Yann Decoopman retweeted
Claude Opus 5.5 is far better developer than I ever was and than I will ever be. No point in competing, just surrendering and transforming my work into everything that is not engineering anymore.
Yann Decoopman retweeted
Jensen closed his GTC Berlin 2026 keynote with this Hermes Agent demo as the finale.
He introduced it by saying the way to build an agent is to ask an agent to build it for you, so he installed and set up a customized Hermes in one prompt.
Then he used Hermes, running on @nvidia RTX Spark, to create a simulated training environment for the @huggingface Microduck. Hermes planned an obstacle course with Nemotron and had GPT-6 Astra build it using FreeCAD, Omniverse libraries, Blender, Newton, and Isaac Lab. Hermes trained thousands of simulated ducks on that course until the validated policy was ready to run on the real robot in the physical world.
At Nous, we believe open source is a practical tool as much as a philosophical ideal. It gives users the ability to build powerful, complex workflows chaining numerous local+cloud models, devices, and tools themselves instead of waiting for a company to enable them.
Yann Decoopman retweeted
Il est quand même pas en train d'essayer de donner une leçon de ce qu'est possible ou pas à Elon Musk ?
Okay, so it's just compressed air.
Easier.
But crappy specific impulse (less than 50 seconds).
Even assuming that 50% of the mass of the car is the compressed air, that gives you less than 10s of useful thrust (with thrust-to-weight ratio larger than 1.0). Probably more like 5 seconds, actually.
So the car must be an empty, lightweight shell containing big compressed air tanks.
Cute demo, but totally impractical.
But of course, you knew that.
Yann Decoopman retweeted
This camera transition during the F1 race was super cool. It looks like they are using AI to sync two cameras — one in the pits and another on top of the car. You can see it when the footage speeds up at the very end.
Yann Decoopman retweeted
At this point I consider @Cloudflare to be a headless company.
I use it everyday (work + side projects) but I only very rarely use the UI. I used it today to store an API key and it felt odd.
Best tools are the ones I don't have to look at because they just work.
Yann Decoopman retweeted
Here's my technique for adversarial code review if I'm driving from Codex: "Review this with claude".
And when I'm in Claude: "Review this with codex".
Models know how to to kick off a review using the cli. Know how to take turns. Know how to settle an argument.
No magic.
Yann Decoopman retweeted
Guys, I don't know how to tell you this, but when you build a website with AI, it doesn't need to be updated.
There's no core PHP app serving the website 24/7 that is full of possible vulnerabilities that need to be patched. There's no theme with vulnerabilities that need to be patched. There are no plugins with vulnerabilities that need to be patched.
Even if you have some kind of underlying CMS, if you do it right, there's no need to worry about constantly updating anything.
The website being served is just HTML, CSS, and JavaScript.
I'm telling you guys, you just don't know how much of a pain in the ass WordPress is until you stop using it. We've all been desensitized to what is required to maintain a WordPress website.
Sometimes you don't know how heavy your backpack is until you take it off.
Yann Decoopman retweeted
I am surprised that people are surprised this is how I prompt Claude.
Talk to Claude the way you would a coworker. There's no secret to prompting. There's no need to be overly scaffolded or prescriptive for most tasks -- give Claude a goal, and it will figure it out.
Back in the Sonnet 3.5 days, your prompt mattered a lot. Nowadays, it's much more important to communicate to the model:
1. What you want it to do
2. How much effort you want it to spend
3. How it should verify that it did the right thing
Yann Decoopman retweeted
Also for the really eagle eyed of you, we now have a binding to query all the data! 🎉
Point your agent at this to build some real fun stuff on workers developers.cloudflare.com/wo…
Yann Decoopman retweeted
Mods are absolutely insane. You can now customize Claude to work and look the way you want by just prompting it.
Each person works differently, so there's no reason why everyone should have an identical Claude experience. Make Claude your own, and share mods as plugins so others can try your mods too.
Yann Decoopman retweeted
You can now mod Claude Code:
- Change how it behaves
- Customize the UI
- Swap in your own features
Write one with a few lines of TypeScript, or have Claude build it for you. Mods ship inside plugins, so you install them with /plugin in the CLI or desktop app.
A few examples: