lol, casual 🤯

IMO, this thread is proof that any generated image/video that is interesting and not just plain slop requires significantly more thoughtful tooling and does require some understanding of aesthetics. The tension of reskilling vs replacement

Boston, Cambridge, Somerville, Worcester

This thread reminds me of the first episode of Black Mirror’s new season (iykyk)

I don’t even do CL research but all of my moots being in Vienna for ACL and sharing so much incredible work is giving me massive FOMO 🥹

It’s funny cuz it’s true!

THAT was a wonderful place to stop??? HBO should straight up delete the whole spinoff and we can pretend it was never cannon

Contribute your part of questions, it will be aggregated with other languages to create a massive multilingual benchmark, in your language

To reflect democratic principles, AI must be built in the open. If the U.S. wants to lead the #AI race, it must lead the #OpenSource AI race. #ArtificialIntelligence #GenerativeAI Hugging Face bit.ly/4mhRQAA

GPT OSS by OpenAI is out on Hugging Face with Apache 2.0 license. Enjoy! huggingface.co/openai/gpt-oss…

So much goes on behind the scenes for major open releases at HF, this is fantastic to hear ❤️❤️❤️

Please attend if you’ll be around!

Hold on this is crazy. TIL

The biggest lever towards ASI is right in front of you (RL environments) but you feel too smart to work on data

New paper idea: Graph Crimes in Evaluations as a form of Epistemic Injustice

The “generational wealth” reaction sort of encapsulates what benchmark chasing and not qualitative measurements of systems (measures like trustworthiness, reliability, happiness) has done to AI. Sometimes these things only become visible after years of longitudinal study which

Tapping the sign (LinkedIn post) again. “Robot death” is real insofar as people’s feelings and emotions are real, and ensuring both reduced emotional codependency and reliable availability is so crucial. Choose systems that you have control over!

GPT-OSS: - 5M downloads in <1 week on Hugging Face 🚀 - 400 new models - already outpacing DeepSeek R1’s launch numbers, and that’s without counting inference calls - also the most-liked release of any major LLM this summer