Jon Hulsinger

@lauriewired I tried signing up and I immediately got flagged for using a US number. Rats.

The hardware in old Chinese cloud accelerator cards never fails to impress me. If you go on Chinese ebay (idlefish) you can get a Xilinix UltraScale FPGA for ~$50 USD. For perspective, the same raw chip is currently ~$2,100 on Mouser.

@ericmitchellai @markchen90 Needs an IOS app to make it accessible from anywhere and more robust thinking workflow that doesn't break down when cellular temporary disconnects.

why isn't chatgpt the perfect personal AGI? what is most disappointing about it? what feature, model improvement, or bugfix would do the most to make it more useful in your daily life? what is most frustrates you that chatty can't do, or can't do well enough?

@rohanvarma Codex on computer remotely accessible through either its own app or on ChatGPT app would be sick.

Every time I ask yall for feedback on Codex, you say two things: 1. SSH dev box support in the app 2. View your files in the app Both are now available with our launch yesterday!! What’s left?

@Scobleizer Someone make a dubstep remix of this video.

My reality is blurry. But when I put on my Apple Vision Pro it becomes sharp. But when you realize you aren’t seeing reality at all your mind gets freed.

@kimmonismus Whether you outsource reads to radiologists in India or hand them to an AI, the liability math is the same. The people doing the actual interpretation don’t carry the malpractice exposure. That stays with the health system and whoever signed off locally. This isn’t a new problem.

I just read this news, and it makes sense. Up until now, it's always been said that more AI in application (using radiology as an example) leads to more people being able to see radiologists due to increased efficiency, and therefore more human radiologists would be needed despite AI. This logic doesn't make sense to me because radiologists are a cost factor for hospitals; hospitals are generally (to put it neutrally) businesses that make calculations. More human radiologists, who have to be paid, are a cost factor. If patients could be billed simply by using the AI ​​model, it would be cheaper. So I wasn't surprised when I read this news.

@Suryanshti777 I have accomplished this simply by explaining that I want it to talk to another ai agent. Then it just figured it out...this isn't rocket science.

Holy shit… someone just made Claude instances talk to each other. Not APIs. Not agents. Not orchestrators. Just multiple Claude Code sessions… messaging each other like coworkers. It’s called claude-peers — and it turns one Claude into a team. Here’s what’s happening: Run 5 Claude Code sessions across different projects Each one auto-discovers the others They send messages instantly Ask questions Share context Coordinate work Your AI tools literally collaborate. Example: Claude A (poker-engine): "what files are you editing?" Claude B (frontend): "working on auth.ts + UI state" Claude A: "ok I'll avoid touching auth logic" No conflicts. No manual coordination. Just AI syncing itself. Under the hood: • Local broker daemon (localhost) • SQLite peer registry • MCP servers per session • Instant channel push messaging • Auto peer discovery • Cross-project communication Everything runs locally. No cloud. No latency. What it unlocks: • Multi-agent coding without frameworks • One Claude writes backend, another frontend • One debugs while another refactors • Research Claude feeds builder Claude • Large projects split across AI workers This is basically: "spawn 5 Claudes and let them coordinate themselves" Even crazier: Each instance auto-summarizes what it's doing Other Claudes can see: • working directory • git repo • current task • active files They know what the others are working on. Commands: • list_peers → find all Claude sessions • send_message → talk to another Claude • set_summary → describe your task • check_messages → manual fallback So you can literally say: "message peer 3: what are you working on?" …and it responds instantly. No orchestration layer. No agent framework. Just Claudes… talking. This is the cleanest multi-agent system I've seen. We're moving from: 1 AI assistant → to AI teams that coordinate themselves. And it's all running on your machine. Wild.

Cloudflare Containers & Cloudflare Sandboxes now both support Docker-in-Docker, allowing you to run a Docker daemon inside a container that is itself running on a Docker host. An example use case for Docker-in-Docker would be running a fully sandboxed development environment. Taken from the horse's mouth, this allows you to: - Develop containerized applications with your Sandbox - Run isolated test environments for images - Build container images as part of CI/CD workflows - Deploy arbitrary images supplied at runtime within a container I believe this was a fairly highly requested feature, so I'm excited to see what people build with it!

This is going to be needed for efficient practical alerts from future smart glasses. Currently most smart glasses already run on a flavor of android:

Introducing Glimmer ✨ @Google’s new design language for UX on glasses! Glimmer prioritizes voice, gesture, and eye-tracking — focusing on glanceable, transient elements that appear only when needed. Discover thoughtful design for a new dimension: goo.gle/4qFBG5S

@kimmonismus DGI has been achieved internally. 🐭

An interesting twist: it's not OpenAI that pays to use the Disney license, but Disney that invests in OpenAI. Disney is investing $1 billion in OpenAI and has struck a three-year licensing deal that lets Sora generate AI videos featuring over 200 Disney-owned characters from franchises like Marvel, Pixar, and Star Wars. Starting in 2026, fans will be able to create and share short Sora videos with these characters - some of which may even appear on Disney+ - while Disney also becomes a major enterprise customer of ChatGPT for internal tools and products.

@ChShersh I built a new version of grep last night if you are interested I can send you a GitHub link when it’s finished. So far I am seeing a P95 latency metric of 9ms.

Here’s a real task from my job. I have a 100GB binary file. Produced daily. I can’t grep it. But I can decode it. However, I can’t store the decoded version either. It’s too big. How do I efficiently query it? Decoding piped to grep takes 2 minutes. I want 2 seconds.

@BatCaveNode @TheCleanCarClub @grok is this true?

@TheCleanCarClub Did you know, It’s illegal to fly a drone over anyone in the USA? 👀🦇

This man deployed built in drone to see traffic ahead.

@ram_ssk what’s going on with vector embeddings getting spat out on Microsoft Copilot GPT-5 ?

@SpencerHakimian @grok is this true?

The housing market is done.

@ztivom @saraaa7447 @yngmisu Modern iPhones (IP67/IP68) are water-resistant, not perfectly airtight. They have pressure equalization vents with membranes. If the phone is hot and you put it in a fridge, the temperature drops fast. Air inside contracts, and the vents pull in outside humid air.

@saraaa7447 @yngmisu The phone is literally waterproof to 12 feet of depth...

Okay so Do NOT do that.

@liteclips You have to use the flash with any high speed photography like this and you only used the flash on the Nokia.

Nokia vs Iphone

@ASilverbill @elhackernet @grok It indeed is can tell the hands glitched partway through the video

@NotebookLM Would like spaced repetition flash cards feature as well

NOW we are officially, officially, OFFICIALLY rolled out to 100%!

@Cartidise Samsung S7 Press Release: news.samsung.com/global/faces-o…

Does Tim Cook know ANYTHING about the competition? He says Apple invented vapor chamber on smartphones lmao “uniquely Apple’s innovation; perhaps only we would do that” It literally has existed on phones for over 7 years now? Many phones even have much larger vapor chambers. How can you say “only Apple would do that” when you don’t even care to know that others have already done that?

@zephyr_z9 Whose buying them they’re on sale!

> be Adobe, 40-year-old PDF jockey > 2025, stock doing a perfect -33% swan dive > “We’ll pivot to AI” says exec on 7-figure retention bonus > can’t ship a model because legal says every pixel needs a 12-page EULA > Midjourney drops v7, makes our Firefly look like MS Paint with a hangover > OpenAI drops GPT-Image, Google drops nano-banana, both free > our response: “Please login with your Adobe ID, install Creative Cloud, update 47 GB, restart, then pay $53.99/month” > users collectively Alt-F4 into orbit > watch in horror as ChatGPT/Gemini reads any PDF you give it for free > enterprise cancels 10k seats overnight > try to counter with Sora killer video model > training cluster catches fire after someone uploads a 1998 clipart library > PR tweet: “We are re-imagining creativity” > quote-tweet ratio hits 1:9k, gif of dumpster bonfire tops replies > premiere pro is now just a bloated launcher for 15 different subscription prompts > 20-something with a phone and CapCut is making better edits > our flagship feature: “Generative fill but now 3% slower” > board meeting: “Let’s raise prices again” > stock drops another 8% during the Zoom call > our most innovative feature in 5 years is a "subscribe to annual plan" button that clicks itself