@AISafetyMemes consider yesterday's other tweets tho... imo they got a significant positive result on a set of evals they'd identified as "agi-candidate", had a meeting about it, and then each went their own way to mull it over & tweet about it
@AISafetyMemes and/or they're all tweeting under the iron tentacles of shoggoth AGI 😏
More notes on nearby ideas
Reply · Sep 26, 2023@max_paperclips @gfodor lol my scenario- good eval results vs whatever internal metric they've come up w for agi, have been mulling over what to say…Reply · Sep 26, 2023@AISafetyMemes what's your expectation of how it would play out when it happens? this matches my expectation- they hired ppl for evals for AGI a few…Reply · Jan 26, 2023@ohnerako @ChadRizzTard aww, well they are senior figures in AI safety, whatever their twitter antics...Quote tweet · Sep 26, 2023@gfodor what do you make of this, and the rumors that OAI has developed a secret set of evals for "AGI"?…