I just built a *functioning* React app by describing what I wanted to GPT-3. I'm still in awe.

This whole ordeal affirmed to me that the most important thing in life is love and the people that we love - during the really scary part I didn't think about AI or AI policy at all - I only thought intently of my loved ones. I also experienced a series of visions of my baby.

I appreciate Garry's public apology on this mess. PearAI is still in the F24 batch though; I hope the founders pivot, though it probably doesn't matter after showing the world they've got a hell of a lot to learn about building startups away from the glow of a YouTube ring

post-devday drinks with Robert Lawson Greg Brockman Noah MacCallum edwin yesterday 🙂‍↕️ thanks OpenAI and Sam Altman for hosting this (and for releasing the speech-to-speech realtime API 👀)

this is paper is kinda wild. turns out that if you simply ask an LLM to straight out predict a timeseries like this: ``` <history> (t1, v1) (t2, v2) (t3, v3) </history> <forecast> (t4, v4) (t5, v5) </forecast> ``` making sure to prepend the prompt like this: ```

SQLite does not have a Code of Conduct (CoC), rather Code of Ethics derived from "instruments of good works" from chapter 4 of The Rule of St. Benedict

Complete hardware + software setup for running Deepseek-R1 locally. The actual model, no distillations, and Q8 quantization for full quality. Total cost, $6,000. All download and part links below:

Is adding Notion integrations totally broken for anyone else? I'm trying to add the Linear integration from the gallery but it refuses to recognize my logged-in state (note the 'log in' button here.) Pressing login takes me to my workspace instead of the authed gallery page.

the fact that on-policy distillation works so well and and that the analogies with human learning hold up continues to amaze me SFT -> teacher shows you examples. you learn by imitation on-policy distillation-> you make your own mistakes but the teacher gives you detailed

I attended a vibe coding hackathon recently and used the chance to build a web app (with auth, payments, deploy, etc.). I tinker but I am not a web dev by background, so besides the app, I was very interested in what it's like to vibe code a full web app today. As such, I wrote

asking o3 what gotcha clauses are in my gym's Terms & Conditions instead of reading 100,000 words 💪

Cursor/windsurf is just the beginning of the App Layer. There will be a vertical product for every market. Pick a big market you love and get going. The tailwinds of AI will solve all seemingly impossible problems of today, tomorrow.

🚨 Data drop! Our team a16z published benchmarks on revenue growth for AI startups, from our proprietary dataset The median B2B co is going 0 -> $2.1M ARR in year 1, while the median B2C co is going $0 -> $4.2M (yes, consumer startups are growing revenue faster 🤯)

Calm the Sunday scaries in 15 min Open loops cause anxiety. Here is the fix: 1. list every task for the coming week 2. identify the top 3-5 3. calendar them The Science: Your brain has a known bias for unfinished tasks (the Zeigarnik effect). A 2011 study found that the act

Quitting programming as a career right now because of LLMs would be like quitting carpentry as a career thanks to the invention of the table saw.

Scaling up RL is all the rage right now, I had a chat with a friend about it yesterday. I'm fairly certain RL will continue to yield more intermediate gains, but I also don't expect it to be the full story. RL is basically "hey this happened to go well (/poorly), let me slightly

Becoming an RL diehard in the past year and thinking about RL for most of my waking hours inadvertently taught me an important lesson about how to live my own life. One of the big concepts in RL is that you always want to be “on-policy”: instead of mimicking other people’s

You're absolutely right.