12
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 14 Jun 2026
12 points (100.0% liked)
TechTakes
2612 readers
12 users here now
Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
founded 3 years ago
MODERATORS
AGI’s ETA: Delayed (again) due to technical difficulties
came across this one when doomscrolling, pretty good article with healthy doses of skepticism. out of three commenters, one appears to be a rationalist (based off the “empirical evidence” and “instrumental convergence” they mention) who tries to argue that we don’t actually need human-like AI for paperclip maxxing, even though every single AI doom argument is about smart-as-a-human and better AIs. Just comes off as goalpost moving
The obsession with utility-maximizing Bayesian agents only gets weirder as time goes on. You couldn't really get further from the paperclip maximizer concept than an LLM, and yet they still won't shut up about unaligned goals and hidden utility functions.
The header image is gold.
This bit is true though; capitalism is already doing a fine job of paperclip maxxing.