I just completed the Horizon Institute's AI policy workshop in DC, a three-day event with the explicit goal of placing safety and technically-minded people in policy roles. As I understand it, they do very little vetting of the particular beliefs about AI that participants hold, trusting instead that putting smart people in good positions will bring about good outcomes.
The mission is awesome, and they are great at what they do. I highly recommend applying for their workshops and fellowships if you are AI policy-curious.
Here are a few of my general takeaways from the event. I omit specific names due to the closed door nature of the event, which benefits from the candor of its guests and participants alike.
I found it heartening that so many people from so many different fields—lawyers, CS grad students, historians, biologists, a feminist literature professor, a screen writer—came to the workshop to kickoff their own path to making AI go well. Going in, I expected something like an "AI as a Normal Technology" risk conference, and instead I got a sober conference that (in the main) understood what's at stake. It has been a positive update for me.
I was also at this event and can confirm everything Ryan said! I had a very similar experience and was a little astonished by how with it the attendees were. The event was under Chatham house rules but I will note that the 'we've all read LessWrong" speaker is very much a DC native not some random EA person who just happened to be there.
I will also recommend in their workshops or applying for their fellowship (closes Aug 30th, begins Jan) if you have any interest in policy, it can be as short as a six month DC 'tour of duty' and I think there's a lot of b... (read more)