AI Circle Daily
https://aicircle.news/archive/2026-09-24?category=leaders&lang=en
Scan to read this page
Nvidia CEO Jensen Huang says AI labs that can't contain their experiments should be shut down, putting OpenAI and Anthropic on notice.
Take: The shovel seller is dictating terms to the miners — offloading safety liability onto model makers while he keeps collecting compute rent. OpenAI and Anthropic get torched.
Huang says AI is not an existential threat no matter how many ex-lab employees sound alarms — but labs that can't contain models should still be shut.
Take: Deny the apocalypse, keep a backdoor warning — classic valuation-protection spin. Who fears regulation most? Check the stock chart.
Anthropic CEO Dario Amodei tells the UN Security Council that poorly managed AI could become a serious risk to humanity.
Take: Rival Altman co-signs on the same stage — two sworn competitors suddenly agree on guardrails, which is really about writing the rules first and raising the bar for newcomers.
OpenAI's Sam Altman and Anthropic's Dario Amodei urge the UN Security Council to adopt international AI standards.
Take: Mortal rivals, instant united front the moment rules get written. Who's locked out? Everyone not yet at the table.
The CEOs of rival labs Anthropic and OpenAI agreed Wednesday that global standards are needed to address AI risks.
Take: Global guardrails sound noble; in practice they're a compliance moat. The louder the giants chant standards, the harder it gets for small players.
After Trump rebuffed a 'globalist scheme' to control AI, Altman and other execs pushed cooperation at the UN, arguing to temper AI's pace to manage risks.
Take: With the White House refusing the global-governance ball, CEOs route around it via the UN — a direct tug-of-war with Trump over who frames AI.
Elon Musk said 'Dario is right' about slowing AI development — even as Tesla and xAI keep ambitious AI goals.
Take: Agreeing with the brakes while flooring the accelerator — PR insurance for the xAI and Tesla roadmaps. Don't take it literally.
Responding to computer scientist Francois Chollet, Musk predicted AI may outperform humans in every field by 2028.
Take: The more aggressive the timeline, the better the xAI fundraising story. Hotter safety debate, more eyeballs on his prophecy — classic attention arbitrage.
On a new episode of The Ezra Klein Show, Nvidia CEO Jensen Huang took aim at catastrophic AI forecasts, saying 'Enough predictions.'
Take: Nvidia's CEO personally suppressing doomsday narratives because panic threatens compute demand — the most expensive PR money can buy.
Jensen Huang says the junior developer problem ends in two years, implying AI will absorb entry-level coding work.
Take: One prophecy rewrites hiring budgets overnight. New grads and CS programs feel the chill first, while firms quietly cut training costs.