Your curated collection of saved posts and media
Wow. Gemini 3.0 on Radiology's Last Exam The first time a general-purpose model has beaten radiology residents with 51% accuracy. Radiology trainees are at 45%. The main significance is that a general model has finally reached a level where it can compete with early-stage human training on a specialized medical exam. Congratulations to @GoogleDeepMind team. @GeminiApp
🔥 Gemini 3.0 vs Radiologists: RadLE Benchmark Results Are OUT! ☠️ Is it game over for Radiology? Let us find out! ⬇️ 🫨 Since yesterday, Gemini 3.0 has been everywhere for crushing benchmarks. My inbox exploded asking: “But how did it do on the hardest visual reasoning benchmar
https://t.co/KHwr77LU2U
https://t.co/eORuGSr816
In case your local affiliates are banning you from hearing something that someone says. https://t.co/xe1VXDCPAe
Yes, he really did post this. It is in multiple news outlets. https://t.co/q1q8PO82Hk
He is a vandal driven by a pathetic little ego. This is the East Wing of the White House today, with no warning, no discussion. All of you who support him are rubes and victims, who revel in your victimhood. Losers. All of you. https://t.co/SBFAIoPloI
https://t.co/HlxGHdSG2n
Connect yourself to Trump or Musk who believe they are above the laws of our land and actually don't care about anyone but themselves and eventually you will suffer the punishment proscribed by our laws. https://t.co/7Ilb0nrg6M
📻 Brazen brute bashes bank! Czar’s zealots embezzle zillions! Cops can't quell criminal caper! https://t.co/LDA7bIswHp is at large for @theworldlabs launch! ‘Shadow’ is the jazz-infused, film noir-style narrative game made for @BenMildenhall & @drfeifei's AI research group 🕵️ https://t.co/xsQg8xqTzC
In 8 weeks we built a game that would have taken 12 months prior to this technology Drops in a week. How-to thread incoming. Thanks @theworldlabs <3 https://t.co/id3BWkuf6f
Dr. Fei-Fei Li (@drfeifei) is known as the “godmother of AI.” For the past two decades, she’s been at the center of AI’s most significant breakthroughs, including: - Spearheading ImageNet, the dataset that sparked the AI explosion we’re living through right now. - Leading work at Stanford Artificial Intelligence Laboratory (SAIL) - Serving as Chief Scientist of AI/ML at Google Cloud - Co-founding Stanford’s Institute for Human-Centered AI - Serving on the United Nations AI Scientific Advisory Board - Being named as Time's 100 most influential people in AI In this conversation, Fei-Fei shares the rarely told history of how we got to today—and what comes next. We discuss: 🔸 The backstory on ImageNet 🔸 Why robotics faces unique challenges compared with language models and what’s needed to overcome them 🔸 Why Fei-Fei believes AI won’t replace humans but will require us to take responsibility for ourselves 🔸 Why world models and spatial intelligence represent the next frontier in AI, beyond large language models 🔸 The surprising applications of Marble, from movie production to psychological research 🔸 How to participate in AI regardless of your role 🔸 Much more Listen now 👇 • YouTube: https://t.co/9uJdY3VreJ • Spotify: https://t.co/2ZzO89EmjF • Apple: https://t.co/pNOGR05t21 Thank you to our wonderful sponsors for supporting the podcast: 🏆 @Figma Make — A prompt-to-code tool for making ideas real: https://t.co/8iUp8fgO4x 🏆 @justworks — The all-in-one HR solution for managing your small business with confidence: https://t.co/ax9QeAVrWE 🏆 @wearesinch — Build messaging, email, and calling into your product: https://t.co/kfaPvA5HQs
It looks demo uses @sparkjsdev to render beautiful 3D Gaussian Splats 🤩 https://t.co/UZrNZbpRfv
Today we are releasing & open-sourcing Segment Anything 3 (SAM 3). It is a state-of-the-art model for image & video segmentation, and builds upon the work of SAM & SAM 2. SAM3 will also power features in Edits, Meta AI, & Facebook Marketplace soon. https://t.co
It looks demo uses @sparkjsdev to render beautiful 3D Gaussian Splats 🤩 https://t.co/UZrNZbpRfv
Researchers are exploring Marble’s generative 3D worlds as a way to rapidly produce simulation-ready environments for robotics without manual scene construction. https://t.co/qxHMzEyP6f
Most VLM benchmarks watch the world; few ask how actions *change* it from a robot's eye. Embodied cognition tells us that intelligence isn't just watching – it's enacted through interaction. 👉We introduce ENACT: A benchmark that tests if VLMs can track the evolution of a home-scale environment from a robot's egocentric view. 🌐https://t.co/2yBqVcGiKk 📄https://t.co/iqe3HgSMnD 1/N
Big news! CDS is joining forces with @NYU_Courant to become the new Courant Institute School of Mathematics, Computing, and Data Science. The new school will include CDS together with the Depts of Mathematics and Computer Science. More info: https://t.co/98YPaUR78I 🧵1/4 https://t.co/rntcYFbkYQ
well someone has been preaching this at us for like 6+ years glad we are past the 'feel the agi' phase and back to building toward human-level intelligence https://t.co/Qp8FyLPaQU
“The thing that happened with AGI and pretraining is that in some sense they overshot the target. You will realize that a human being is not an AGI. Because a human being lacks a huge amount of knowledge. Instead, we rely on continual learning. If I produce a super intelligent
If a foreign adversary snuck into our Federal budget and cut Education and Science Research the way we’re cutting it ourselves — strategically undermining America’s long-term health, wealth, and security — we would likely consider it an act of war. https://t.co/qshkf8Vllu
Recent media misreport — about Meta’s AI orgs and (oddly) myself — clarifications: 🧪 FAIR is Meta’s long-term research lab — not GenAI, not MSL 🔍 We do open research with public data, no access to GenAI/MSL infra 😅 I’m not bald ⏳ No complaint — just asking folks to be patient https://t.co/fZoejFSk4e
No matter how AI evolves overnight—tech, career, how it may impact me—I remain committed to using "physics of language models" approach to predict next-gen AI. Due to my limited GPU access at Meta, Part 4.1 (+new 4.2) are still in progress, but results on Canon layers are shining
SAM 3D is helping advance the future of rehabilitation. See how researchers at @CarnegieMellon are using SAM 3D to capture and analyze human movement in clinical settings, opening the doors to personalized, data-driven insights in the recovery process. 🔗 Learn more about SAM 3D: https://t.co/WAtASpkTdY
NEW data I've received: Just 5% of people detained by ICE since October 1 have had violent criminal convictions, 3/4 had no criminal convictions at all. Most "criminals" had immigration, traffic, and vice offenses. Not the "worst of the worst"... https://t.co/KASOlWz1v9
this is what Anthropic wanted > that “AI cyberattack” article was the setup > they walk into congress with it > push their anti-opensource agenda into regulation > make it harder for people to own their intelligence resist by making sure everyone knows NOT to trust them https://t.co/nkD8105GsU
Anthropic is testifying to Congress in December after catching the first real AI-assisted cyberattack. Cyber defense is becoming AI vs AI, and every serious government is going to start building and hoarding compute in response
The misleadingly named “Hygiene Hypothesis” is more accurately framed as the “Old friends hypothesis.” We co-evolved with friendly bacteria & some parasites. We did NOT co-evolve with the *crowd infections* of mass travel, mega-cities, & indoor confines. 4/ https://t.co/6Lvxm0vRAr
Read more details, with links to lots of research here: Scientists Just Connected the Dots Between Viruses and… Everything https://t.co/vwrPwSP2Hw 7/
For a long time, catching frequent viruses was considered both inevitable and harmless. But it turns out that common, seemingly-mild viruses have disturbing long-term health impacts. 1/ https://t.co/ya2BgrTAQC
"People who go all in on AI agents now are guaranteeing their obsolescence. If you outsource all your thinking to computers, you stop upskilling, learning, and becoming more competent. AI is great at helping you learn." @jeremyphoward @NVIDIAAI https://t.co/s2ZIeHK3sq 2/
I wrote my first blog post in 2015 & was very surprised when it went viral. Here I reflect back on 10 years of blogging and some of my most impactful posts, on themes of challenging conventional wisdom, debunking AI hype, & democratizing ML 1/ https://t.co/38PnTQsP6Y
Deep learning is not just for images and text (2018). Companies such as Pinterest and Instacart are also applying it to tabular data, the type of data you might normally put in a spreadsheet. 4/ https://t.co/2anbHkJXCK
In 2018, I was nervous to disagree w Google CEO & head of AI when they said all data scientists need automated NN architecture search. However, I was invited to keynote the AutoML workshop as a result and 7 yrs later my critique has proven valid. 5/ https://t.co/OGgTabXibf
AI ethics is not just a theoretical topic (2019). I was (and still am) alarmed about the harms being caused to human beings by AI systems irresponsibly applied to healthcare, employment decisions, policing, & more. 6/ https://t.co/l4pqBe6duF
(2022) My husband & I never planned to homeschool, but we unexpectedly found that our child thrives with this approach. Fits with a theme of challenging conventional wisdom and often doing things differently. 7/ https://t.co/20as8kCBnz
(2024) A 3-part series on how AI is being applied to try to predict T cell binding. Accurate predictions would impact a range of treatments, but this problem is far from solved. 8/ https://t.co/dP7sD0gvQd
