zuck is a smart guy, he knows how much people can forget or deform past events

📚 Our book "Machine Learning in Quantum Sciences" — a guide to the AI revolution in quantum physics — is coming soon! 🚀 🤖 This unique collaboration was led by the incredible Prof. Anna Dawid Ania Dawid, with insights from top experts in ML & AI in physics. 🔬 Stay tuned!

golden nugget here

Thermo Matrix Exponentials has been published in Physical Review Research 🔥 On a thermodynamic computer, the matrix exponential occurs very naturally through the temporal covariance driven by the noise - a polynomial speedup over digital computers! journals.aps.org/prresearch/abs…

Check out our new paper on thermodynamic optimization! 🔥 We show how K-FAC, a powerful and scalable second-order optimizer, can be accelerated using thermodynamic hardware, bringing us closer to efficient GPU-thermo co-processing! This work pushes the boundaries of

This first one took me a good few minutes 😄

Oops 😄

How much GPU hours were spent in making an LLM defeat the Pokémon League

completely wild chart

Especially true for vibe coding with fully automated git commands

Sorry but this is a bad plot. Answering any question takes 15 seconds (with an error bar in that order of magnitude) ?

How is it possible that the font is uniform? They must be using some trick to paste the text at specific position no?

Filez moi des memes Français 🇫🇷 à faire en version Ghibli 🙏

The fact that ELIZA did better than 4o should be enough to rebutt most agi claims for all these systems

Doesn’t this just point to them simply having access to the problems?

Hey London! Join us during London Tech Week on Wednesday, June 11 for a hands-on workshop led by William Falcon ⚡️, creator of PyTorch Lightning and CEO of Lightning AI and Luca Antiga ⚡️, CTO of Lightning AI. Learn how to create lightning-fast, scalable AI agents that are production-ready from

When does Europe stop coping and funding companies with unsupported (not to say bs) claims? This is by far not the only case en.ara.cat/misc/openchip-…

Oof

LLMs have largely solved breadth (you can code in any language…) but so far have not been so good with depth (…only if you have a good foundation in at least one language, and only if you don’t do things that are too complex)

This is a really good analogy