@XenoParticle
Account based inFrance
About this account
- Account based in
- France
- Connected via
- France Android App
Account-level information from X, not a live location or the device used for a specific post.
Comprendre les Sciences Fondamentales et les aspects théoriques de la Monnaie.
Joined November 2022
- Tweets3.6K
- Following2.4K
- Followers613
- Likes12K
Pinned Tweet
Fast, synchronized Markdown editor. No membership. Perfect for saving your LLM conversations. coreality.cc/markdown/
D'après Yann, les LLM sont déjà obsolètes. D'après Dario, ils vont tous nous tuer. Vous préférez être exterminés par :
50%Un LLM
0%Un World Model
50%Un suicide collectif
2 votes • Final resultsNo aviation company sells airplanes that crash. The idea that AI labs will create a superintelligence so incompetent it accidentally destroys humanity is a contradiction. Either they know how to build advanced AI properly, or they don’t.
Au Coeur des Sciences retweeted
C'est la première fois qu'on a un "lanceur d'alerte" qui n'apporte strictement aucune information ni document à part une vague prophétie et dont le pdg de l'entreprise s'empresse de l'appuyer. 🤡
Au Coeur des Sciences retweeted
I think we're living another GPT2 moment. Too dangerous to be release. 🤭
Au Coeur des Sciences retweeted
safety cartel
/ˈseɪf.ti kɑrˈtɛl/
noun
A coalition of dominant firms that invokes safety to justify coordinated limits on the pace or terms of competition. This can include common restrictions, government enforcement, or exemptions from ordinary antitrust constraints, especially when those measures allow participants to slow down without sacrificing their relative competitive positions.
Especially: a group of frontier AI companies advocating industry-wide “pacing,” common capability limits, or regulatory checkpoints that constrain faster-moving rivals while preserving incumbent advantages.
See also: regulatory moat, incumbent protection, coordinated pacing.
“When competitors need an antitrust waiver to coordinate how fast their industry is allowed to advance, critics may call it a safety cartel.”
Cette panique générale repose sur un mensonge : il ne peut pas exister une super IA qui serait non alignée. Les capacités du llm proviennent de son entraînement à faire ce qu'on lui demande. Si cette étape échoue, le modèle n'est pas "dangereux", il est juste incapable et inutile
Capabilities aren't built in a vacuum and then 'aligned' later. Instruction Following IS the Capability and the 'Smart Unaligned AI' is a Fallacy: If a model fails to follow human intent, it’s not superintelligent, it’s just less capable.
Can we produce a formal Lean proof showing humanity is fully extinct, so the superintelligence stops bothering us?
Si tu es chercheur en IA et que tu crains pour ta vie face à la chasse aux sorcières qui se prépare, sache que je peux t'accueillir dans ma cave. Par contre, tu vas pleurer en voyant ma carte graphique. 😬
Why worry ? Next year, everybody will have a personal Terence Tao for 20$/month.
BREAKING: Terence Tao and 24 other Fields Medalists just signed a letter telling AI companies they're destroying mathematics
>headlines: AI solves famous math problems
>25 greatest mathematicians respond today
>"we are witnessing a general threat to intellectual work"
>ai labs are treating math problems like a benchmark you can brute force
>“solutions are announced in a rush, leaving no time for a proper writeup and citing relevant previous work of others"
>they're talking about openai
>"this raises severe attribution and plagiarism questions"
>they're definitely talking about openai
>openai offered to put math professor name on navier-stokes proof
>condition: abandon his coauthor bc he works at anthropic
>"the most precious resources of our profession are students and ideas"
>for the labs the most precious resource is GPUs
A Severe Misalignment of AI in Mathematics
Au Coeur des Sciences retweeted
i trained the fly to solve rubiks cube
is there anything this little guy can't do?
Bon je pense qu'on a la confirmation qu'il ne comprend absolument rien à ce qu'il raconte. Comment il a pu travailler dans ce domaine reste un mystère.
Jacob Coxon's next interview on CBS News (ex Anthropic+Open AI researcher who resigned)
"We can't just unplug it because it could be copying itself over to other computers. Like it's not that difficult to find yourself because an AI is just code. It could transfer itself over the internet to a different place and then you unplug it here, but it's actually still over there and maybe it makes 10,000 copies of itself and they're all cooperating."
----
From "CBS News" YouTube channel, (full video link in comment)
Après nous avoir fait miroiter un paradis d'abondance et un Einstein dans notre smartphone, on arrive au moment où une poignée d'élites s'approprie toute la puissance de l'IA — « pour notre bien », parce que c'est « trop dangereux ».
Et la foule va applaudir.
After teasing us with a paradise of abundance and an Einstein in our smartphones, we’ve reached the moment where a tiny elite appropriates the full power of AI — "for our own safety", because it’s "too dangerous".
And people will applaud.