just me, developer infra for ai stuff @AIatMeta; former @pytorch, @docker; all my tweets are my own

California, USA
Joined April 2009
Fable inheriting Astras sloppy seconds.
Fable 5.5 auditing the codebase I let Astra maintain
92
It’s incredible how quickly Muse has become a must need app for me for so many things. It’s my easy app for just about everything.
92
Italian airports are a special brand of hell
87
Can confirm it works out pretty great
Alexandr Wang (@alexandr_wang) says he has more than 200 direct reports at Meta's superintelligence lab, and that it is by design: "Yeah, I think more than 200." "One of the design principles around this was making it very flat. Effectively, all the research that we hired into this group called TBD, they report directly to me." "TBD is like literally anti-bureaucracy." "So I don't need to be micromanaging you to do brilliant work. You're gonna be able to do that on your own." "We have this kind of like pod structure where we have various pods and there's various technical leads of the pods who tie break on the technical decisions." "From a quote unquote people management perspective or from a bureaucracy perspective, we're just anti-bureaucracy." "I would be lying if I said I was the best manager to more than 200 people." "They don't need managers. They need a great environment." "They just need the room to cook." - Alexandr Wang, Meta Chief AI Officer, in an interview with David Senra published October 11.
1
1
101
I very much miss my portal. Video calls with my parents on their portal were the highlight of my week during Covid.
Meta used to sell "Portal," it was their own version of Apple's "HomeView" product. It wasn't just one product. They had a lineup of four products with similar functionality to Apple's HomeView. They also had great integration with Meta's products, such as Facebook and WhatsApp, and even supported YouTube and Zoom. I had one, and it was quite useful at the time. It failed. Now tell me, why would Apple's HomeView do any better?
4
120
Today we are launching Muse Wonder and Muse Wonder Full, our best models yet, with state-of-the-art performance on personal and professional agentic tasks. They are built for long running, multi step work. Muse Wonder is available today in the Meta Model API and Muse Code, and Muse Wonder Full in Muse Code.
33
53
5
438
19,091
Already see a lot of people discounting Muse Wonder because it wasn’t RL’d in 3d stuff like Astra and Opus. 3d game demos are cool! They don’t represent real code work at all. Wonder is an incredible code model. Don’t sleep on it.
214
80
39
3,902
970,630
MUSE WONDER this is the culmination of over a year of research and infrastructure work in msl since we originally shipped muse spark in april, we have continued to scale our models across pre-training and RL, now reaching the frontier 🏆 see our model's improvements on the artificial analysis index:
216
312
170
4,541
434,410
introducing muse wonder our new frontier model. state-of-the-art at personal agents and professional agents. our largest and most capable model yet—rivals the big bois
521
686
376
10,091
2,752,019
Come and join this great event to learn about the latest technologies in AI infra.
Announcing the first SGLang Summit: November 12–13 at Fort Mason, San Francisco. Join us for two days of talks, discussions, and deep dives into frontier AI infrastructure, silicon, models, and applications. Register: sglang.io/summit
1
10
105
8,572
A lot of people forget that our job was never to write code. Our job is to solve people’s problems.
1
1
98
Walking around FiDi in SF feels like being in a completely different town.
97
it’s ok sarah muse is thinking about you too
123
90
60
2,663
601,715
openai cfo: muse is going to better, sorry, not muse, dots is going... youtube.com/watch?v=jXvxK87F…
14
19
15
383
61,178
Muse helped schedule me an electrician this morning to fix a dimmer switch issue. I didn’t even have to do anything besides give it my task rabbit login.
1
170
I guess the PRD took 6 months to get approved and they still needed to ship for someone’s promo case.
Antigravity 2.0 now features a dedicated planning mode, just like the Antigravity CLI. Type /plan and the agent steps back to think through the task, and conduct extensive research, before generating an implementation plan for your review. The agent will ask for your approval before it begins executing on the task. You can also ask for a plan naturally in your prompt, for a lighter version of a plan.
213
I’ve seen a couple of posts about this so wanted to demystify. Today, every Muse user gets a free computer in the cloud. It's a real computer, and we’ve designed the security architecture of the Muse Secure VM carefully so you and your Muse can do almost anything you could with a computer sitting under your desk while keeping you and the system safe from threats like prompt injection. We wrote about this at length in our security blog post – security.muse.ai. Activity in the “runtime cell”, which you share with your Muse is unfettered, but sensitive actions are all overseen by the Sentinel, which runs outside of that cell. Similarly, all sensitive secrets - like the passwords you enter into Muse’s secure credential storage - are also stored outside the runtime cell. The runtime cell gets its own root filesystem (including a full Ubuntu linux image) separate from the host filesystem where your other more sensitive data lives. Because it is isolated from the sensitive stuff that runs on the same box, this means that we can, and do, offer users full visibility and control over the files in the runtime cell. Just as you can when you install Linux on your home computer, you can poke around and see all the files that make the system work - both debian system files and the binaries and data files that implement the parts of Muse which run in the runtime cell. This was a very deliberate choice - your Muse Secure VM truly is your own computer in the cloud. You can install software in it, write and compile code, use the browser to surf the web: it is your own Linux box that you can operate as you choose with your Muse. Poking around in this computer doesn't give you any privileged access to Meta infrastructure, or to other people's data If I may geek out a little here for a second… As a kid I loved to take things apart to see how they worked. As a teenager I got into computers and soon found myself drawn to C:\WINDOWS\SYSTEM and the system registry, later Slackware’s /dev/, /proc/ etc – I could see how the system was laid out and as I explored what DLL files and .so files actually did, I gradually became able to meld the computer to my own will. We’re really proud to be able to put a real computer in millions of people’s hands with a similar level of transparency. We built a file explorer right into the Library tab of the UI. We want you to be able to see the markdown files Muse writes while it thinks about how to serve you better, and explore the internals of the system if you’d like to. So, when you ask your Muse to show you its entire filesystem, and receive gigabytes of files you’re seeing the full contents of the runtime cell. It’s yours to explore and enjoy! If you’re not a geek like me, or simply want to download the data that you personally have created directly with your Muse, we added a feature for that too in Settings > Data controls > Download your agent data.
255
330
196
3,885
846,527
AI might be melting my brain but at least it can't take away my neovim muscle memory.
1
98
If you can’t shop on Amazon using muse then I guess I won’t shop on Amazon as much. 🤷🏽‍♂️
86
Agents should follow up an auto-mode denial with a question tool asking you to explicitly authorize permission for the thing that got denied. It drives me crazy that my agent is asking me to run commands in a separate terminal. Like bro, that's what I have you for.
4
1
324