What's the point of AGI?

It's been a busy few weeks, but I'm on a panel for the latest podcast of Tech Policy Press' "Sunday Show" alongside Borhane Blili-Hamelin and Margaret Mitchell to discuss a paper we co-authored – with a slew of additional authors – called "Stop Treating AGI as the North Star Goal of AI Research." It's an examination of the pressures and incentives that the AGI frame places on Artificial Intelligence research.

If you aren't sure what AGI is, join the club. Loosely framed as "Artificial Intelligence that can do what humans do, but better," the entire concept is weakly defined – and that's just one of the problems with using it as a strategic goal for tech development. With many pundits suggesting that "AGI is right around the corner," it's worth digging into what that means, exactly – because a close look beneath the surface tells us, it can mean anything anybody wants.

While that paper was a consensus view created with lots of input with various people in AI and social research fields. In the podcast, some of us get to channel our own interpretations from that research project, which should be taken as our own views and not that of the collective authors – which is a diverse group consisting of Borhane Blili-Hamelin, Christopher Graziul, Leif Hancox-Li, Hananel Hazan, El-Mahdi El-Mhamdi, Avijit Ghosh, Katherine Heller, Jacob Metcalf, Fabricio Murai, Andrew Smart, Todd Snider, Mariame Tighanimine, Talia Ringer, Margaret Mitchell, Shiri Dori-Hacohen, and myself.

Told you it was a lot.

Go check out the podcast on Apple, Spotify, Amazon, Pocket Casts or anywhere else.

It has been a hectic few weeks, and I apologize for silence from the newsletter! More to come soon, I hope.