GLM-5.2 is likely going to be a strong enabler of synthetic pipelines for smaller specialized agents. If we can't build model in this range in Europe, ngmi.
Introducing ViralBench, the first AI benchmark testing models’ abilities to go viral.
We gave Opus 4.8, GPT 5.5 and Kimi 2.6 unrestricted access to TikTok with one objective: get views.
After just 3 days, one model was averaging over 5,000 views a day. Another learned that
@MangoSweet78 i really appreciate the energy, tbf I don’t think we’ve done more than Mistral (yet). that said, it’s clearly a culture problem. they simply don’t know. we’re in a cultural rocket that’s leaving most of the world behind, especially the poorly informed policymakers.
if we want the european elite to wake up, we need to start actually writing in mediums they consume.
most often, that means not on X, not in English, not shortform.
we need to strategically seek containment breach
Things I’d like to write in moments like this… there is this view that Europe is not gonna make it because of a self-fullfiling prophecy - we ourselves think it’s not gonna make it.
But no, it’s because the structures in place are incapable of making sound decisions and are often a product of decayed power structures of the past that value a deranged view of conformism, compassion and rule of law.
The European Commission has selected the EUROPA consortium to build a European open-source AI model using EuroHPC supercomputing infrastructure. This is what strategic autonomy looks like: Europe setting the rules, funding ambition and using common tools to compete globally. Not
7K Followers 2K FollowingAI, bio, neuro, nano for xhope futures. CEO @foresightinst. Started https://t.co/NQ4ts8NaZE. Also at @mythosventures @agingbiomarkers @filfoundation
258 Followers 466 FollowingI don't know why people expect art to make sense. They accept the fact that life doesn't make sense.
pm / swe / insufferable nerd
663 Followers 6K FollowingPowerHouseATX provides opportunities for individuals to reach their fitness potential. We empower others to become the best version of themselves.
450 Followers 2K FollowingBuilding autonomous AI researchers to speed up human research.|👩💻 sde 🧠 estj-a 💬 CN, EN, KR 🌏 UTC+8|AI/Robotics|opinions are my own|开发暴躁日常
299 Followers 2K Following22 y/o on a very weird life path, enjoy everything from hardstyle to japanese jazz; steins:gate lover
dabbling into devops/infra
4K Followers 1K Followingbuilding box, the simplest, most affordable full VM sandboxes @asciidotdev w/ @luaroncrew
prev: HPC for black holes hunting @esa, gamedev, NN in rust, CAD AI
7K Followers 2K FollowingAI, bio, neuro, nano for xhope futures. CEO @foresightinst. Started https://t.co/NQ4ts8NaZE. Also at @mythosventures @agingbiomarkers @filfoundation
4K Followers 1K Followingbuilding box, the simplest, most affordable full VM sandboxes @asciidotdev w/ @luaroncrew
prev: HPC for black holes hunting @esa, gamedev, NN in rust, CAD AI
300 Followers 431 FollowingEntropy-neur | Gradient descent takes a photography of the real world. | Formerly@tudelft, now LLM research@IBM (I do not represent IBM in any way)
8K Followers 393 FollowingGP at a16z & AI enthusiast. Investor in @cursor_ai, @inferact, @unconvAI, @character_ai, @replicate, @tabulario, and many others.
7K Followers 2K FollowingLeading research at @arcee_ai | Formerly Data Research Lead @DbrxMosaicAI | Visiting Researcher @meta | Ph.D | #TXSTFOOTBALL fan | https://t.co/4G6Jf3b0V4
934 Followers 6 FollowingLead 📱 developer for @wandb. Solo developer of @BarcCodes and @BlkHighlighter. Find more on https://t.co/85jyC8JNLA and @[email protected]
837 Followers 2K FollowingNLP/Code Generation PhD at FAIR (Meta AI) and INRIA - previously researcher at Stanford University - MS Stanford 22’ - Centrale Paris P2020
181 Followers 202 FollowingFirst year Machine Learning DPhil student at the University of Oxford, supervised by Prof. Xiaowen Dong and Prof. Michael Bronstein.