Months earlier than transferring away from his function as CEO of Google DeepMind, Demis Hassabis held discussions with Trump administration officers and leaders of different AI labs about forming a brand new unbiased {industry} security physique. In response to a report by Wall Avenue Journal, folks acquainted to the matter revealed that Hassabis has been advancing the thought of a company that might codify security guardrails and greatest practices for growing synthetic normal intelligence (AGI). He has likened the proposed entity to the Worldwide Atomic Vitality Company, a nongovernmental watchdog for nuclear vitality.
Demis Hassabis’ assembly with US officers
The report additional provides that earlier this summer time, Hassabis met with Treasury Secretary Scott Bessent and Trump tech adviser Michael Kratsios to preview elements of his plan. He later revealed an essay on X calling for a U.S.-led, industry-funded requirements physique to evaluate AI fashions, much like the Monetary Trade Regulatory Authority. Below his proposal, federal companies and nationwide laboratories would work with AI firms to check fashions related to nationwide safety and designate “frontier-class” methods.The world’s largest AI labs are racing towards AGI, however leaders together with Hassabis, Sam Altman of OpenAI, and Dario Amodei of Anthropic have voiced considerations about dangers. Altman not too long ago proposed a U.S.-led worldwide discussion board, whereas Anthropic endorsed a worldwide settlement to gradual growth and confirm compliance. Hassabis’s efforts signify a few of the most concrete steps by a number one lab official to determine security mechanisms.
Management transition at Google
Final week, Google introduced a administration shake-up at DeepMind, with Hassabis stepping down as CEO and changing into chairman of DeepMind and chief scientist of Alphabet. In a memo, he wrote: “I’ve been working in the direction of AGI my complete life and now, like a lot of you, I really feel it’s shut at hand… I’ve determined that now’s the precise time handy over my day-to-day operational duties.” Alphabet CEO Sundar Pichai stated Hassabis’s new function permits him to focus totally on shaping the way forward for AGI.Critics argue voluntary and self-regulatory efforts are inadequate to deal with AI’s safety and financial dangers. Lawmakers are calling for necessary mannequin critiques and stronger authorities oversight. The Trump administration has elevated scrutiny of AI’s safety dangers however continues to maintain evaluations voluntary, leaving the talk over regulation unresolved as AGI growth accelerates.
