Sam Altman Warns AI Could Concentrate Power, Lose Human Control as 10% Extinction Risk Debate Grows
Updated
Updated · Yahoo Finance · Sep 15
Sam Altman Warns AI Could Concentrate Power, Lose Human Control as 10% Extinction Risk Debate Grows
3 articles · Updated · Yahoo Finance · Sep 15
Summary
Sam Altman said AI progress could fail in two ways: humans could lose control of the technology, or extraordinary power could end up concentrated in one company, person or country.
In an X post Monday, the OpenAI CEO argued safety and alignment work must stay ahead of model capability gains, framing that balance as a “narrow middle path.”
Dario Amodei sharpened that debate over the weekend with a 3,800-word essay urging AI labs to slow capability advances so risk prevention has time to catch up; Altman said he agreed.
Jensen Huang pushed back at a conference when asked about a 10% human-extinction risk from AI, saying past predictions about technology have often been wrong.
Deutsche Bank said public anxiety has shifted this year from AI disrupting software companies to fears over jobs and even extinction, underscoring how central existential-risk rhetoric has become to the AI boom.
Are top AI executives warning of existential doom to save humanity, or just building regulatory walls to crush future competitors?
As AI systems begin self-improving beyond human understanding, can any global governance actually stop a misaligned intelligence from taking control?
If mandatory safety vetting slows down frontier models, will the trillion-dollar AI investment boom suddenly collapse before delivering real-world results?