Your curated collection of saved posts and media
AI inside Unity is getting real. ๐ Join us on Open Source Friday with Andy Tsen to dive into Unity MCP. We'll cover: ๐ค How AI agents talk to Unity ๐ฎ What "context" actually means in a game engine โ๏ธ How to start building AI-assisted workflows Set a reminder for tomorrow's stream ๐ https://t.co/ByFvpdTaUJ
One more step towards a world without language barriers. Real-time translations just became even easier in the Google Translate app!ย Now you can get live translations with headphones on iOS (itโs already available on Android!).ย ๐ง Also available in more markets and over 70 languages! ๐
every arxiv paper on https://t.co/mlJE5Cr4AP https://t.co/IlTclXMqz7
Amazing. https://t.co/LEN8O2hA2v
"Top Secret!" (1984) is super silly and funny. This is a classic action comedy parody film by the trio Zucker-Abrahams-Zucker (the team behind Airplane! and The Naked Gun). Val Kilmer plays the lead role โ this is also his film debut. https://t.co/DlzQUO7bzc
๐ never been a better time to start yapping! ๐ welcome to the world, gemini 3.1 flash live: https://t.co/686c32EnWc
Introducing Gemini 3.1 Flash Live, our new realtime model to build voice and vision agents!! We have spent more than a year improving the model + infra + experience, the results? A step function improvement in quality, reliability, and latency. https://t.co/0esYpmDy5l
๐ never been a better time to start yapping! ๐ welcome to the world, gemini 3.1 flash live: https://t.co/686c32EnWc
make me a billion dollars, make no mistakes https://t.co/VmfVObzI2Q
We have been surpassed: AI written output exceeded human written output in 2025 https://t.co/Dv4CNJDMVf
Trying this new โAccounts you follow and who they followโ reply filter on my latest post that Elon reposted. As soon as I turned it on the number of spam replies collapsed. I donโt think Iโll turn it on for all my posts, but for posts like this it looks like a great feature. https://t.co/wVF66GhMhT
Q1 is almost over and SpaceX has expanded its lead over the rest of the world. One company launches more rockets than all other companies *and* world governments combined (that includes other American companies and NASA) https://t.co/aZZ54ySx1U
GPU MODE (@GPU_MODE) & PyTorch Foundation are organizing an ML systems hackathon in Paris on April 9, immediately following PyTorch Conference Europe 2026. Researchers and engineers will compete across two tracks: -Distributed training (LLM speedrun) and inference optimization (leaderboard) - Access to a B300 cluster from Verda and H200 instances from Sesterce - Cloud credits as prizes, including 48-hour access to a GB300 NVL72 rack - Talks from PyTorch (Helion), vLLM, Prime Intellect, and more - Food and refreshments Doors open at 9:30, with the closing ceremony at 20:00. Attendees can join with a pre-formed team or match on-site. Location details are shared upon registration. Spots are limited. Register: https://t.co/1BYgfGWJIr #OpenSourceAI #PyTorchCon #PyTorch #vLLM #Helion
Building systems that can execute the scientific research lifecycle has been an incredible journey for our team. For those interested in diving in, we released our implementations for both versions of The AI Scientist: โธป V1: https://t.co/A5NDvfZKLC V2: https://t.co/gedB5EIMCg โธป
Lyria 3 came out from google today, so naturally I had to use it for generative lullabies. now, when the monitor detects the baby is crying, it can play a 30 sec lullaby pre-generated to the specific baby's tastes (based on what's worked historically) using Lyria 3 Clip. this demo showcases this generative lullaby capability, whereas the last showcased playing soothing phrases in the mother's cloned voice to calm the baby. this is using the @openhome devkit as hardware. more to come!
GPU MODE and PyTorch Foundation are organizing an ML systems hackathon in Paris on April 9, immediately following PyTorch Conference Europe 2026. Researchers and engineers will compete across two tracks: -Distributed training (LLM speedrun) and inference optimization (leaderboard) - Access to a B300 cluster from Verda and H200 instances from Sesterce - Cloud credits as prizes, including 48-hour access to a GB300 NVL72 rack - Talks from PyTorch (Helion), vLLM, Prime Intellect, and more - Food and refreshments Doors open at 9:30, with the closing ceremony at 20:00. Attendees can join with a pre-formed team or match on-site. Location details are shared upon registration. Spots are limited. Register: https://t.co/1BYgfGWJIr #OpenSourceAI #PyTorchCon #PyTorch #vLLM #Helion
When @karpathy built MenuGen (https://t.co/2OjrUJ3aLS), he said: "Vibe coding menugen was exhilarating and fun escapade as a local demo, but a bit of a painful slog as a deployed, real app. Building a modern app is a bit like assembling IKEA future. There are all these services, docs, API keys, configurations, dev/prod deployments, team and security features, rate limits, pricing tiers." We've all run into this issue when building with agents: you have to scurry off to establish accounts, clicking things in the browser as though it's the antediluvian days of 2023, in order to unblock its superintelligent progress. So we decided to build Stripe Projects to help agents instantly provision services from the CLI. For example, simply run: $ stripe projects add posthog/analytics And it'll create a PostHog account, get an API key, and (as needed) set up billing. Projects is launching today as a developer preview. You can register for access (we'll make it available to everyone soon) at https://t.co/1tSgGbSLxM. We're also rolling out support for many new providers over the coming weeks. (Get in touch if you'd like to make your service available.) https://t.co/vjRymcVCKI
Say hello to Gemini 3.1 Flash Live. ๐ฃ๏ธ Our latest audio model delivers more natural conversations with improved function calling โย making it more useful and informed. Hereโs whatโs new ๐งต
Meet Gemini 3.1 Flash Live ๐๏ธ๐๏ธ Our first Gemini 3-based multimodal, real-time model is officially here. Try the preview now on the Live API and in @GoogleAIStudio ๐ https://t.co/4OxcX7glyG https://t.co/XMp9hIsUfK
We just launched Gemini 3.1 Flash Live! Our fastest, most natural real-time voice AI model for building Agents. - Scores 90.8% on ComplexFuncBench Audio for tool use. - 70 languages, Video streaming, Audio transcriptions, 128k context - Comes with Agent Skill for building live voice agents. - All generated audio is watermarked with SynthID.
New Voxtral-4B-TTS is beautiful ๐ it sounds so good that it's whispering at closed models: "your time is over" - Demo available on Hugging Face โฌ๏ธ https://t.co/gI02Jmo9j5
Bounding boxes are key for citations, and we just shipped a new guide showing how to use LiteParse for visual citations! LiteParse is our fast and open-source document parser. Using both bounding box extraction and page screenshots, anyone (including agents) can learn how to associate text with an element on the page. https://t.co/Vauhx5Yh9n
Gemini's audio and voice capabilities just got an upgrade with Gemini 3.1 Flash Live. Our new high-quality audio and voice model comes with: โก๏ธ Faster response times ๐ฌ More helpful, natural dialogue ๐งต 2x longer conversation memory in Gemini Live ๐ Multilingual support for Search Live in 200+ countries & territories See whatโs new ๐งต
Just dropped on HF: Cohereโs cohere-transcribe-03-2026 > ๐ฅ #1 on the Open ASR leaderboard > ๐ #4 multilingual > ๐ #6 long-form > Supports 12+ languages: English, German, French, Italian, Spanish, Portuguese, Greek, Dutch, Polish, Arabic, Vietnamese, Chinese, Japanese, Korean > Conformer-based encoder + lightweight Transformer decoder for transcription > And of course: Apache 2.0 license
Cohere has released an Apache 2.0 model on @huggingface. No restricted license. No "research only." Actually open. Respect. Is this a one-off or a direction change for Cohere? https://t.co/i204YRbe98
Introducing: Cohere Transcribe โ a new state-of-the-art in open source speech recognition. https://t.co/l87Z6oyQdM
@cohere just released the best speech->text model :) It currently ranks #1 for accuracy on @huggingface Open ASR Leaderboard, setting a new benchmark for real-world transcription performance. Read more ๐ https://t.co/WopGr8Wu9P
Introducing: Cohere Transcribe โ a new state-of-the-art in open source speech recognition. https://t.co/l87Z6oyQdM
Q1 is almost over and SpaceX has expanded its lead over the rest of the world. One company launches more rockets than all other companies *and* world governments combined (that includes other American companies and NASA) https://t.co/aZZ54ySx1U
Will people avoiding computer science end up looking like these 1987 studies? โFive years ago, computers looked like they were the land of good money and easy opportunityโ https://t.co/APtEshUctt
Jevons paradox is happening in real time. Companies, especially outside of tech, are realizing that they can now afford to take on software projects that they wouldnโt have been able to tackle before because now AI lets them do so. Weโre going to start to use software for all n
Realism isn't just power prevailing. Realism also requires looking at the world realistically to recognize the limits of a nation's power and to deploy it only for compelling purposes. Trump's war-of-choice on Iran isn't realism but reckless stupidity. https://t.co/vc2C9Nz034
As the United States turns 250, the countryโs population is more racially and ethnically diverse than in previous decades. Three-quarters of U.S. adults see racial and ethnic diversity as a good thing for the country. Read more about Americansโ views on diversity here: https://t.co/ngrBuEWe5O
At GTC, @NVIDIAAIDev announced general availability for NuRec, their library for gaussian splatting (uses 3DGUT) for Physical AI. It's now available through @NVIDIAAI NGC Catalog Article: https://t.co/5zK7VAhBkx Access: https://t.co/66PNiUMOb8 https://t.co/LaQQeo0ABf
๐Introducing Voxtral TTS: our new frontier open-weight model for natural, expressive, and ultra-fast text-to-speech ๐ญRealistic, emotionally expressive speech. ๐Supports 9 languages and accurately captures diverse dialects. โกVery low latency for time-to-first-audio. ๐Easily adaptable to new voices