@ESYudkowsky @perrymetzger @alyssamvance is there a specific plan for research / experiments you'd do given more time and resources? or is your message that you believe AI research is net negative and should be put on hold indefinitely until someone comes up with new ideas to tackle alignment?
@ESYudkowsky @perrymetzger @alyssamvance hm, an answer of sorts perhaps? https://x.com/ESYudkowsky/status/1627123446672150530
More notes on nearby ideas
Reply · Oct 21, 2023@foomagemindset @MelindaBChu1 @bitcloud yeahhh, this is one of the reasons the big Pause is foolish- don't have a plan, can't even figure out how to…Reply · Jun 9, 2023@ask_ourself I don't think AGI to ASI is simple or fast like ppl say no, and also I think alignment research is best done alongside actual…Reply · Mar 4, 2023@MichaelTrazzi agree but I don't think this is "way behind" when you hear Safetyists asking for 10 to 50 yr pause for alignment research firstReply · Feb 25, 2025@ESYudkowsky tfw the most interesting alignment research results are the ones where researchers try to make the models *bad* from their default good…