As said, Open-Endedness is going mainstream. Great article by Nathan Benaich and Alex Chalmers on Air Street Press!

Really excited to share that I'm hiring for a Research Scientist position in our team! If you're interested in the kind of cognitively-oriented work we've been doing on learning & generalization, data properties, representations, LMs, or agents, please check it out!

Open-endedness is all you need! Great read from Alex Chalmers and Nathan Benaich at Air Street Press outlining the challenges in building truly useful AI agents that can deal with the conplexity and unpredictability of the real-world. They highlight open-ended learning as an

Great example of how AI will accelerate the scientific process! Congrats on the seed round Karl Moritz Hermann and Marc G. Bellemare! "You dump a thousand documents in, say you want this, that, and the other data out of them, and Reliant pores through them and finds that information —

Some impressive new quantum chemistry research from Google DeepMind in Science Magazine! We share a gold standard, deep learning approach for computing how molecules behave at the quantum level when stimulated e.g. by light. ⚛️💡deepmind.google/discover/blog/…

What do you really know about AI? Explore humanity's most transformative technology in this compelling collection of short essays from Professor Tim Rocktäschel. Out 26 September: geni.us/ArtificialInte…

🐝🌻Summer Eggcelent🌻🐝 Mind you, I pointed out this was the early version of these concepts at the start of the year. x.com/_rockt/status/…

5/ Google DeepMind's Genie model introduced a latent action model that can infer hidden actions in between video frames. for ex - learning to "jump" over obstacles by training on platformer videos. this helps predict the next frame in an interactive video

People were dunking on this sort of stuff for being too jank *this year* and the fact we're at this point already is insane x.com/_rockt/status/…

Nice work disentangling learnability from room for improvement (aka regret). The two do not necessarily coincide! This has always been a crucial sticking point for regret-based environment design. When you have a clear metric for success rate p, the "learnability" metric

Life update: I’ve joined the UK AI Safety Institute on the Safeguards Analysis team! We’re working on how to conceptually and empirically assess whether a set of safeguards used to reduce risk is sufficient.

We’re presenting AlphaProteo: an AI system for designing novel proteins that bind more successfully to target molecules. 🧬 It could help scientists better understand how biological systems function, save time in research, advance drug design and more. 🧵 dpmd.ai/3XuMqbX

The ICML talk on Genie is now available. Great presentation by Jack Parker-Holder and Ashley Edwards! Congrats to the whole team again on the Best Paper Award! Well-deserved. Video: icml.cc/virtual/2024/o…

Excited to announce AlphaProteo, our breakthrough AI system for designing novel proteins that bind to a target molecule or protein. It has the potential to accelerate our understanding of biology, and aid the discovery of new drugs, the development of biosensors, & much more!

The popular science series 10 Things You Should Know by SevenDials Orion Publishing has a new member: Artificial Intelligence! Out September 26th. You can pre-order it at geni.us/ArtificialInte…

🏆 Best Paper Award: UCL researchers show weaker AI models can assess stronger ones through debate, achieving 76% accuracy. This breakthrough paves the way for future AI alignment without relying on human expertise. Congrats to CDT student co-authors akbir. & Laura Ruis

Do you want a copy of my next book for free? The MIT Press @mitpress.bsky.social has arranged a giveaway via Goodreads, and you can sign up for a chance to win a free copy.

Super proud to have been able to work with you Patrick Lewis! Does this improve my Bacon number?

Are you working on applying LLMs to games in some way? Well, we have the place you will want to submit your work to, namely our special issues on LLMs and Games in IEEE Transactions on Games. Deadline: December 1.