DeepMind CEO Calls for AI Standards Body

DeepMind CEO Demis Hassabis wrote a lengthy and widely-shared X article (which I’m happy to summarize for you here!) laying out the case for an #Artificial Intelligence Standards Body. He compares this to the Financial Industry Regulatory Authority (FINRA), a public-private partnership between government and industry that oversees brokerage firms and exchange markets. (also comparable to the concept to the Motion Picture Association, or MPA.) Hassabis argues that AGI is “probably only a few short years away,” and he wants to ensure that frontier models get evaluated for national security dangers. As well, he suggests that the labs producing these models agree to some basic safety standards. The guidelines themselves seem largely reasonable, but many responders point out that, even if the US adopts such a standard, there’s no guarantee that the rest of the world (particularly #🇨🇳 ChinA.I. 🤖🧠🦾🤖 ) would play along. https://techcrunch.com/20..