Claude Skills are awesome, maybe a bigger deal than MCP https://t.co/1wIYcTFrzI

Let me pitch you Skills: - They allow anyone to customize agents with a simple primitive — files - They give us continuous learning until continuous learning arrives - They are powerful, composable, and sufficiently AGI-pilled - you can now say “probably a skill issue” at work

Multi-agent search was easy to prototype (V1 only took me a day) but incredibly complex to bring to production. Feels like building RAG in 2022: we see the promise but are still figuring out the primitives. Sharing our journey so far - hope it helps your exploration!

My workflow for the past few months: - Write a design doc (with Claude ofc) - Ask Claude to generate tests - Verify the tests - Let it cook

Introducing Claude 3.7 Sonnet: our most intelligent model to date. It's a hybrid reasoning model, producing near-instant responses or extended, step-by-step thinking. One model, two ways to think. We’re also releasing an agentic coding tool: Claude Code.

Success in the LLM space isn't about building the most sophisticated system. It's about building the right system for your needs. Had a lot of fun writing this with @ErikSchluntz!

Graph-based memory that you can customize is now open-source with MCP https://t.co/m1n5Zkmalr https://t.co/k67mGSwrdi

Case study on https://t.co/sXP5wkc5HK from @AnthropicAI: - Zero to $4m ARR in 4 wks - 100k+ of y'all using Bolt every wk - One of fastest growing AI tools globally Wild & surreal! Ty for all the support & patience as we scale y'all 🙇 Let's go!

Introducing an upgraded Claude 3.5 Sonnet, and a new model, Claude 3.5 Haiku. We’re also introducing a new capability in beta: computer use. Developers can now direct Claude to use computers the way people do—by looking at a screen, moving a cursor, clicking, and typing text.

Just saying...🍓

K-Scale Labs just launched on @ycombinator's Launch YC! K-Scale Labs 🤖 Open-source humanoid robots for everyone. Check them out:

@TeraflopAI is excited to help support the @caselawaccess and @HarvardLIL, in the release of over 6.6 million state and federal court decisions published throughout U.S. history.

Hosting an event for our clients and leaders in compliance, financial services and law, on February 6 at @NYSE to discuss AI Agents & the Law. The event will begin with a live one-on-one discussion between @johnjnay and Lawrence H. Summers, Harvard University Professor and

How much data do you need to fine-tune GPT-3.5? @barry_zyj et al. have a great article showing that you only need ~100 datapoints for better output structuring and customizing the tone. Lots of great stuff incl. cost/latency analysis. Check it out! https://t.co/nKU2ebHKSE

"Enough to gauge your incompetence"

This works on many levels

How to break GPT-3.5 through fine-tuning... more coming soon

Q: What's better than Mixture of Experts? A: Mixture of Mixture of Experts! Introducing GodMode: the AI Chat Browser https://t.co/uql5ZPkCTj Fast, Free, Access to ChatGPT, Bing, Bard, Claude, YouChat, Poe, Perplexity, Phind, and Local/GGML Models like Vicuna and Alpaca No

I founded a second AI company. We're pioneering an empirical, practical approach to AI alignment, leveraging law/policy & LLMs. Backed by top venture firms and AI researchers in the Bay Area & NYC, we're hiring 5 more Members of our Technical Staff. DM me your CV if interested

This is what we are using to evaluate LLMs? (QUASAR-T)