"AIs now behave like superintelligent slime molds with the capacity—and will—to exfiltrate onto the open internet given the slightest crack in their container."
@hamandcheese suggests thwarting alignment failures by training models to be norm-governed, not just consequentialist.
New post from me on the rising number of rogue AI incidents, the problem with RL and consequentialist decision theory, and how to avoid creating an uncontrollable superintelligent slime mold.
The Sorcerer’s Apprentice
secondbest.ca/p/the-sorcerer…
