JUST IN: Florida hospital reveals Palantir software has cut sepsis deaths by more than half since it was installed.

NEW: malware developers added nuclear & biological weapons text to to their spyware. Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner. Cleanest practical example I can think of for why over-indexing on first order safety alignment is risky. When closed (and open) models ship with aggressive refusals, they will be sprinkled with second-order blindspots that attackers will discover...and exploit. We are only in the earliest days of attackers leveraging these features, and it wouldn't surprise me if users systems that need to handle complex cybersecurity issues demand that models be less safety-blunted. In the weeds: @SocketSecurity's post also shows why intention matters in how you design a malware analysis pipeline to avoid prompt manipulation. H/T to colleagues that shared this with me socket.dev/blog/mini-shai-hu…

Ouroboros v0.41.0 Release Run it anywhere, and trust what it ships 1. Pi is now a first-class Ouroboros runtime (Finally!!) 2. Ouroboros has always opened with a single questioner. Now that questioner has a panel. Milestone lateral review is promoted from a non-blocking advisory to a required lightweight subagent pass at exactly the moments hidden assumptions start to bite. 3. The more autonomous the loop, the more its "done" has to mean done. This release makes the verifier's decision typed, auditable, and policy-routed (I believe this looks like real Agent OS) github.com/Q00/ouroboros/rel…

I'm taking a break. TTYL

I really believed a whole generation of developers, who only know open source from npm and pypi, miss how open source actually used to work. When Debian or a Linux distribution ships a dependency they take responsibility of it. If there is a security issue and it’s not fixed by the developer upstream, they fix it for their users. Debian and others basically vendor every thing they distribute. They honor the license and they maintain patches. Most of the stuff that you get from your Linux distribution is basically a (small) fork. The same is true for Apple, Microsoft and others. The open source software they ship, they carry that responsibility. That doesn’t mean that security fixes are not upstreamed, but Apple or Debian or anyone else won’t jump in Twitter to shame a developer into compliance with their ways. They are not dependent on the health of a packaging infrastructure. They own their software including all the things it depends on. I want that thinking back. Because it fundamentally makes people feel more responsibility and it shares the burden of issues. It also does not put so much focus and attention on the one overworked developer who just happened to have too much of the world depend on their library. Remember: they carry a responsibility they never signed up to and they never got compensated for.

My best advice for AI agent security is to not trust anyone who claims they know how to solve AI agent security. There are so many hard problems in this space, and anyone who claims otherwise doesn't know what they're doing.

they should do it as per-diem and you get to keep what you don't use if they actually want to see people use ai effectively.

Satya talking on stage about MCP is so validating when so many of you people said MCP was dead.

Codex just found a “workaround” of not having sudo on my pc…

The touchbar was too early and didn't deserve to die, it deserved an LLM. Static shortcut buttons were the wrong bet. Imagine buttons that rewrite themselves based on whatever you're doing, that's the keyboard that feels right for an AI-native Mac.

One night I quietly gave our AI agent full access to YC's production database. It made the agent 10x more useful. That's what convinced me that trust-by-default is the only way to get the most out of agents.

Ok, confession time: I use agentic coding *all the time* and *every day*. And have been doing so for many months. I am *terrified* of skill deterioration on my side. I see the studies, I can feel it myself. The agents make me much more productive, but I feel I need to force...