Your curated collection of saved posts and media
We’ve made three new AI fluency courses, co-created with educators, to help teachers and students build practical, responsible AI skills. They’re available for free to any institution. https://t.co/nK2D3W5YcU
We’re also announcing a new Higher Education Advisory Board, which helps guide how Claude is used in teaching, learning, and research. Read more about the courses and the Board: https://t.co/TorRcYMHnd
New Anthropic research: filtering out dangerous information at pretraining. We’re experimenting with ways to remove information about chemical, biological, radiological and nuclear (CBRN) weapons from our models’ training data without affecting performance on harmless tasks. https://t.co/YUBlLKIL2c
One concern is that filtering CBRN data will reduce performance on other, harmless capabilities—especially science. But we found a setup where the classifier reduced CBRN accuracy by 33% beyond a random baseline with no particular effect on a range of other benign tasks. https://t.co/24xCQBjejh
If you’re interested in joining us to work on these and related issues, you can apply for our Research Engineer/Scientist role (https://t.co/x3G4F5qVWv) on the Alignment Science team.
Medical research moves faster with GPT-5. Professor @DeryaTR_ demonstrates its impact. https://t.co/0WN0gZY0Io
@jxnlco Look inside the environments/ directory of atropos to find out https://t.co/FZLYzPzubi
@rosstaylor90 @paperswithcode We havent built agentic environments yet but we are building an extremely flexible RL environments framework called Atropos - Building up infra still and finding what works before moving on to more advanced trajectories :) https://t.co/FZLYzPzubi
Why is no one talking about this? This is why I don't use an AI browser You can literally get prompt injected and your bank account drained by doomscrolling on reddit: https://t.co/aGN8xrdZtD
AI agents that can browse the Web and perform tasks on your behalf have incredible potential but also introduce new security risks. We recently found, and disclosed, a concerning flaw in Perplexity's Comet browser that put users' accounts and other sensitive info in danger. http
With $2.2 billion in CHIPs grants Intel's received so far, the total investment adds up to about $11.1 billion, which is close to 9.9%. Source: https://t.co/DZLY7SeIPk
He pointed out on X, "There's zero evidence of AI consciousness today. Source: https://t.co/0Eo2dsp6tj
Terms and conditions apply. Source: https://t.co/37NjKkEV1o
🚀📈 @exostellarcloud helps enterprises optimize their cloud infrastructure with #AI. At AWS NYC Summit, @TonyShakib discussed how AWS has expanded Exostellar’s customer reach. Join us at AWS Summit Los Angeles to find out more from AWS experts. 👉 https://t.co/I5BfDqng9F https://t.co/KIRGpUDmbU
https://t.co/2QoL4x5Fj3 https://t.co/yPjR5g3uLm
On Saturday at the Buildathon hosted by AI Fund and https://t.co/zpIxRSuky4, over 100 developers competed to build software products quickly using AI assisted coding. I was inspired to see developers build functional products in just 1-2 hours. The best practices for rapid engineering are changing quickly along with the tools, and I loved the hallway conversations sharing tips with other developers on using AI to code! The competitors raced to fulfill product specs like this one (you can see the full list in our github repo; link in reply): Project: Codebase Time Machine Description: Navigate any codebase through time, understanding evolution of features and architectural decisions. Requirements: - Clone repo and analyze full git history - Build semantic understanding of code changes over time - Answer questions like “Why was this pattern introduced?” or “Show me how auth evolved” - Visualize code ownership and complexity trends - Link commits to business features/decisions Teams had 6½ hours to build 5 products. And many of them managed to do exactly that! They created fully functional applications with good UIs and sometimes embellishments. What excites me most isn’t just what can now be built in a few hours. Rather, it is that, if AI assistance lets us build basic but fully functional products this quickly, then imagine what can now be done in a week, or a month, or six months. If the teams that participated in the Buildathon had this velocity of execution and iterated over multiple cycles of getting customer feedback and using that to improve the product, imagine how quickly it is now possible to build great products. Owning proprietary software has long been a moat for businesses, because it has been hard to write complex software. Now, as AI assistance enables rapid engineering, this moat is weakening. While many members of the winning teams had computer science backgrounds — which does provide an edge — not all did. Team members who took home prizes included a high school senior, a product manager, and a healthcare entrepreneur who initially posted on Discord that he was “over his skis” as someone who “isn't a coder.” I was thrilled that multiple participants told me they exceeded their own expectations and discovered they can now build faster than they realized. If you haven’t yet pushed yourself to build quickly using agentic coding tools, you, too, might be surprised at what you can do! At AI Fund and https://t.co/zpIxRSuky4, we pride ourselves on building and iterating quickly. At the Buildathon, I saw many teams execute quickly using a wide range of tools including Claude Code, GPT-5, Replit, Cursor, Windsurf, Trae, and many others. I offer my hearty congratulations to all the winners! - 1st Place: Milind Pathak, Mukul Pathak, and Sapna Sangmitra (Team Vibe-as-a-Service), a team of three family members. They also received an award for Best Design. - 2nd Place: David Schuster, Massimiliano Viola, and Manvik Pasula. (Team Two Coders and a Finance Guy). - Solo Participant Award: Ivelina Dimova, who had just flown to San Francisco from Portugal, and who worked on the 5 projects not sequentially, but in parallel! - Graph Thinking Award: Divya Mahajan, Terresa Pan, and Achin Gupta (Team A-sync). - Honorable mentions went to finalists Alec Hewitt, Juan Martinez, Mark Watson and Sophia Tang (Team Secret Agents) and Yuanyuan Pan, Jack Lin, and Xi Huang (Team Can Kids). To everyone who participated, thank you! Through events like these, I hope we can all learn from each other, encourage each other, invent new best practices, and spread the word about where agentic coding is taking software engineering. [Original text: https://t.co/wJbQMrnZdL ]
The products that teams worked on at Buildathon: https://t.co/iA48xG9yU2
my longtime friend, AI legend @AndrewYNg joined me and @eladgil on @NoPriorsPod this week -- he thinks AI can really help with investing, but it can't get at some of the people judgement or transparency borne of relationships (yet) https://t.co/iaYwkxLR5w
Enjoyed chatting w @AndrewYNg * Bleeding Edge of Agentic AI * Will Models Bootstrap Themselves? * Vibe Coding vs. AI Assisted Coding * Successful Founder Profiles * Next Industry Transformations https://t.co/FCrIY0P6jn
Jon asks why the GOP's "one is too many" logic applies only to immigrants and not mass shootings https://t.co/gkddPI7qyl
Jon asks why the GOP's "one is too many" logic applies only to immigrants and not mass shootings https://t.co/gkddPI7qyl
NYT: The Trump administration on Friday ordered that all construction stop on Revolution Wind, a $4 billion wind farm off the coast of Rhode Island that is already mostly built. Matthew Giacona, the acting director of the Bureau of Ocean Energy Management, issued a letter on Friday to Orsted, the Danish company building the wind farm, ordering it to “halt all ongoing activities” because of unspecified issues. This. Is. Insane.
52 million jobs have been created since the end of the Cold War. 3 Republican administrations. 3 Dem administrations. 96% of jobs created were under DEMOCRATIC administrations. The last three GOP administrations have one thing in common: RECESSIONS. https://t.co/Wc4KbSuOCD
@LiorOnAI i watched the talk and tl;dr is @ylecun believes generative models/LLMs don't scale because they're trying to predict without understanding they're predicting pixels but they don't understand the underlying physics and world he believes world models + JEPA are the answer (1/2) https://t.co/8fyuHTsPPQ
👀 lack of perception depth in AI A four-year-old has seen more data than an LLM! interesting talk by @ylecun https://t.co/ZnbUh4Y14e
One of the first pruning methods for neural nets came in 1989: Optimal Brain Damage by @ylecun et al. “We … derive a class of prac- tical and nearly optimal schemes for adapting the size of a neural network.” They pruned weights with low 2nd-derivative of the loss. https://t.co/WzIDQExX87
@mingyang2666 Congrats on the acceptance, Mingyang! That's a very timely and interesting study. Btw, the below paper by @cheeesio in your bibliography doesn't seem to exist (at least I wasn't able to find it). Did you perhaps mean to cite https://t.co/syQWIDG6np? https://t.co/mpKEnGEpbp
@mingyang2666 @cheeesio No worries. You might also want to double check the other references as some of them may have different names/authors, e.g., this one: https://t.co/iS5j18CPV6
~55 million foreigners on visas ~25 million illegal immigrants 80/340 = 24% 1/4 of the country isn’t American. https://t.co/XuCdYqplKL
BIG NEWS: The United States of America now owns 10% of Intel, one of our great American technology companies. This historic agreement strengthens U.S. leadership in semiconductors, which will both grow our economy and help secure America’s technological edge. Thanks to Intel CEO @LipBuTan1 for striking a deal that’s fair to Intel and fair to the American People.
@NoBanksNearby a new definition of “vibe” that I’m genuinely excited about https://t.co/R1mjoYe9q7
A useful thing that GPT-5 can do that wasn’t previously possible before powerful AI is to monitor complex topics by asking it to give you scheduled reports. Example: I have a weekly report on “reproducible, benchmarked evidence of autonomous or recursive self‑improvement in AI” https://t.co/Q4MeLVWxpp
Turn any LLM into AI agents that can reason, retrieve knowledge, and use tools in a few lines of code. > Build lightning-fast Agents that can generate text, image, audio and video. > Add memory, knowledge, tools and reasoning as needed. > Run anywhere, open-source. https://t.co/psgkuizNKX
