Joined February 2015
Photos and videos
Pinned Tweet
at #MSBuild this year agents learned how to click and play a piano - here's @TheChainsmokers @msdev on Windows! powered by @trycua Cua Driver
4
4
37
3,532
Francesco retweeted
coding is a clear step up from glm-5.1. we've pushed context out to 1M and put serious work into memory. the longer the horizon and the messier the task, the more the model shows its edge.
Jun 13
Intelligence should be open, accessible, and ready to build with, empowering every developer, everywhere. GLM-5.2 is now available to all GLM Coding Plan users, including Lite, Pro, Max, and Team plans. docs.z.ai/devpack/latest-mod… As our new flagship model, GLM-5.2 delivers powerful coding capabilities, usable 1M-context support, and continued strengths in long-horizon tasks. API and Chatbot services will launch next week. The model will also be officially open-sourced next week under the MIT License. The future of AI is open, and it belongs to the people.
50
27
980
68,397
Francesco retweeted
Jun 12
## The Future Codebase After the PR dies, after the Code Review dies, i am seriously wondering if Git needs to die next. roughly 20-40% of code spend is just managing and updating merge conflicts. necessary evil? or legacy "horseless carriage"? cargo culting the past? we don't do line by line merge conflicts when we collaborate with human colleagues - instead we chat, suggest edits, do side comments, and an owner ships it. btw we also don't do CI/CD even collaborating on documents with serious legal/financial implications. maybe the future codebase looks more like a Notion or Linear database than .git objects. It will be less efficient, but more scalable. exactly the Salty Lesson.
72
6
123
28,934
Francesco retweeted
you can fork the code you can't fork the design, the taste open-source is best when built with integrity
you can fork the code. you can't fork good taste. back to building 🛸 - F
1
2
10
1,501
you can fork the code. you can't fork good taste. back to building 🛸 - F
1
3
16
2,486
so.. event canceled?
You waited and Claude Fable 5 is finally here. It's Mythos-class and the most capable model we've ever GA'd, and we're inviting you to be amongst the first to build with it. We believe powerful AI can compress decades of progress into years - disease eradication, massive GDP acceleration, life extension - but AI doesn't do any of that on its own. People do. Edison once said "genius is 1% inspiration and 99% perspiration" but that ratio is inverting. The perspiration is getting automated. Humans have ideas. Agents implement, test, and evaluate them an order of magnitude faster than before. It's becoming clear that much of what advances the frontier is automatable and progress is mostly a function of tools and resources: -how fast you can run experiments -how many at once -how quickly you get answers Every important technology became important because someone picked it up early and aimed it at a problem that mattered. This Saturday, June 13th, with our friends at @cerebral_valley in SF, we're giving you a day to do exactly that. Pick up the problem that's been nagging at you and do something about it. It will be a day of building and learning with @AnthropicAI co-founder Jared Kaplan, Research PM Jeremy Hadfield, and technical leaders from @Replit, @periodiclabs, and @_hex_tech. With over $150K in Claude credits for winners, this is the day you go after: -the idea you've been sitting on -the broken workflow in your industry -the tool you wish existed The longer and more complex the task, the better. Come build something that matters and we'll see you in person on Saturday 👇
1
8
1,715
enjoy the 3 days of training data
8
814
3 days of god-tier progress. now back to sonnet 3.5
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…
2
14
971
Francesco retweeted
I got cua working with Hermes. Solid combo. x.com/khalidwarsa/status/206…

Replying to @superdoteng
Hermes desktop running MiniMax M3 is another solid option. It still needs enhancements like opening new window/tab instead of modifying existing ones or maybe a dedicated profile like Claude, but so far so good.
2
11
587
here’s your reminder to run your agents in the background and keep your main thread free. A simple “run in the background” rule in your prompt or main .md is usually enough
 * if you’re still prompting your agents for human-demanding work
2
6
564
🫶
the OG computer use. fan for 2 yrs 🫡
8
419
Francesco retweeted
We're excited to announce Diana Hu (@sdianahu) as YC's newest Managing Partner. Diana co-founded Escher Reality (YC S17), which was acquired by Niantic, where she shipped AR to the 100M people playing Pokémon GO. Since returning to YC as a partner, she has worked with nearly 230 companies that are now worth a combined $7 billion. Few people have built a startup from zero and also shipped at global scale. Diana has done both. ycombinator.com/blog/diana-h…
71
47
918
401,580
just as you used to plan for technical debt during a sprint, it’s more important than ever to revisit your agent workflows regularly
~Once a month I take an afternoon to completely revisit my Codex setup in light of new features and/or an improved model What I landed on yesterday is really exciting. I feel like I'm actually getting >50% out of Codex what is possible Working on a blog, but it includes: - a home thread that manages a todo list that is always open - a home thread that manages other threads - heartbeats on almost all of them Still WIP, but the vision is to simplify everything: show what matters to uniquely me (because an agent can't do it), show me what my agents are up to, and convert my provided goals/tasks/intentions to the right threads who are responsible for those workstreams In essence, this "home thread" has become my context manager (context: most my work is not coding)
1
8
736
Francesco retweeted
Replying to @_ARahim_ @bcherny
only boomers fix typos in prompts. llms perfectly understand you even if you mistype.
358
195
5,296
160,045
CUA 2.0
Jun 10
1/2 The Cua team is excited to join @aiDotEngineer World’s Fair on July 1 in SF to talk Computer-Use 2.0. CUAs are moving beyond screenshot cursor loops toward window-scoped loops orchestrated by coding agents. We’ll share what changed and what becomes possible next.
17
838
Francesco retweeted
Jun 10
1/2 The Cua team is excited to join @aiDotEngineer World’s Fair on July 1 in SF to talk Computer-Use 2.0. CUAs are moving beyond screenshot cursor loops toward window-scoped loops orchestrated by coding agents. We’ll share what changed and what becomes possible next.
2
6
27
2,920
maybe model preservation should be a thing too
ngl sometimes i miss sonnet 3.5
5
544
Francesco retweeted
I won’t be surprised if @trycua’s cua-driver becomes the standard protocol for computer use. You can run background computer-use agents and literally let any AI agent drive your machine and the developer experience is awesome already for agents and humans (if thats relevant anymore) :D It already speaks MCP, so Claude Code can use it today. Would love to see it land first-class! @bcherny @trq212 @delba_oliveira
Replying to @davis7
Install @trycua and tell it to use the cua-driver skill. Works like Codex's computer use but can be used with any model. Integrated it with my Ghostty based IDE (ghostex.dev) and have been very happy with it.
1
4
15
1,988
users are already finding Fable works far better with cua-driver than their raw computer-use. we've offered to help the CC team wire it up for free. door's still open @bcherny @trq212 @alexalbert__ @_catwu @felixrieseberg @delba_oliveira lets chat - intros to the team welcome 🫶
Replying to @davis7
Install @trycua and tell it to use the cua-driver skill. Works like Codex's computer use but can be used with any model. Integrated it with my Ghostty based IDE (ghostex.dev) and have been very happy with it.
2
6
35
5,538