Arnaud Benard
@arnaudai
ceo & co-founder Galileo AI (acquired by Google)
I’m excited to share that: 1. Galileo AI has been acquired by @Google. 2. We launched today the next generation of our product, powered by Gemini: Stitch More on this big news below.
We’re spinning up a few new roles on the Jules team. If you’re into AI agents and dev tools, DM me. Looking for: – Front-end / full-stack (React ecosystem) – macOS / iOS – Product designer Must have experience building for developers.
Very cool use of @stitchbygoogle
I came across the @SuperteamNG bounty by @airbillspay and immediately loved the product. Paying for airtime, data and other services with stablecoins, and built on solana $SOL ecosystem is noteworthy. I worked together with @ThePromiseOkeke to deliver a complete redesign and full…
The biggest disruption in no-code isn’t AI itself. It’s the shift from paying to host your app to paying to generate it. Hosting is now table stakes. Pre-AI tools will start to fall behind if they don’t adapt quickly.
New 🚢 for Stitch, pretty excited about the improvements to the Experimental mode.
Stitch release alert ✨ • Experimental Mode limit increased to 100 generations/month • Faster workflow with suggested responses in the chat thread • Latest 2.5 Pro model in Experimental mode • Translation support for 30+ countries • Enhanced generation consistency and…
AI talent, low income taxes, and no capital gains tax. 🇨🇭is a strong contender to become Europe’s AI hub.
It's all change in AI in 🇨🇭. When I moved here from the UK 11 years ago, for the quiet life, I didn't imagine there would be this much action! Exciting times.
We have an ambitious vision for Stitch, we're just getting started. We've heard all your feedback on the product. Lots of improvements coming up soon on design quality, variety and iteration capabilities. Maybe even a new model 👀 Code is our underlying artifact, we will make…
Today's deep-dive is on @stitchbygoogle. Stitch was born out of a desire to make app building more accessible to everyone. A product that serves both engineers and designers, Stitch can generate both UI designs and frontend code and export them to Figma or HTML respectively.
The reception for @stitchbygoogle from the community has been amazing! Huge thank you to everyone who has tried it out. We're building this for you, so we'd love to know: what's your top feature request? Let us know below! 👇
Meet Stitch by @GoogleLabs, the easiest and fastest product to generate great designs and UIs. 🧵 stitch.withgoogle.com
Big news: Galileo AI has been acquired by Google! When @arnaudai and I started Galileo in 2022, designing interfaces with text prompts felt like a wild dream—ambitious and ahead of its time. And we brought that vision to life. To everyone who believed in us, especially the…
I’m excited to share that: 1. Galileo AI has been acquired by @Google. 2. We launched today the next generation of our product, powered by Gemini: Stitch More on this big news below.
We're experiencing a lot of traffic with @stitchbygoogle right now: some generations will fail. We're on it.
Mario Draghi's new report on EU competitiveness doesn't mince words. "Across different metrics, a wide gap in GDP has opened up between the EU and the US, driven mainly by a more pronounced slowdown in productivity growth in Europe. Europe’s households have paid the price in…
Dear Mario Draghi, a year ago, I asked you to prepare a report on the future of Europe’s competitiveness. No one was better placed than you to take up this challenge. Now, we are eager to listen to your views ↓
We're very early in inference scaling, but it seems like one of the most interesting research areas in LLMs. We went from investing a lot in pre-training to post-training, so it's only logical that the next step is more compute at inference time.
OpenAI’s Strawberry, LM self-talk, inference scaling laws, and spending more on inference Whether or not scaling works, we should spend more on inference. interconnects.ai/p/openai-straw…
It's official: an open model now matches the performance of GPT-4.
Our Llama 3.1 405B is now openly available! After a year of dedicated effort, from project planning to launch reviews, we are thrilled to open-source the Llama 3 herd of models and share our findings through the paper: 🔹Llama 3.1 405B, continuously trained with a 128K context…
Decided to start a new blog series about model architectures in the era of LLMs. 😀 Here's part 1 on broader architectures like Transformer Encoders/Encoder-Decoders, PrefixLM and denoising objectives. 😄 A frequently asked question: "The people who worked on language and NLP…