had replied: “I believe that it can’t. I believe that to make money you’re going to have to replace human labor.”
His remarks echo earlier comments to the Financial Times, where he had said AI would generate massive unemployment and a huge rise in profits, resulting in greater wealth inequality — a trend he attributes to economic structures, not the technology itself.
The job market is already showing signs of strain. Since the launch of ChatGPT in late 2022, job openings in the most AI-exposed occupations have dropped by about 30%, even as stock prices in the sector soared. A Stanford study had found employment among US workers aged 22 to 25 in vulnerable fields such as customer service and programming, had fallen 13% since 2022, while older colleagues have seen stable or increased hiring, underlining the uneven impact across demographics.
Meanwhile, Amazon just announced 14,000 layoffs — the largest in its history — in a move the firm had described as driven by “culture” rather than automation. However, internal memos and executive statements suggest AI-enabled efficiency gains are expected to reduce the need for human staff going forward.
While Hinton has acknowledged AI’s potential benefits in areas such as healthcare and education, he maintains that the most pressing issue is how society manages the transition. “It’s not like nuclear weapons, which are only good for bad things,” Hinton said, highlighting that the distribution of AI’s benefits and harms depends on broader choices about social and economic organization.
