Hassabis’ AI Standards Idea Gets Support—What’s Next?

Way to go, Demis Hassabis!

The Google DeepMind co-founder and CEO won plaudits from much of the AI industry—including rivals OpenAI CEO Sam Altman, https://x.com/demishassabis/status/2077066418177151381?s=20">Microsoft CEO Satya Nadella, Block https://x.com/jack/status/2077043492065325458?s=20">co-founder Jack Dorsey and https://x.com/aplusk/status/2077033327098495144?s=20">investors Ashton Kutcher and https://x.com/chamath/status/2077050883070726409?s=20">Chamath Palihapitiya—for proposing the https://demishassabis.substack.com/p/a-framework-for-frontier-ai-and-the-dawning-of-a-new-age">creation of a new AI standards group funded by industry but working with government. 

Hassabis’ idea is that this new group could review frontier AI models to ensure they don’t pose a danger in areas such as cybersecurity and biological threats or in any other way (https://www.theinformation.com/briefings/google-deepminds-demis-hassabis-calls-new-u-s-led-ai-standards-body?rc=9byxri">more here).

Industry peers generally called the proposal “thoughtful,” although Nadella labeled it “important” and Palihapitiya described it as “well reasoned.” (Kutcher settled for “agreed.”) Unfortunately, those accolades won’t get us far.