surrealist theory of international relations
Milan Weibel 🔷
@weibac.bsky.social
375 Followers
863 Following
computer and political science student | intern @ LatAmGPT project | 🏳️🌈 clarity is a public good weibac.github.io
-
-
This site would probably benefit a good deal from community notes. That's maybe the only great twitter feature
-
LLMs are great at picking up and echoing vibes, positive or negative
-
the LLMs are mating now
What if we could evolve AI models like organisms, letting them compete, mate, and combine their strengths to produce ever-fitter offspring?
Excited to share our new paper, “Competition and Attraction Improve Model Fusion” presented at GECCO 2025 (runner-up for best paper)!
-
there's considerable overlap between AI experts and the supposed "AGI cult"
ilya sutskever was a coauthor of alexnet and chief scientist at openai, geoffrey hinton has warned about existential risk from AI, stuart russell wrote both the canonical textbook on AI and a book on the alignment problem
-
also: there's considerable overlap between AI experts and the "AGI cult"
ilya sutskever was a coauthor of alexnet and chief scientist at openai, geoffrey hinton has warned about existential risk from AI, stuart russell wrote both the canonical textbook on AI and a book on the alignment problem
-
side point but running LLMs and selling subscriptions / API credits for people to use them is actually profitable, what makes AI companies spend more than earn is their R&D investment into training bigger models
-
genuine question is this request borne out of a desire for people to write "correctly" or out of a dislike for the dropping of capitalization/punctuation as a subcultural marker?
-
yea this guy rolled a 0 on persuasion
-
i think the difference here is that he's looking at voting like a consequentialist action intended to bring about a less bad outcome (thus saving the romani) rather than an expression of moral support
him not considering third options though is quite dumb, politically and rethorically
