Bill Gates Says the AI Era Will Be Turbulent — and That We're Not Ready

Bill Gates has published a nearly 6,000-word essay on his personal website, gatesnotes.com, titled "The turbulent AI era is here. The choices we make now are critical," shifting from his earlier optimism about artificial intelligence to a deeply pessimistic assessment of its implications for the collective future. The essay, covered by The Verge, The New York Times, and the Financial Times on August 26, 2026, marks a stark departure from the hopeful tone of his 2023 essay "The risks of artificial intelligence," which framed AI risks as real but manageable. The Verge

Gates wrote that "the transition to this new AI era will be one of the most turbulent times in human history" and warned that the world is not remotely ready for what AI will bring. "We are not preparing for it," he stated. He framed the stakes in stark terms: "AI will either be the greatest equalizer ever invented, or the worst source of injustice." Firstpost reported that Gates argues AI is advancing through several risk thresholds, though the essay does not enumerate them precisely. Firstpost

A central claim is that comparisons of AI to past technological shifts are misleading. Gates argued that analogies to the PC revolution, a comparison he himself drew frequently in earlier writing, understate the pace and scope of what is coming. He asserted that current performance issues in AI models — hallucinations (when an AI generates confident but false information) and reliability gaps among them — will be fixed quickly, removing the technical friction that today limits deployment in high-stakes domains like healthcare, finance, or criminal justice.

On labor, Gates was unflinching. He predicted that many jobs will disappear permanently, with both white- and blue-collar roles at risk of replacement, leading to widespread, permanent unemployment and significant social upheaval. He urged political leaders to act on AI's labor impact before unemployment rises sharply, communities are hurting, and public trust has eroded.

Gates proposed two specific policy mechanisms. First, taxing AI tokens and robots to discourage companies from replacing human staff and to bolster social funds. (An AI token here refers to the units of text or data that an AI model processes; taxing them would mean levying a cost on each unit of AI output.) Second, reserving specific categories of jobs for humans, a domain he called "Human Reserved." The essay did not specify which jobs would fall under that designation or how the token tax would be structured administratively.

On governance, Gates said existing institutions are inadequate for managing AI. He called for countries to establish dedicated bodies able to set priorities across government agencies. At the international level, he said the community must build a global organization to manage risks that cross borders, though he did not define those risks or offer concrete design details. He suggested the framework could draw inspiration from international nuclear, aviation, and ozone layer-protecting agreements. He acknowledged that such a framework would require US-China cooperation but did not outline a mechanism for achieving it, and urged countries to begin setting up the organization now.

Gates also said that "if someone had a credible plan for slowing down AI advances globally, I would likely support it," a position that places him closer to the AI safety and slowdown advocacy camp than to the accelerationist posture common among leading AI lab executives. The Verge characterized the essay as a bleak departure from the hopeful excitement about AI in his 2023 essay, though Gates insisted it is not too late "to ensure that the good outweighs the bad." The Verge

The evolution is traceable through his published writing. In July 2023, on gatesnotes.com, Gates argued that AI risks were real but manageable, warning that an AI-drafted essay could be as riddled with prejudices as factual errors, and advising readers to check the AI's biases as well as their own. In December 2024, reviewing the book "The Coming Wave," he noted that questions people ask him about AI often boil down to "What should I be worried about, and how worried [should I be]?" In January 2026, in a Facebook post, Gates said his wish for the world regarding AI in 2026 was "a feeling of Hope." The August 2026 essay retains a thread of that aspiration but is framed by alarm rather than confidence.

The gap between Gates' proposals and current policy reality is wide. Token-level taxation of AI usage has no precedent in any major economy. Job reservation frameworks for humans exist in limited forms in some countries for specific sectors, but none have been applied to AI displacement at scale. A global AI governance body modeled on the IAEA (the International Atomic Energy Agency) or ICAO (the International Civil Aviation Organization) would require treaty negotiations that, as Gates himself acknowledged, depend on US-China alignment that is not currently in evidence. The essay offers no mechanism for resolving that impasse.

Gates' framing of AI as categorically different from prior technology transitions is worth examining. The PC revolution, the commercial internet, and the smartphone shift each displaced specific job categories while creating others, sometimes on a compressed timeline. What Gates is now arguing is that AI compresses the displacement window further, broadens the scope of affected occupations beyond what prior transitions touched, and may not generate replacement work at sufficient scale or speed. That is a testable claim in the coming years, though the essay presents it as near-certain.

The broader context here is that the most consequential element may not be the pessimism itself but the specificity of the policy asks. Gates is one of the few figures in technology with both the institutional credibility and the philanthropic infrastructure to move a governance conversation. His call for a token tax and a Human Reserved job category gives policymakers concrete proposals to react to, even if neither is likely to be enacted in anything like its proposed form. The essay's value may lie less in immediate legislative traction than in shifting the Overton window — the range of ideas considered politically acceptable — of what counts as a respectable AI policy position. A figure who spent decades championing software-driven productivity now arguing that the pace of AI advancement may exceed society's adaptive capacity is a signal that the debate's center of gravity has moved.