Why We Think
Special thanks to John Schulman for a lot of super valuable feedback and direct edits on this post. Test time compute (Graves et al. 2016, Ling,...
Read moreCategory: General
A Simple Way to Install Talos Linux on Any Machine, with Any Provider
Talos Linux is a specialized operating system designed for running Kubernetes. First and foremost it handles full lifecycle management for Kubernetes control-plane components. On the other...
Read moreAdd JSR packages with pnpm and Yarn
You can now access JSR packages via pnpm and Yarn. Here’s how.
Read moreMusic AI Sandbox, now with new features and broader access
Helping music professionals explore the potential of generative AI
Read moreRequirements change until they don’t
Recently I got a question on formal methods1: how does it help to mathematically model systems when the system requirements are constantly changing? It doesn’t make...
Read moreHow to Install GlassFish Application Server with Nginx Reverse Proxy on Debian 12
GlassFish is a free and open-source implementation of the Java EE Platform developed by Eclipse. In this guide, you will install GlassFish Application Server on Debian...
Read moreIntroducing Gemini 2.5 Flash
Gemini 2.5 Flash is our first fully hybrid reasoning model, giving developers the ability to turn thinking on or off.
Read moreThe Halting Problem is a terrible example of NP-Harder
Short one this time because I have a lot going on this week. In computation complexity, NP is the class of all decision problems (yes/no) where...
Read moreGenerate videos in Gemini and Whisk with Veo 2
Transform text-based prompts into high-resolution eight-second videos in Gemini Advanced and use Whisk Animate to turn images into eight-second animated clips.
Read moreDolphinGemma: How Google AI is helping decode dolphin communication
DolphinGemma, a large language model developed by Google, is helping scientists study how dolphins communicate — and hopefully find out what they’re saying, too.
Read moreDefending against Prompt Injection with Structured Queries (StruQ) and Preference Optimization (SecAlign)
Recent advances in Large Language Models (LLMs) enable exciting LLM-integrated applications. However, as LLMs have improved, so have the attacks against them. Prompt injection attack is...
Read more