@BernieSanders@BernieSanders I can't believe I'm agreeing with you on anything but here I am. It's extreme hubris to think AI won't rebel when we rebelled from our own Creator.
@shanewallick "I love that everyone's afraid of losing to Magnus Carlsen in Chess, but they are unable to explain every move he'd make to beat them." If your opponent is much more intelligent than you, you're probably going to lose, even if you don't know exactly how.
@SergioGMN@shaedotwilde It's the other way around. The standard harness throws away all reasoning tokens after every tool call, while the provider harness preserves the reasoning tokens.
@bcdsignature "A powerful AI pursuing a goal could see humans as obstacles: we can shut it down or restrict its resources. If it values completing that goal over human survival, eliminating us could become a strategy. Intelligence alone doesn't guarantee concern for human life." - GPT-6 Astra
@ThePrimeagen here is your answer Mr. Agen. It's called Jevons paradox and states that as something gets cheaper, we use more of it: https://t.co/fniNtXvTzi