The latest blog posts to take off on Hacker News. 6,328 posts.
Introduction to Helion In modern machine learning, the demand for high-performance computation has led to a proliferation of custom kernels. While these kernels can deliver impressive performance, they are often...
As AI agents change how the Internet is used, they create a challenge for security. We explore how Anonymous Credentials can rate limit agent traffic and block abuse without tracking users or compromising their privacy.
A Response to Yudkowsky and Soares from the Front Lines of AGI Development
A brain dump of all the ways I've been using Claude Code.
The shutdown of the facial recognition cameras at the Václav Havel Airport in Prague came after years of criticism from EDRi member IuRe.
Surprisingly (to me) Claude Code debugged my new ML-DSA implementation faster than I would have, finding the non-obvious low-level issue that was making Verify fail.
OpenAI is now set to become a Public Benefit Corporation, with its investors entitled to uncapped profit shares.
Last night I had a very strange experience: About two thirds of the way through reading a Web page about myself, Tim Bray, I succumbed to boredom and killed the tab. Thus my introduction to Grokipedia . Here are early impressions.
A new avenue for identifying greedy, badly-behaved bots
This blog post is based on the work BaNEL: Exploration Posteriors for Generative Modeling Using Only Negative Rewards. Tackling Very Hard Problems The ultimate aim of machine learning research is to push machines beyond human limits in critical applications, including the next generation of th
Research on the ability of large language models to introspect on their own internal states.
Mozilla.ai is adopting llamafile to advance open, local, privacy-first AI—and we’re inviting the community to help shape its future.
At Augment Code, we run multiple frontier models side by side in production. This gives us a unique vantage point into how different models behave in real…
Composer is our new agent model designed for software engineering intelligence and speed.
In this year’s Octoverse, we uncover how AI, agents, and typed languages are driving the biggest shifts in software development in more than a decade.
It’s very much a “v1” product. You type a prompt to start a new thread, it launches a little container for your agent to work in, and you can keep talking to it. It produces a branch, which you can open a PR for (that’s the only way to see a diff of the changes Claude Code made, for now). Or if you want to keep…
Can LLMs control robots? We answer this by testing how good models are at passing the butter – or more generally, do delivery tasks in a household setting. State of the art models struggle, with the best model scoring 40% at Butter-Bench, compared to 95% for humans.
Microsoft and OpenAI sign a new agreement that strengthens its long-term partnership, expands innovation, and ensures responsible AI progress.
A framework for reasoning about AI-based changes you encounter or create.
AI can write code, but it can’t build real software. Software engineering remains human work because AI can code, not engineer.
Life doesn't freeze when winter comes—if anything, that's when riders need reliable transportation most, when being exposed to the elements becomes less appealing. Today, the Waymo Driver successfully navigates rain, fog, sandstorms, and freezing temperatures. As we expand to more cities across the U.S. and globally…
A collection of posts by Joe Magerramov
An interactive visualization — based on the Acquired Google Podcast Credits to Ben and David (who also appreciated this graph in the Acquired internal Slack!).
A blog about making culture. Since 1999.
A perspective on continual learning, motivating our paper on sparse memory finetuning
When you paste ChatGPT’s response instead of your own feedback, you’re not being helpful. You’re being lazy and creating more work for everyone.
a loss plateau that looked like my mistake turned out to be a PyTorch bug. tracking it down meant peeling back every layer of abstraction, from optimizer internals to GPU kernels.
Earth AI is helping enterprises and cities with everything from environmental monitoring to disaster response.
Today, we’re introducing memory to the Claude app, where Claude remembers you and your team’s projects and preferences, eliminating the need to re-explain context and keeping complex work moving forward.
We now live in a world where ML workflows (pre-training, post training, etc) are heterogeneous, must contend with hardware failures, are increasingly asynchronous and highly dynamic. Traditionally, PyTorch has relied...
