There are not many figures in the AI industry who can credibly tell its most powerful executives to be quiet. Yann LeCun is one of them. The Turing Award-winning scientist (who helped build the very foundations of modern deep learning) offered something increasingly rare in the current AI moment: a calm, dissenting voice grounded in four decades of actual research.
His message was blunt. “Don’t listen to CEOs,” he told Axios. “They have a vested interest in propping up the power of the products they sell.”
LeCun argued that AI tools, whilst powerful, are “still not very good at reasoning,” and warned that “there is a long history of researchers in AI having a widely optimistic view of when machines will become more intelligent than humans.” His conclusion: human-level AI is not coming any time soon.
The interview comes at a pivotal moment. For those trying to make sense of conflicting signals, from trillion-dollar infrastructure commitments to apocalyptic workforce forecasts, LeCun’s perspective carries particular weight. He is not a sceptic of AI’s potential. He is, arguably, one of its architects. That is precisely what makes his pushback so significant.
The 20% forecast
The most direct clash has been with Anthropic CEO Dario Amodei. Amodei has suggested that sectors such as technology, law, consulting and finance could face major disruption at the entry level, estimating that as many as half of such roles could be affected within one to five years.
LeCun is having none of it. He described the 20 per cent job-loss prediction as “ridiculously stupid,” arguing that current AI systems are nowhere near capable of replacing the bulk of white-collar work.
LeCun blasted Amoei on social media earlier last month: “Dario is wrong. He knows absolutely nothing about the effects of technological revolutions on the labour market. Don’t listen to him, Sam, Yoshua, Geoff, or me on this topic.”
Instead, he directed people towards economists who have spent careers studying exactly this question.
He is not alone in that scepticism. Nvidia CEO Jensen Huang has previously said he “pretty much disagree[s] with almost everything” Amodei has said on the subject. Yet the drumbeat from frontier lab leaders has continued regardless. Microsoft AI chief Mustafa Suleyman has suggested the white-collar disruption Amodei describes could be realised within just 18 months.
Meanwhile, current data does not obviously support the catastrophist case. US Bureau of Labour Statistics figures from March 2026 showed unemployment at 4.3 per cent, whilst Anthropic’s own labour tracker found no broad unemployment split for AI-exposed occupations.
The clearest signal so far is a roughly 14 per cent drop in job-finding rates for workers aged 22 to 25 in AI-exposed roles, a real and meaningful shift, but a long way short of the civilisational disruption being described in some quarters.
LeCun’s concern with the current generation of large language models is well documented, and he has put significant capital behind an alternative vision. Having left Meta in late 2025, he launched AMI Labs. Headquartered in Paris and explicitly positioned as a credible frontier AI company outside the US-China binary, backed by a $1.03 billion seed round. The company is built around world models rather than LLMs, on the conviction that language models alone will never reach genuine human-level reasoning.
The infrastructure buildout is real. The productivity gains from AI tooling are real. What LeCun is challenging is the conflation of those genuine advances with labour-market forecasts that happen to serve a particular commercial narrative. In an industry where hype and substance have rarely been harder to separate, that distinction may be the most important one to hold on to.
RELATED STORIES
ITW 2026
Over 2000 organisations from 120 countries made their mark at ITW 2025, powering the future of global connectivity and digital infrastructure.
