Avalon is the world's fastest 3D simulator for RL agents. All baselines train on 1 GPU in ~1 day.
We want academic researchers to be able to study aspects of intelligence missing from today’s models, even w/o access to large-scale compute.
Get started: https://t.co/Z73eF0SuvF
In 2020, @michael_nielsen & I began a 2-month project to write: "how would we fund science?"
2 years & 40,000 words later, it's become: "how can the culture & institutions of science actually change, and ultimately become self-improving?"
Our answer: https://t.co/QnxTbfemTA
Bluesky is building a social protocol. We released “ADX” (the X stood for Experiment) in May. Now that the design is starting to solidify, we’re renaming it to the “Authenticated Transport Protocol” — the “AT Protocol.”
https://t.co/Q0XUF8OhVD
It takes a lot of human ratings to align language models with human preferences. We found a way to learn from language feedback (instead of ratings), since language conveys more info about human preferences. Our algo learns w just 100 samples of feedback. Check out our new paper!
@keerthanpg@jxmnop Half of that in a year would be only 2-3% of Nvidia’s datacenter revenue. But I agree it may have a big effect for Nvidia! It speeds up the compute arms race by creating fomo for companies (like Meta) to spend even more than that to catch up… begun, the compute wars have
Excited to share a peek at what I've been working on recently: we're building networks of little sensors and using ML to help wind farms produce more energy!
It’s technically possible for our online identities and relationships to belong to us the way an email address, phone contact book, or password keychain does. So I think it will happen, one way or another. It’s just a matter of time.
Catch @DarrylBarnhart speak @NVIDIAGTC today about our exciting work @LatentSpaceAI on scaling LLM with retrieval! 📚
There’ll be an open QA in 30 mins: https://t.co/C42yjpjEi1
@iamtrask That’s true, we don’t have the counterfactual :) I think the best outcome would have been to push forward with a promising technology, but with a heavy focus on safety (as opposed to halting development). As far as I know, France did a good job on this front with nuclear energy.
...though Newton did spend ~20% of his time working on alchemy, and we've had some open problems in AI alchemy for a while 😂 https://t.co/SLnEQDwZcL (4/4)
Here’s another memetic isomorphism:
Crypto is more like physics, and AI is more like biology – not in “biologically inspired” sense, but in the nature of possible solutions.
So using the wordcel / rotator basis, Darwin is a wordcel and Newton is the ultimate rotator. (1/4)
16. It's fascinating how the two axes seem to imply each other:
Pure tinkering is fast and effective, but it is unprincipled, so lacks large-scale legitimacy, so can only cooperate through centralization
Theorycel thinking is slower, but well-suited for decentralization
Would Darwin/Newton gravitate more towards AI or crypto today?
Newton seems like he’d be more interested in snark hunting, while Darwin would work on scaling laws. (3/4)