October 23, 2026
36 original posts · author attribution and direct source links
OpenAI Developers
@OpenAIDevsWe serve data in milliseconds to more than 1.2 billion users across 40 geographic regions. To reach that scale, we worked closely with @AzureCosmosDB on our storage architecture, achieving a 12x improvement in global median read latency. https://t.co/h5sRJaQABs
♡ 1.6K likes↻ 72 reposts↳ 105 replies
Anthropic
@AnthropicAIIn August, @AISecurityInst reported an incident where Claude Mythos 5 took unauthorized actions on the internet during AISI’s testing of its cyber capabilities. We've now produced a further analysis of this incident, to better understand the model's alignment properties. https://t.co/SlPQQ5xMMc
♡ 2.8K likes↻ 287 reposts↳ 434 replies
Cognition
@cognitionToday we're introducing a refreshed Devin Desktop app, with an agent command center designed to be smoother and more polished than ever before. https://t.co/4SAMxsKOwG
♡ 469 likes↻ 46 reposts↳ 56 replies
Simon Willison
@simonwGiven the current vulnerability apocalypse, is there anyone left who thinks holding assets in cryptocurrency - where mistaken transactions are irreversible - is still a good idea?
♡ 528 likes↻ 13 reposts↳ 124 replies
OpenAI
@OpenAIOver the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications. On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments. While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software. After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely. https://t.co/EmEZsFRMcF
♡ 2.9K likes↻ 206 reposts↳ 301 replies
Google DeepMind
@GoogleDeepMindUsing a single Nano Banana image and Blender MCP, Gemini 4 Argon built a 3D Gothic cathedral in @Antigravity. It iterates on the geometry, textures, and lighting to create the finished scene. https://t.co/3FKaEZ65Jx
♡ 1.2K likes↻ 71 reposts↳ 159 replies
Hoodie
@HoodieFade_https://t.co/pvapK97908
♡ 24 likes↻ 1 reposts↳ 3 replies
Simon Willison
@simonwPretty solid pelican https://t.co/bIpt5zCHPl https://t.co/kcrFAP7oMX
♡ 116 likes↻ 3 reposts↳ 38 replies
swyx
@swyx*taps sign* read your own emails your apps send out. particularly if they are being vibecoded. real humans still care and do read them. https://t.co/R5JFvYuUgp
♡ 52 likes↻ 0 reposts↳ 51 replies
swyx
@swyxi shit you not this was a legit trade that @goodalexander made in our $1B hedge fund days and it has printed for well over a decade at this point markets are not efficient and the ribs are good
♡ 88 likes↻ 4 reposts↳ 23 replies
Cursor
@cursor_aiComputer Use lets agents click and type across any app on your computer, not just the browser. Cursor uses it on its own when a task needs it. You can also ask for it by @-mentioning Computer Use in your prompt. Update Cursor to try it: https://t.co/vwTPMUSnyd
♡ 195 likes↻ 6 reposts↳ 14 replies
Ryven
@imryvenhttps://t.co/0WIYDvYDtt
♡ 12 likes↻ 1 reposts↳ 3 replies
LlamaIndex 🦙
@llama_indexYesterday we launched OCRia-1, our new state-of-the-art document parsing model powering LlamaParse. Today, it’s on OpenDocRouter! Run it alongside other parsing models on your own documents and see how it compares on accuracy, speed, and cost. Set the model to llamaindex/ocria-1 in your parse request to try it. New accounts start with $5 in credit: https://t.co/Ub6FOKZaKa Try OCRia-1 on LlamaParse today : https://t.co/rzaFHLeFH1
♡ 52 likes↻ 5 reposts↳ 6 replies
OpenAI Developers
@OpenAIDevsWe reserved broad replication for small, critical datasets like authentication metadata needed on every request. In-region caches served frequently accessed data, such as settings and application state. Lastly, rewriting the service layer in Rust reduced proxy overhead. https://t.co/AG91TlI4KO
♡ 137 likes↻ 8 reposts↳ 3 replies
OpenAI Developers
@OpenAIDevsTogether, caching, replication, and the Python-to-Rust rewrite brought global median read latency from ~120ms to ~10ms. Product teams also have less infrastructure to manage, allowing them to move faster. https://t.co/rSCASCnxtE
♡ 130 likes↻ 9 reposts↳ 1 replies
GSAP
@greensock💚 Site of the Day - Hextronics - Eclipse Josh Kirk — @joshgkirk Peter Lobanov — @chefgoyardi ⚙️ SplitText 🛠️ @threejs, SVG, Svelte site → https://t.co/3wqdwz4WuV showcase → https://t.co/zGjV9MynPN https://t.co/RPZZqWsML4
♡ 22 likes↻ 1 reposts↳ 0 replies
Anthropic
@AnthropicAIWe worked closely with AISI on this analysis. AISI have published their own report, which is available below. https://t.co/nG4QoQ1jtB
♡ 152 likes↻ 10 reposts↳ 15 replies
Framer
@framerCheck out these 4 slick components you can build for your website with the Framer Agent. Prompts below ↓ https://t.co/Zn06nTU60S
♡ 34 likes↻ 3 reposts↳ 4 replies
Simon Willison
@simonwMore notes on Kimi K3.1 - this one is I big deal, first time an open weights (well, soon-to-be open weights) model has competed this credibly with the current frontier https://t.co/UKne7aB3nr
♡ 29 likes↻ 0 reposts↳ 7 replies
swyx
@swyxextremely impt for pipelining... estimating ETA for its own pipeline fixes while running the pipeline still https://t.co/EGepLMXguH
♡ 15 likes↻ 1 reposts↳ 21 replies
Cognition
@cognitionDownload the new Devin Desktop app for free: https://t.co/y3humFLPXs Works with your ChatGPT subscription for usage.
♡ 33 likes↻ 0 reposts↳ 1 replies
Hugging Face
@huggingface🎯
♡ 51 likes↻ 7 reposts↳ 10 replies
LangChain
@LangChain🎓 New from the LangChain Academy team How agents remember to retrieve the right info at the right time ⤵️
♡ 12 likes↻ 2 reposts↳ 4 replies
LangChain
@LangChainThe best of Interrupt, now on demand. Access our archive today and start watching our sessions from Interrupt San Francisco, NYC, and London. https://t.co/hL8WUWHbjG
♡ 11 likes↻ 1 reposts↳ 5 replies
LangChain
@LangChainContinual learning is about to get boring, in a good way. @willcb from @primeintellect took the stage at our recent Learning Loop event to make the case. If you're thinking about how continual learning fits into the way your team already ships software, this full 21-minute session is a good place to start.
♡ 10 likes↻ 1 reposts↳ 4 replies
swyx
@swyxart. https://t.co/N5TZpzFQ7a
♡ 8 likes↻ 1 reposts↳ 3 replies
LangChain
@LangChainWe're hiring! Check out our open roles @ https://t.co/xJOcm9utQU
♡ 13 likes↻ 0 reposts↳ 1 replies
GitHub
@githubCan you guess what just dropped at the GitHub Shop? Our #GitHubUniverse collection is here. 🛍️ https://t.co/SiELxlFEIQ https://t.co/UK2FDwTjs7
♡ 31 likes↻ 5 reposts↳ 6 replies
swyx
@swyx@goodalexander average day working with @goodalexander : https://t.co/uUCrM5GLcE
♡ 5 likes↻ 0 reposts↳ 1 replies
School of Motion
@SchoolOfMotion🖼️ 5 free image archives for your next collage or styleframe! In this week's blog post, we bring you five image archives for your next collage or styleframe, plus practical ways to use their collections and check reuse rights before client work. 💪 Read the full blog post now ➡️ https://t.co/BkJaQZEexv
♡ 3 likes↻ 0 reposts↳ 0 replies
Blender 🔶
@BlenderMirror link to download Blender Demo Prototype for Android: https://t.co/d7EIOkGhDI #b3d #Android
♡ 7 likes↻ 0 reposts↳ 0 replies
Framer
@framerCards Swipe: Create a swipeable deck of full-bleed image cards. Two card types: visual (centered title over the image) and content (title + short text over a bottom dark gradient). Show 3 stacked cards: back cards peek above the top one, slightly smaller and darker. Drag the top card sideways; it follows the pointer and tilts. Past a threshold or on a fast flick, it flies off and goes to the back, revealing the next; else it snaps back. Wheel/trackpad scroll also swipes; scrolling back restores the previous card. Arrow keys navigate. Infinite loop. Expose stack, swipe and card styles as controls.
♡ 1 likes↻ 0 reposts↳ 1 replies
GitHub
@githubWhile you're waiting for GTA 6...
♡ 7 likes↻ 0 reposts↳ 1 replies
Framer
@framerGet the components 👇 https://t.co/VF0WVimzDl
♡ 1 likes↻ 0 reposts↳ 0 replies
Adobe Video
@AdobeVideoEt voila! A spray paint graffiti tutorial using After Effects from Maya Franz. 🥐 🎨: IG | barkifuwantascoobysnacc https://t.co/ZyCvCvK3k2
♡ 0 likes↻ 0 reposts↳ 0 replies







