Might sound small, but hoping this will be big improvement for building LLM apps.

if you could improve one thing about chatgpt, what would it be?

Go Kraftful! Best way to summarize product feedback.

My city 🥰

Can female founders operate in founder mode? @BChesky said founders’ intuition comes from being the parent of the company. As VC Backed Moms, we know it well. As founders. And parents. Today we’re proud to announce our non-profit status, supporting nearly 500 badass founders🧵

Klarna CEO saying they are firing their SaaS providers even the "systems of record" that we thought were impossible to rip out. gone this is.....wild.

Large language models, chain of thought, reinforcement learning. It's all finally coming together. Our new model OpenAI o1 is here, and it's really good at reasoning.

OpenAI o1 is trained with RL to “think” before responding via a private chain of thought. The longer it thinks, the better it does on reasoning tasks. This opens up a new dimension for scaling. We’re no longer bottlenecked by pretraining. We can now scale inference compute too.

It's fascinating how well o1 generalizes across different subjects.

First step was to scale the model size. Next step is to scale test time compute. LLM -> TTC.

o1 model doing better than PhD-level humans

Training of o1 has been conducted by an amazing team. By far the best team I've seen in my life. If you are an outstanding researcher or engineer who wants to study frontiers of reinforcement learning for reasoning, we still have a few roles we want to hire for o2 😉

swyx.o1 OpenAI tylercowen Catherine Brownstein Mario Krenn I asked GPT o1 to do the homework for the hardest class at Harvard, Math 55, and it got a 90%:

What will be possible once inference flops / watt drop by another few OOMs?

That feeling when ChatGPT o1 accomplishes in 1 hour what took you about a year in your PhD: youtu.be/M9YOO7N5jF8?fe…

The E.U. Went Too Far stratechery.com/2024/the-e-u-g…

Advanced voice model is finally rolling out this week. Post-training this model was an amazing project to work on as I transform from a CV researcher into a truly multimodal one. Join the team if you are excited to figure out the best interfaces for intelligence.

Voice mode getting better