My 7061 Saved posts

New Anthropic research: Tracing the thoughts of a large language model. We built a "microscope" to inspect what happens inside AI models and use it to understand Claude’s (often complex and surprising) internal mechanisms. https://t.co/PboGlLFnHG

After reading like 15 pages of a PDF of this book I found online, I immediately ordered a physical copy. I can't stress just how good it is. The 10 year-old me would never imagine he'd one day choose a book on the oil industry instead of gaming... https://t.co/08JVq4ewxj

To fellow writers who have been PROCRASTINATING so bad today: think of your audience. I mean *actually* visualize their face/imagine what they think as you're typing (your essay-paper-thesis-book-what have you) down. Such an effective remedy to writer's block—at least for me.

4o image generation has arrived. It's beginning to roll out today in ChatGPT and Sora to all Plus, Pro, Team, and Free users. https://t.co/pFXDzKhh2t

📣 Today, we’re introducing Gemini 2.5, our most intelligent AI model. An experimental version of Gemini 2.5 Pro is available now in the Gemini app for Gemini Advanced users: https://t.co/973tlgjcKq Let’s get into this update ⬇️🧵 https://t.co/AWccU7rrtb

Think you know Gemini? 🤔 Think again. Meet Gemini 2.5: our most intelligent model 💡 The first release is Pro Experimental, which is state-of-the-art across many benchmarks - meaning it can handle complex problems and give more accurate responses. Try it now →… https://t.co/bFcx0IlY24

Introducing Gemini 2.5, our most intelligent AI model. Our first release, an experimental version of 2.5 Pro, unlocks state-of-the-art performance in math and science. 🔥 Learn more 🧵 https://t.co/aoe7egliJb

It's finally here: Brampton Brampton is the world's most intelligent, creative, and fastest model. Brampton dramatically outperforms Grok 3, Claude 3.7 Sonnet, and GPT 4.5. Reply with "brampton" for early access. https://t.co/K7g2mq17SC

🚀 DeepSeek-V3-0324 is out now! 🔹 Major boost in reasoning performance 🔹 Stronger front-end development skills 🔹 Smarter tool-use capabilities ✅ For non-complex reasoning tasks, we recommend using V3 — just turn off “DeepThink” 🔌 API usage remains unchanged 📜 Models are… https://t.co/QVuPwCODne

The world bends faster than you think. Get crystal clear on what you want. Move like it’s already yours. No hesitation. No backup plan. Just full pressure. Suddenly, things shift. People show up. Doors open. It looks like luck. It’s not.

Today we are announcing ARC-AGI-2, an unsaturated frontier AGI benchmark that challenges AI reasoning systems (same relative ease for humans). Grand Prize: 85%, ~$0.42/task efficiency Current Performance: * Base LLMs: 0% * Reasoning Systems: <4% https://t.co/k9o7t031fN

we built Cursor for 3D modeling. https://t.co/dWNPEOkS6Q

Introducing Together Chat! Use DeepSeek R1 (hosted in North America) & other top open source models to do web search, coding, image generation, & image analysis. Available today for free! https://t.co/OLHwCd7Tpj

Best thing I’ve read recently. https://t.co/FOksO0RDkR

After 7 months, I finished my essay on high agency. So I asked myself: "What's the most high agency way of promoting it?…

After 7 months, I finished my essay on high agency. So I asked myself: "What's the most high agency way of promoting it?" Last night, I took over the iconic Times Square in New York for $0. Hilarious story... https://t.co/U5xcF4707E

“Steven, I don’t have a big marketing budget to advertise my app” No problem. 1: Open TikTok and search your keywords 2: Save viral videos from your competition 3: Recreate those viral videos yourself 4: Post 30+ and see what works for you 5: Find a winner based on conversion

If you have a child that can articulate themselves at an early age and exudes confidence , you have an obligation to ensure that the environment you place them in does not break this out of them.. you cannot surround them with adults with poor self esteem and ignorance

life is too short. execute those ideas, & hug your mom more often.

a trap younger ppl fall into & end up becoming "lazy" is that due to high info/knowledge exposure, they get to cultivate much more taste than skill their skill severely lags behind taste & by not being afforded environments that satisfy their taste & simultaneously being unable…

This article title goes extremely hard https://t.co/4JfYHjbd41

being able to ask any question i can think of with zero judgement has genuinely helped keep my childlike curiosity alive https://t.co/2rJ3GKWjNG

Beberapa abad lalu sudah disarikan Imam Syafi'i. Menuntut ilmu punya enam syarat utama: kecerdasan, semangat, kesungguhan, biaya, bimbingan guru, dan waktu yang lama. https://t.co/xmh1OfjikV

new one minute blog: big Mario marketing https://t.co/VQRyXpxV6u

Meta recently made a 1 character change to their codebase which saves the equivalent of ~15,000 servers in capacity per year 🤯 https://t.co/pGZ1JjCwLc

The future is unevenly distributed. https://t.co/ww6WTozUes

In mitochondria, respiratory supercomplexes (blue) pump protons across crista membranes to create an electrochemical gradient, fueling adenosine triphosphate (ATP) synthase (pink) to regenerate ATP—the energy currency of life. Now, cryo–electron tomography captures these… https://t.co/DKk8fBW47j

Meet Liner Deep Research! ↳ The most accurate AI-powered deep research ↳ One simple question, 360 degree in-depth reports ↳ 10 free reports daily Liner Deep Research is here to help students, researchers, and professionals save time conducting researches and verifying… https://t.co/iQJ6edfCCH

Hot take: Anybody should have the possibility to knock at a University door saying "I did this cool thing working at home the last three years, can you give me a phd title" Three experts would review the request, we would have a three-day "defence" to avoid cheats. 1/2

All hands on deck. Figma Slides is now available to everyone https://t.co/kJyp3WdjAa