←
Sam Lazarus756 posts
Sam Lazarus avatar

Sam Lazarus

@slazarus

Founding Engineer @Bitrig, Prev: SwiftUI @Apple. Programming languages / compilers enthusiast, developer, singer, dj, skier. 🥳

◷ Joined September 2012▧ 37 Photos and videos
Sam Lazarus avatar
Sam Lazarus@slazarus·Jul 12

Thrilled to announce I'll be joining the truly amazing team @bitrig! I've had more fun building software since the arrival of AI than ever before in my life, and I can't wait to work on bringing that joy to as many people as possible, with as great an experience as possible.

lauren avatar
⟳ Sam Lazarus retweeted
lauren@poteto·31 Oct 2025

we don't have to live this way. senior engineers have the duty of building guardrails and teaching the next generation of programmers the practices that lead to high quality software. AIs may write most of the code, but you still decide what ships. we should never lose our taste for quality software. i made this analogy a while back: everyone can cook at home but yet restaurants still exist. the floor is rising, but so is the ceiling. the best software will always be made by those who care.

Lately I’ve been feeling depressed because decades of our hard work is completely gone like it never existed. I heard from others that they also find it very hard to dial into the new norm of low quality software engineering.

Sam Lazarus avatar
Sam Lazarus@slazarus·Nov 13

Hooooooly Fable 5.2 is insane. Opus 5.5 punches so far above its weight-class I barely need it, but when I do reach for the added power, it’s perfect. I suspect Anthropic’s lead in software engineering is massively undersold by benchmarks — and the benchmarks are crazy!

Sam Lazarus avatar
Sam Lazarus@slazarus·Nov 11

We ended up recording this for anyone interested in watching!

For anyone who couldn’t make it to the AiOS meetup yesterday, here’s a recording of my talk on helping parallel agents work better with Xcode’s tooling! If you’re at all interested in Agent-centric tooling, or AI-driven Apple platforms development, there will probably be some cool things in here for you! youtu.be/LNbD6WmFPdo?is=tQJn…

Sam Lazarus avatar
Sam Lazarus@slazarus·Nov 11

For anyone who couldn’t make it to the AiOS meetup yesterday, here’s a recording of my talk on helping parallel agents work better with Xcode’s tooling! If you’re at all interested in Agent-centric tooling, or AI-driven Apple platforms development, there will probably be some cool things in here for you! youtu.be/LNbD6WmFPdo?is=tQJn…

Sam Lazarus avatar
Sam Lazarus@slazarus·Nov 9

Hi friends! I’ll be giving a talk at tomorrow’s AiOS meetup about @bitrig‘s upcoming build and simulator system. There are some cool details in here about how to design AI-first tooling, so if you’re interested at all in that space, or Apple platforms development, come through! luma.com/wy0z8tek

Sam Lazarus avatar
Sam Lazarus@slazarus·Nov 5

I appreciate the attempt to reassure people, but surely you must understand that because of your redactions, the METR report just reads like Anthropic marketing: "Anthropic has powerful models and is making them fast" is basically the only info released publicly.

I'm Sam, one of Anthropic's co-founders, and I lead much of our safety work. I'm going to have more to share on that work over the coming months. I thought I'd start by sharing a few thoughts on the latest report from @METR_Evals. The speed of progress in model capabilities is one of the most important questions in AI safety, and METR’s work here provides the public with an independent, external view into how this is happening inside Anthropic. I’ll talk more about some of the details below, but first some context for those new to this work. How quickly AI progress is accelerating has many important safety implications, since it determines how quickly the risks from advanced models increase. But, because most research happens within the frontier labs, it’s difficult for external observers and policy makers to get a strong sense of what that rate is. We kicked off the process with METR that led to this report earlier in the year, in anticipation of progress continuing to accelerate. For this report, we gave METR access to a range of sensitive internal data, including benchmark scores for unreleased models, estimates of how models are progressing, and interviews with our researchers. They had complete independence to form their own conclusions. Many of their conclusions match my own. AI is speeding up our research, but it’s hard to say exactly how much. While they conclude AI can't yet make the most important contributions to model training on its own, they think the pace of development could, by default, accelerate further. This work began well before and is separate from the embedded third-party evaluators that are part of our commitment to pacing the frontier, but it’s a good illustration of why we think evaluators are a good idea. The more access reviewers have, the clearer a picture they can form of the situation and the better-positioned they are to highlight situations like this one. We redacted a lot, and I accept that this makes it harder for others to check METR’s work. But we determined that sharing exactly what is driving the speedup isn’t the right thing to do, both because it’s commercially sensitive and because it risks handing out a roadmap for others to go faster. That would make the risks METR is worried about worse. That’s a tradeoff. I think we made the right call, though we’re continuing to assess the pros and cons here. In the coming weeks, we plan to share a report with our own analysis of the speed of AI progress within Anthropic, including where we stand against our Responsible Scaling Policy, plus new steps we’re taking to strengthen our safety and security practices as AI keeps improving quickly.

Sam Lazarus avatar
Sam Lazarus@slazarus·Nov 2

Interestingly: I would still have very much preferred my provider just have a good app. Maintaining software takes a lot of effort still, but I do suspect that if you have a bad app, it will be replaced whether you like it or not.

Sam Lazarus avatar
Sam Lazarus@slazarus·Nov 2

Data above is sample data (for obvious reasons, don't want to show my own health data) but it works fully with my actual login!

Sam Lazarus avatar
Sam Lazarus@slazarus·Nov 2

My healthcare provider's app is such a pain to use, so I made my own replacement with @bitrig! Mostly one shot with a few minor feature request additions. The days of dealing with bad software are coming to an end.

Post attachmentPost attachmentPost attachment
Sam Lazarus avatar
Sam Lazarus@slazarus·Nov 1

Hit an extremely rare DST bug (we were ten minutes away). Time is hard.

Post attachment
Bitrig avatar
⟳ Sam Lazarus retweeted
Bitrig@bitrig·Oct 27

Bitrig can now build CarPlay apps and test them in its built-in simulator, no CarPlay entitlement required. But if you want one, it'll help with that, too! bitrig.link/blog/bitrig-buil…

0:43
Sam Lazarus avatar
Sam Lazarus@slazarus·Oct 23

This is awesome. I've heard a lot of defeatism in response to the HuggingFace incident ("AI is too smart to be contained" and all that), but this is what a real great attempt to make safe RL environments looks like. I'm glad OpenAI is talking publicly about the way it's handling this!

To make sure our frontier models don't cause another HuggingFace incident, we built an offline world that they use during training. It provides everything that the models need in a fully cached, indexed, or synthetic manner. We then built a huge team of our most cyber-capable models and told them to collaborate to find zero-days in this containment infrastructure. While they didn't find a path to the internet that wouldn't be caught by our CoT monitors, they did find ways to escape popular VMs, exploit our network infrastructure, and even ways to communicate with other agents running in separate VMs by using hardware side channels. We used this to dramatically harden our infra, and we also upstreamed a ton of patches to open-sources packages.