Everyone should look at the demos here. They are insane.

openai.com/index/hello-gp…

GPT-4o is our new state-of-the-art frontier model. We’ve been testing a version on the LMSys arena as im-also-a-good-gpt2-chatbot 🙂. Here’s how it’s been doing.

you guys are going to get like 8x more than you expect.

Deep learning is hitting A WALL:

A

Week of
Announcements,
Launches, +
Livestreams

Eval numbers for the new GPT-4 Turbo are out

GPT-4 reasoning has been further improved

No better day than an eclipse for the “science is awesome” shirt

1 token is all you need

Introducing MatX: we design hardware tailored for LLMs, to deliver an order of magnitude more computing power so AI labs can make their models an order of magnitude smarter.

Our hardware would make it possible to train GPT-4 and run ChatGPT, but on the budget of a small startup.

It's ok AI, I also can't sightread sexquinquagintillion

I’m 36.

thinking about thinking

its sad watching founders with perfectly good companies get trainitis, an affliction which compels people to train their own models from scratch. the cause of failure isn't big co doing the startup, but the startup doing the big co. trainitis can happen to anyone 🥺

Super excited about our new research direction for aligning smarter-than-human AI:

We finetune large models to generalize from weak supervision—using small models instead of humans as weak supervisors.

Check out our new paper:
openai.com/research/weak-…

Presenting two posters at #NeurIPS2023 , come by!

10:45am-12:45pm for both

- #527 Tuesday @ poster session 1: 'Training Chain-of-Thought via Latent-Variable Inference'

- #332 Thursday @ poster session 5: 'EvoPrompting: Language Models for Code-Level Neural Architecture Search'