filipe
264 posts
- Replying to @willdepueEveryone feeling like they wanna cancel the holiday and just get back to work
- Happy to finally be able to talk publicly about this project! I'm glad to have had a chance to work on what I think is one of the most exciting developments in LLM deployment and usage so far :)We are adding support for plugins to ChatGPT — extensions which integrate it with third-party services or allow it to access up-to-date information. We’re starting small to study real-world use, impact, and safety and alignment challenges: openai.com/blog/chatgpt-p…
00:00 - We've just release our ICML paper on combining differentiable fluid simulations and graph neural networks to predict fluid flows. Joint work with Thomas Economon and @zicokolter. Paper: arxiv.org/abs/2007.04439 Poster (requires ICML registration): tinyurl.com/icmlcfdgcn
- Let me go cancel my Quora account, haven’t used it in years anyway…
- We have reached an agreement in principle for Sam Altman to return to OpenAI as CEO with a new initial board of Bret Taylor (Chair), Larry Summers, and Adam D'Angelo. We are collaborating to figure out the details. Thank you so much for your patience through this.
- Someone make a keyboard app out of this.Fun with the system message on GPT-4.
- Presenting our work on differentiable physics for learning and control at #NeurIPS2018 today: goo.gl/XezmES Watch the spotlight at 09:55 at Room 220 CD inclusive-or stop by our poster (#38) during poster session A from 10:45 to 12:45 PM at Room 210.
- Replying to @willdepueI was one step from buying it and cancelled because of that. I’m literally just waiting for an equivalent competitor that charges a fixed price (even if slightly more)
- maybe ea should go back to its Singer roots: help the people most in need today, quit trying to make impossible calculations of how actions now will butterfly into gigantic effects in the future
- Listening to @dwarkesh_sp now feels like discovering one of those indie bands in the 00s before they exploded I'm leaving it registered here that I subscribed so I can claim my coolness points later -- you should too!To help pay the bills for my podcast, I'm turning on paid subscriptions on Substack No major content will be paywalled - please don't donate if you think twice before buying a cup of coffee But if you have the means & have enjoyed my podcast, I would appreciate your support 🙏
- Replying to @filipeabperesEmbedding a differentiable fluid simulation into a graph net makes it able to generalize to previously unseen physical behavior, while predicting flows faster than a full simulation. Additionally, it is able to operate directly on CFD meshes, unlike regular convolutional methods.
- Replying to @tszzlhaving no AGI makes the mission of "ensuring it benefits all humanity" vacuously easy, gotta give it to them
