Thanks for pulling out these quotes, I found this useful!
Interesting, the hearing was much more directly focused on core AI Safety issues than I expected. I find myself continuing to get more optimistic about some dynamics like whether AI Safety discussions are happening in serious venues, but continue to be uncertain about some dynamics like open source / decentralized systems and fast takeoff scenarios.
FWIW, the parts I quoted were much more focused on that than the parts I didn’t quote – it’s not like the hearing as a whole was this focused on core issues.
Thanks for pulling out these quotes, I found this useful!
Interesting, the hearing was much more directly focused on core AI Safety issues than I expected. I find myself continuing to get more optimistic about some dynamics like whether AI Safety discussions are happening in serious venues, but continue to be uncertain about some dynamics like open source / decentralized systems and fast takeoff scenarios.
FWIW, the parts I quoted were much more focused on that than the parts I didn’t quote – it’s not like the hearing as a whole was this focused on core issues.
Fair point! Even seeing these things mentioned at this level from the execs and some senators seemed like a positive step to me.