Suha

The number of OS primitives that CUDA breaks is pretty impressive. :-)

We're hiring for Bombadil! Come work with me on the future of browser testing — specification languages and temporal logic, JS runtimes, bundlers, and WASM frontends in Rust, and lots and lots of property-based testing. Apply here: antithesis.com/company/career… Retweets appreciated.

Minecraft will really add something like “more breedable cows” and the craziest people you know will instantly say “aha, this lets us integrate third-order ODEs, now we can solve PSPACE problems in a tenth of the time”

10 things i learned about harness engineering - how to get the most out of coding agents TODAY (shouts out to @0xblacklight for the excellent post)

@trq212 I think part of it is lots of skill builders/publishers are not building them with deep llm intuition

one of my coworkers hard-coded a (restricted) WASM interpreter into the weights of a transformer, losslessly x.com/ChristosTzamos…

Legend of program verification Tony Hoare passed away at 93. I only ever met him once in @Cambridge_Uni as a student about 20 years ago and never had the privilege to collaborate with him, although I shared 3 co-authors.

LLMs are now superhuman at reward hacking our kernel competitions Natalia Kokoromyti, was #1 on last problem of the NVFP4 competition for around 10 min before we scrubbed the reward hack I know of very few humans who can write such a hack gpumode.com/news/reward-ha…

✨New demo: what if vibe coding felt more visual? @brian_lovin @maryrosecook and I did a game jam using Notion as our "IDE": launching Cursor agents from a task board, and making a custom image for each task 😎 The demo shows 3 ideas for the future of agents: 1) Agents should collaborate across apps. Each app has its focus--Notion AI is good at drafting specs and organizing tasks; Cursor is good at coding. So let them specialize! Today we're launching a new integration where Notion AI can kick off Cursor Cloud Agents to do coding tasks. The Cursor API accepts natural language prompts, so I think of this as "cross-app sub-agents" -- it's kinda cute how it resembles humans hiring outside contractors 😊 BTW: the parallelism of cloud agents is incredibly freeing for creativity, but it also creates a new problem: sooo much work to keep track of! Which brings us to the next idea... 2) Agent orchestration is a data visualization problem. A powerful frame for designing agent UIs is to think of the chat transcripts as the "raw data" and ask: what visual projections might help people make sense of this data at scale? We need to engage our human GPUs -- our visual processing -- to understand what the computer GPUs are doing for us! One thing we can do is use AI to populate traditional UIs like progress bars and status updates. But there are also new possibilities now... For example: when you have a lot going on, it can be hard to identify tasks just by text titles. So we tried generating an AI image for each task -- turns out this helps a lot by giving it a unique visual identity! And of course, it also just makes it super fun to build with friends 😃 Speaking of friends... 3) The future of coding is collaborative. Sometimes it feels like IC engineers are being reduced to middle managers: shuffling information between the team's context and the coding agents that they individually manage. The solution: bring all the people and agents into one shared space, with shared context and visibility! In the video you can get a glimpse of how this feels. Mary, Brian and I record ourselves chatting about ideas, and then we use AI to turn that conversation into a list of tasks on a shared board. As the ideas get built in parallel, we can all monitor progress and review the work together, nothing is siloed. My main takeaway from this game jam was: damn, creativity with friends, at the speed of conversation, is incredibly fun. --- Our goal here is to let anyone use Notion as a fun and creative "software factory" to build software together with your team. Give the Cursor integration a shot and let us know what you think! (AI Image gen in Notion isn't GA yet, but coming soon and already out to some users) And let me know if you'd want a template or more detailed instructions on the setup we showed in this demo...

MARCH 7TH 2026 THE GREAT SANDBOX SYMPOSIUM SAN FRANCISCO, CA real builders, getting together to try all the sandbox tech, comparing them across different dimensions + sharing research with each other at the end + no credits, no prizes, no judging, just hacking and learning see you there @jheitzeb @AITinkerers @blaxelAI @e2b @daytonaio

We’re excited to release TorchLean which is the first fully verified neural network framework in Lean. The Lean community has largely focused on pure mathematics. TorchLean expands this frontier toward verified neural network software and scientific computing. With the recent release of CSlib, we see this as another step toward a fully verified ML stack. We support features: 1. Executable IEEE-754 floating-point semantics (and extensible alternative FP models) verified tensor abstractions with precise shape/indexing semantics 2. Formally verified autograd system for differentiation of NN programs Proof-checked certification / verification algorithms like CROWN (robustness, bounds, etc.) 3. PyTorch-inspired modeling API with eager-style development + export/lowering to a shared IR for execution and verification Project page: leandojo.org/torchlean.html Paper: [2602.22631] TorchLean: Formalizing Neural Networks in Lean Work done @Robertljg, Jennifer Cruden, Xiangru Zhong, @huan_zhang12 and @AnimaAnandkumar. #MachineLearning #ScientificComputing #Lean

Pumped to be @runsybil running all things recruiting! We just opened a Head of Engineering role in our NYC office. Check it out below! jobs.gem.com/runsybil/am9ic…

Paged Out! #8 is out! pagedout.institute @pagedout_zine In "An AWKward Modem" (p. 28), I show how to write a tiny modem in 5 lines of AWK and shift it into the near-ultrasonic range. 🔊

This is an incredible post by @kylemathews that everybody should read If you only read one thing this week make it this

this is 100% true btw but at the same time, all the best engineers I know on this train are doing something so far from removed from what "big vibe coding" is pushing

"... code is only accepted if the AI proves mathematically the specs are fulfilled ..." The Universalis approach is to consider every AI as adversarial, and untrusted, and hence we cannot trust the proofs that are generated by an AI. So given a spec, the AI produces an implementation together with a proof that the implementation implements the spec. Then we check the proof independently, not using AI, and only then accept the code.

AI-assisted code review should from on be referred to as Clauditing

@seanhn I’m incredibly curious when (or if!) the progress in having LLMs check their own work and arguments (e.g., instead of calling out to Lean) will transfer over to correctly reasoning about the validity of vulns that don’t have strong external verifiers

@seanhn Nice! An exploit is just a proof of a vulnerability indeed

What mathematicians call "literature review" should be familiar to you as "vulnerability research". Or, put another way: erdosproblems.com is currently the best benchmark for LLM capabilities in finding 0days.