We’re honored to join the new AI Safety Institute Consortium led by National Institute of Standards and Technology and U.S. Commerce Dept.. AI agents that take actions in the real world will have significant implications, and policy plays a critical role in shaping these technologies as we build them. We look forward to

🎙️ New Generally Intelligent podcast episode! We interviewed Tri Dao (author of FlashAttention & Chief Scientist Together AI) and discussed: - Taking a contrarian bet on recurrent connections over attention - Using data augmentation to encode knowledge into models -

“We want to make it much easier for relatively technical people to make software. And then much easier for non-technical people to make software. And then eventually much easier for everyone to make software. And then we won’t think of what we’re doing as making software.” — Our

If you're sad about not getting into NVDA years ago, go work with this team--they're great, and we're very excited to use their chips as soon as they're ready

Finally, someone making it easy to interact programmatically with browsers! I've really been enjoying using Browserbase recently. Kudos to Paul for making such a nice product

Introducing the research preview of our foundation model, Character-1. Available today at hedra.com (on desktop and mobile). * Infinite duration (30s for open preview) * 90s generated per 60s (if our H100 supply holds) * Expressive talking, singing, rapping

Today we're releasing: - Cleaned up (and extended) versions of 11 public NLP benchmarks - An open source method for automatically discovering scaling laws - A guide to bringing up a 4000 GPU cluster from bare metal - ...and more, see below!

This was such a fun conversation with swyx 🔜 NeurIPS + #DevWritersRetreat and Jonathan Frankle I'm really proud of the work our team did, and it was great to get a chance to share some of the details.

Try to answer the question below. Confused? You’re not alone. 🍓🫐 When combing through LLM evaluation sets, we found many ambiguous, subjective, or unanswerable questions like this one. This is why it's hard to score 100% on some eval sets. How can a model get this right?

Had a great time at the AI Engineer World's Fair — thank you swyx 🔜 NeurIPS + #DevWritersRetreat and @LatentSpace for hosting

Largely correct, though the stochastic parrot people also seem to be taking the (weird) position that billions of dollars and thousands of smart people working on something will be insufficient to make significant improvements in that thing

Guess what isn't subject to the same rate limits? The slack web client. Maybe time to start scraping the data you need and make an unofficial API library?

Writing code is just the start. To move beyond prototypes, we need agents that plan, write specs, run tests, follow style guides, and catch bugs before you do. @JoshAlbrecht shared how we're tackling this with Sculptor at AI Engineer World’s Fair: youtu.be/x_1EumTaXeE?si…

Sculptor: the missing UI for Claude Code 🎨 Imagine running 5 Claudes in parallel, safely in containers, while you stay in flow. Then bring their work straight into your IDE to test/edit together. This is how one developer ships like a team. Try it with Sonnet 4.5!

Excited to host this workshop next week! We'll be sharing some lessons we've learned from building Sculptor--and we’re eager to hear how you’re using agents too.

It's a battle of the agents! ⚔️ Run 5 agents with different approaches to the same task. Toggle between them with one click, watch your app change instantly, then pick the winner and commit. That's Sculptor's Pairing Mode.

Happy SF Tech Week! Join us this Thursday for a workshop on how to use agents to write high-quality code. We'll share our favorite workflows with Sculptor and other tools, and would love to hear yours too. 🗓️ Oct. 9, 1–2:30pm RSVP: partiful.com/e/RW3jchCRNPpK…

Come help us share Sculptor with the world!