@willdepue We don't believe in the scenarios Liron describes where a model training overnight suddenly becomes superintelligent, agentic, and malicious, takes over the data center and begins an attack on humans.
More notes on nearby ideas
Reply · Sep 19, 2021@BlckCatBlckSky All part of the Superintelligence's plan to convince humans to willingly turn over control to a more reasonable beingReply · Aug 29, 2023@samswoora @ToBeAtAll and slowly! at least compared to fears of superintelligence breakout overnight in a datacenterReply · Apr 21, 2023"I'm dreaming of a glorious transhuman future just as much as all the e/acc bros. But considering the dangers of a superintelligent AI wouldn't it…Reply · Oct 16, 2024@mriiirx oh like the question of what is intelligence and how will superhuman ai play out, my brain melts down trying to model it because i can't…Artificial intelligence