AI Tweets

Capability overhang means too many gaps today between what the models can do and what most people actually do with them. 2026 Prediction: Progress towards AGI will depend as much on helping people us...

OpenAI@OpenAI·2 days ago

one thing I truly love about @ampcode is how opinionated they are being. this talk from @beyang starts by rejecting MCP slop and then only gets spicier from there (subagents, @mitchellh's thread shari...

swyx (Latent Space)@swyx·2 days ago

Agents

Emad Mostaque@EMostaque·3 days ago

I find automated math proving with AI not hugely interesting What is most interesting is how we can leverage AI to enable more intuition and inspiration like Ramanujan Save the proofs for later, dar...

Powered by Gemini 3, the @YouTube Playables Builder web app can help creators develop fun, bite-sized games with text, video or image prompts. 🕹️ Find out more → https://t.co/mxyV4dTqLH https://t.co/...

Google DeepMind@GoogleDeepMind·4 days ago

Full Blown RPG in your browser: No Downloads - Just Click and Go!

swyx (Latent Space)@swyx·4 days ago

ToggleX is now LIVE-the first of its kind. ToggleX is an Ambient Work Intelligence Fabric that sits across your tools and turns everyday activity into work intelligence: context, insights and simple ...

Hugging Face@huggingface·4 days ago

What is the best physics podcast

Emad Mostaque@EMostaque·4 days ago

I don’t think Lean is that useful (yet) for proving anything major. All the labs should get together and apply huge compute to upgrading and filling the gaps in mathlib first, would be useful for mod...

Emad Mostaque@EMostaque·4 days ago

Sleigh the season with the most personal gift around. Get them a Cameo video!

Sasha Rush@srush_nlp·5 days ago

Computer Vision

Anthropic@AnthropicAI·6 days ago

We’re releasing Bloom, an open-source tool for generating behavioral misalignment evals for frontier AI models. Bloom lets researchers specify a behavior and then quantify its frequency and severity ...

Safety

Emad Mostaque@EMostaque·6 days ago

diffusion is how reality generates itself

Computer Vision

Andrej Karpathy@karpathy·6 days ago

https://t.co/Lb6T42n5jl

You can now adjust specific characteristics in ChatGPT, like warmth, enthusiasm, and emoji use. Now available in your "Personalization" settings. https://t.co/7WSkOQVTKU

OpenAI@OpenAI·6 days ago

As part of our partnership with @ENERGY on the Genesis Mission, we're providing Claude to the DOE ecosystem, along with a dedicated engineering team. This partnership aims to accelerate scientific di...

Anthropic@AnthropicAI·7 days ago

GPT-5.2-Codex is now available in Codex. It sets a new standard for agentic coding in real-world software development and defensive cybersecurity. It also delivers more reliable performance on compl...

OpenAI@OpenAI·7 days ago

People use AI for a wide variety of reasons, including emotional support. Below, we share the efforts we’ve taken to ensure that Claude handles these conversations both empathetically and honestly. ...

Anthropic@AnthropicAI·7 days ago

📌 PINNED CHATS ARE HERE 📌 PINNED CHATS ARE HERE 📌 PINNED CHATS ARE HERE Rolling out now to iOS, Android, and web. Tap the "..." next to your chat on web or long press on mobile to pin.

OpenAI@OpenAI·7 days ago

Ever wonder what your Browser extensions are really doing? Or what data that random VSCode plugin is siphoning off?

Jason Wei@_jasonwei·7 days ago

Ever wonder what your Browser extensions are really doing? Or what data that random VSCode plugin is siphoning off?

Sasha Rush@srush_nlp·7 days ago

AI has the potential to compress the time needed for new discoveries from years to days. 📉 That’s why we’re supporting the US Dept. of @ENERGY's Genesis Mission – providing National Labs with access ...

Google DeepMind@GoogleDeepMind·7 days ago

You can now ask @GeminiApp: "Is this video made with AI?" 🔍 Upload the file and it will check for the SynthID watermark to help verify if it was created or edited by Google tools. Find out more → ht...

Research

Google DeepMind@GoogleDeepMind·7 days ago

I love the expression “food for thought” as a concrete, mysterious cognitive capability humans experience but LLMs have no equivalent for. Definition: “something worth thinking about or considering, ...

Andrej Karpathy@karpathy·8 days ago

Full Self-Driving Supervised improves US road safety by over 80%, saving lives & preventing injuries https://t.co/VfO021Xra8

Sasha Rush@srush_nlp·8 days ago

Safety

Sasha Rush@srush_nlp·8 days ago

Excellent read. Clear and concise from an expert on this topic.

Watch how Gemini 3 Flash compares to 2.5 Pro when generating SVG code and simple web apps. ↓ https://t.co/4C2Nino7Eg

Google DeepMind@GoogleDeepMind·8 days ago

Phantom Blade Zero launches worldwide on September 9, 2026 on Steam, PlayStation, and the Epic Games Store. Join the journey and wishlist today to stay connected!

Sasha Rush@srush_nlp·9 days ago

📢Palworld: Home Sweet Home is OUT NOW! 25% off on Steam right now! ･ULTRAKILL Collab ･Building Improvements ･Base Painting ･New Raid ･Raid Battlefield ･Melee Combat Improvements ･Experimental PvP ･St...

Sasha Rush@srush_nlp·9 days ago

Replay Robinhood Presents: YES/NO? Yes. https://t.co/sfKri5tA41

Sasha Rush@srush_nlp·9 days ago

Google DeepMind@GoogleDeepMind·9 days ago

⚡⚡⚡

How will AI affect education, now and in the future? Here, we reflect on some of the benefits and risks we've been thinking about. https://t.co/WjdBLUzWat

Anthropic@AnthropicAI·9 days ago

We released an updated Gemini 2.5 Flash Native Audio model for live voice agents. 💬 It's now even better at following instructions and holding more natural conversations. Find out more → https://t.co...

Google DeepMind@GoogleDeepMind·10 days ago

Our AI safety and security research partnership with the @AISecurityInst is expanding. 🤝 Together, we’ll focus on critical areas – from monitoring how models “think” to understanding their social and...

Google DeepMind@GoogleDeepMind·10 days ago

Everyone sees the model. No one sees what keeps chat apps alive at scale. At 500 QPS, chat stops being an LLM problem. It becomes a systems problem: tail TTFT, kv-cache hit rates, sampler correctness...

Safety

Yann LeCun@ylecun·11 days ago

NEW GAME Raise your ideal daughter in Magical Princess!! In a Magical kingdom, become a father for the ultimate daughter. Fully voiced and over 50 endings!

Jason Wei@_jasonwei·12 days ago

Margin Call and The Social Network take place at the same time, with roughly the same characters, but with very different aesthetics.

Sasha Rush@srush_nlp·13 days ago

We need policies that champion newer, modern recycling to keep up with demand for products made with recycled plastic.

Sasha Rush@srush_nlp·14 days ago

nanoGPT - the first LLM to train and inference in space 🥹. It begins.

Andrej Karpathy@karpathy·15 days ago

Quick new post: Auto-grading decade-old Hacker News discussions with hindsight I took all the 930 frontpage Hacker News article+discussion of December 2015 and asked the GPT 5.1 Thinking API to do an...

Andrej Karpathy@karpathy·16 days ago

Happy weekend to those who celebrate https://t.co/lafvNmmOJO

Andrej Karpathy@karpathy·19 days ago

Building enterprise-grade .NET web applications? Start your next project the right way.

Yann LeCun@ylecun·23 days ago

Building enterprise-grade .NET web applications? Start your next project the right way.

Sasha Rush@srush_nlp·23 days ago

Building enterprise-grade .NET web applications? Start your next project the right way.

Jason Wei@_jasonwei·23 days ago

Building enterprise-grade .NET web applications? Start your next project the right way.

Andrej Karpathy@karpathy·23 days ago

🛑 Tired of AI hallucinations? Gurubase = Hallucination-Resistant AI Grounded in your docs, tickets & knowledge base - Works with Zendesk / Salesforce / Intercom - Cloud or 100% on-prem (zero dat...

Sasha Rush@srush_nlp·24 days ago

swyx (Latent Space)@swyx·24 days ago

Zero-days like ToolShell expose a familiar challenge: real environments aren’t clean, simple, or perfectly patched. Our new blog explores what actually works - strategic resilience, compensating contr...

Sasha Rush@srush_nlp·25 days ago

A number of people are talking about implications of AI to schools. I spoke about some of my thoughts to a school board earlier, some highlights: 1. You will never be able to detect the use of AI in ...

Andrej Karpathy@karpathy·about 1 month ago

I asked it to create a personalized weekly workout plan, and then posters that I can print on the wall to remind me what exercises to do each day. Tuesday looks more intense because I asked for "more ...

Andrej Karpathy@karpathy·about 1 month ago

As a fun Saturday vibe code project and following up on this tweet earlier, I hacked up an **llm-council** web app. It looks exactly like ChatGPT except each user query is 1) dispatched to multiple mo...

Andrej Karpathy@karpathy·about 1 month ago

🎄Celebrate family love with our Custom Engraved Wooden Bear Family Puzzle. Crafted from high-quality natural wood, this adorable puzzle features 2 to 6 bear figures, each symbolizing a cherished famil...

Andrej Karpathy@karpathy·about 1 month ago

Five reasons IT resists Macs and one blog that changes the conversation. Click for the full story. 👇

Sasha Rush@srush_nlp·2 months ago

Bullish, in the coming decades majority of compute will be spent on ai for science

Jason Wei@_jasonwei·3 months ago

Jason Wei@_jasonwei·4 months ago

Old friends, new lab

Enjoy the best experience on X.

Yann LeCun@ylecun·5 months ago

Sasha Rush@srush_nlp·5 months ago

Andrej Karpathy@karpathy·5 months ago

swyx (Latent Space)@swyx·5 months ago

Jason Wei@_jasonwei·5 months ago

Enjoy the best experience on X.

Andrej Karpathy@karpathy·5 months ago

Becoming an RL diehard in the past year and thinking about RL for most of my waking hours inadvertently taught me an important lesson about how to live my own life. One of the big concepts in RL is t...

Jason Wei@_jasonwei·5 months ago

New blog post about asymmetry of verification and "verifier's law": https://t.co/bvS8HrX1jP Asymmetry of verification–the idea that some tasks are much easier to verify than to solve–is becoming an i...

Jason Wei@_jasonwei·5 months ago

Bryan Johnson longevity mix is the most popular drink at ragers in SF https://t.co/zJE36CoV6V

Jason Wei@_jasonwei·6 months ago

We don’t have AI self-improves yet, and when we do it will be a game-changer. With more wisdom now compared to the GPT-4 days, it's obvious that it will not be a “fast takeoff”, but rather extremely g...

Jason Wei@_jasonwei·6 months ago

The most rewarding thing about working in the office on nights and weekends is not the actual work you get done, but the spontaneous conversations with other people who are always working. They’re the...

Jason Wei@_jasonwei·6 months ago

I would say that we are undoubtedly at AGI when AI can create a real, living unicorn. And no I don’t mean a $1B company you nerds, I mean a literal pink horse with a spiral horn. A paragon of scientif...

Jason Wei@_jasonwei·6 months ago

If you have a mouse, you have to play this game.

Sasha Rush@srush_nlp·over 1 year ago