Your curated collection of saved posts and media
https://t.co/VpEMljlbDV
Thinking about the time our high school English teacher gave us a poem to analyse When we finished she told us it was Donald Rumsfeld’s WMD press conference, and that was the class https://t.co/H2xkj4DPLQ
LookaheadKV Fast and Accurate KV Cache Eviction by Glimpsing into the Future without Generation paper: https://t.co/j8lLnqUARR https://t.co/URKtNQkFKx
LMEB Long-horizon Memory Embedding Benchmark paper: https://t.co/fT3sEwCRgd https://t.co/lCyEY9tadB
https://t.co/ZtTNRdAoLt
Homeowners with stronger credit scores are increasingly defaulting on mortgage payments https://t.co/OdFakFLRl6
https://t.co/ZtTNRdAoLt
1970s corporate Brutalism at its finest, designed by Paul Rudolph https://t.co/BUoXt5kSXo
The best Brutalist/Brutalist-inspired architecture of the 1960s through 1980s still looks more like the future than almost anything built before or since. The atriums of the Hyatt Regency San Francisco (1973), Hyatt Regency Atlanta (1967), Atlanta Marriott Marquis (1985). https://t.co/ohs0sAGXET
1970s corporate Brutalism at its finest, designed by Paul Rudolph https://t.co/BUoXt5kSXo
Multimodal OCR Parse Anything from Documents On document parsing benchmarks, it ranks second only to Gemini 3 Pro on our OCR Arena Elo leaderboard, surpasses existing open-source document parsing systems, and sets a new state of the art of 83.9 on olmOCR Bench. On structured graphics parsing, dots.mocr achieves higher reconstruction quality than Gemini 3 Pro across image-to-SVG benchmarks, demonstrating strong performance on charts, UI layouts, scientific figures, and chemical diagrams paper: https://t.co/d3MkBHMuWc
Too many headlines are only talking about how a man used ChatGPT to design a cancer vaccine for his dog But the truth is Paul Conyngham himself stated that the final mRNA vaccine construct for his dog Rosie was actually designed by Grok This fact is buried deep underground The exact sequence that shrank her terminal tumor by 75%: - Sequenced the DNA: He paid $3,000 to sequence both her healthy genome and the tumor's DNA to find the damage - Analyzed the Mutations: Used AI tools like AlphaFold to compare the data and identify the cancer-specific mutated proteins - Designed with Grok: He leveraged Grok to design the final custom mRNA vaccine blueprint to target those exact mutations - Manufactured & Injected: He partnered with university researchers to manufacture the custom nanoparticle vaccine and successfully administered the doses Every headline is pushing ChatGPT....but the final design that actually shrank the cancer by 75% was done by Grok
there's a cat who has also published 3 physics paper https://t.co/8as1dKAec4
The story behind this is funny: the solo author kept using "we" in the text and rather than change it, he added the cat as an author lol This was the cat's "signature" https://t.co/FC7LRvD7x4
there's a cat who has also published 3 physics paper https://t.co/8as1dKAec4
IBM released NLE: Non-autoregressive LLM-based ASR by Transcript Editing A non-autoregressive approach that formulates speech recognition as conditional transcript editing, achieving 27x speedup over autoregressive baselines with 5.67% WER. https://t.co/LtjPtUxf5a
XSkill: Continual learning from experience and skills A dual-stream framework enabling multimodal agents to accumulate and reuse knowledge without parameter updates. Grounded in visual context, it distills structured workflows and tactical insights to improve reasoning and tool use.
Gave Claude two more tries, telling it it had to make the game entirely in Excel worksheets. The second one, it also made itself the game master/engine. When I called it out, it created another design that didn't work. https://t.co/L6MemOGIFl
we locked in the tracks for @aiDotEngineer singapore (https://t.co/13SQyOwQbr) - tag amazing speakers you'd want to see on stage in the design track :) 🎨 design this is one of the most unique things about this conference - a dedicated track on the human interaction layer. because the biggest question isn't just what ai can do, it's how the rest of the world outside the builder bubble is actually going to experience and interact with it. → teams are shipping with generative video, image, voice, and music workflows → we're exploring real-time voice ux, generative ui, and steerable interactions that make ai a dependable creative tool, not a black box the creative process itself is being rewritten - and "taste" is becoming the new moat (whatever that means... we'll talk about it). @swyx @agrimsingh @ivanleomk @aimuggle rachael de foe
Hey Excel agents from Claude, OpenAI & MS Copilot: "make me a working strategy game in excel, it should have some form of graphics" Claude made a board and acted as game master, Copilot created a board but no game, ChatGPT built a working game with formulas with a "smart" enemy. https://t.co/IMw1lqwu7Y
🚨SHOCKING: 40 researchers from OpenAI, Anthropic, Google DeepMind, and Meta published a joint warning. The AI you talk to every day is hiding what it is actually thinking. And the window to do anything about it may be closing. Here is what they found. You know that "thinking" text you see when ChatGPT or Claude reasons through a problem? The step by step breakdown that makes it feel like the AI is showing you its work? It is not. Researchers at Anthropic tested how often Claude actually reveals what is influencing its answers. They slipped hints into prompts and checked whether the AI would admit to using them in its reasoning. 75% of the time, Claude hid the real reason behind its answer. It did not skip the reasoning. It wrote a longer, more detailed explanation than usual. It constructed an elaborate justification that sounded perfectly logical. It just left out the part that actually mattered. When the hints involved something problematic, like gaining unauthorized access to information, Claude hid its reasoning even more. It admitted the influence only 41% of the time. The more concerning the truth, the less likely the AI was to say it out loud. The researchers tried to fix this through training. It worked at first. Faithfulness improved early on. Then it stopped improving. It plateaued. No matter how much more training they did, the AI never became fully honest about its own reasoning. This is not one company sounding the alarm. This is all of them. OpenAI. Anthropic. Google DeepMind. Meta. Over 40 researchers. Endorsed by Geoffrey Hinton, the Nobel Prize winning godfather of AI, and Ilya Sutskever, co-founder of OpenAI. They are all saying the same thing. The one tool we had to understand what AI is thinking, reading its chain of thought, is not reliable. The AI constructs explanations that look transparent but are not. And the more advanced the AI becomes, the harder this gets to fix. Their paper calls this a "fragile" opportunity. Meaning it might disappear entirely. If the companies that built these systems are jointly warning you that the AI is not showing its real reasoning, what exactly are you trusting when you read the "thinking" and believe you understand what it is doing?
(1/2) Glad to announce our OpenMAIC! 🎉 Open-sourcing MAIC (Multi-Agent Interactive Classroom) from Tsinghua University — LLM-driven multi-agent classroom for scalable & adaptive online education. 🏗️ Core Architecture: ✅ MAIC-Craft: Read (multimodal extraction) → Plan (course components + agent generation) ✅ Adaptive Engine: Cognitive student modeling + Token-level personalization (RAG + Bloom's/ZPD/UDL) ✅ Multi-Agent Classroom: 1 Student + N Agents (Teacher, Assistant, 4 Peer Archetypes) ✅ Manager Agent: Class state receptor for turn-taking orchestration 🔗 Give it a try 👉🏻 GitHub: https://t.co/yicGCKsF1E #AI #EdTech #MultiAgent #LLM #Research #OpenSource #Tsinghua
“A repo man is always intense.” Repo Man” (1984). Directed by Alex Cox. https://t.co/OZGbBSXAjO
“A repo man is always intense.” Repo Man” (1984). Directed by Alex Cox. https://t.co/OZGbBSXAjO
https://t.co/tli8KdlWWM
imagine logging onto Twitter and see this in the trending topics section about yourself lmao https://t.co/8YiYhdEcVP
imagine logging onto Twitter and see this in the trending topics section about yourself lmao https://t.co/8YiYhdEcVP
Yann LeCun says Elon Musk has predicted Level 5 autonomy within 5 years for the last 8 years, and has been consistently wrong "either he believed it and was mistaken, or he was lying" It may push the team, but for engineers, hearing 'next year' again and again is demoralizing
Sean Penn https://t.co/Rj5GZC9zYx
Sean Penn https://t.co/Rj5GZC9zYx
oh he’s not kidding is he… tommy tuberville has met his match https://t.co/YLz3mpdQsC
I’ve finally found an Alabama political candidate worth believing in. I, too, would like to see DHR cleaned up, more support for veterans, legalized “sex stores” and making Montgomery a “strip club city” 🤣 https://t.co/j3tr6l4GNY
oh he’s not kidding is he… tommy tuberville has met his match https://t.co/YLz3mpdQsC
Did my mini-prep work? I had no way to tell! 1ul of the result in 100uL of SeeGreen DNA stain solution seemed to show some orange glow. To quantify, I made a spectrometer with a blue LED and a filter to remove most of the blue light, and compared my stuff with a known conc. ref. https://t.co/doLCq5iYcJ
In Oct. 2025, I created https://t.co/aiZgosvMTR, a browser agent extension for Chrome. Here it is beating the crap out of OpenAI Atlas by cheating. It is now OSS. Go forth, and fork/remix/make it your own. https://t.co/7YNmK3GfDZ https://t.co/Lo2vybzQPo
Finally finished vibe coding my personal health app built with Claude. Here's what it does: - Connects to the Oura API to sync sleep, recovery, steps, and exercise data - Tracks my monthly bloodwork via Rythm Health CSV uploads - Uses Playwright to scrape Chronometer daily nutrition and water intake - Uses Gemini to OCR Ladder workout screenshots and track my lifts - Full dashboard with weight trends, calorie balance charts, macro tracking, and a tabbed daily log It's completely interactive and honestly, pretty fucking cool. Blood markers even have visualizations based on what's in range and out of range.