Saber Darabi

(1) Today we're releasing Muse Spark 1.1 -- a strong agentic and coding model at a very low price. It's available through our new Meta Model API and in Meta AI.

"The old way of doing science is over." Derya Unutmaz, immunologist and professor at The Jackson Laboratory, says science is moving from slow experiments to AI-agent loops. The real shift is labs becoming autonomous systems for biology, chemistry, physics, and drug discovery. Open intelligence turns discovery into a networked process.

you know, this makes me a bit thankful to Americans for making English the universal language. French was OK but France didn't have Anglo demographics so we'd have more fragmentation. I'd have been unable to talk to the Chinese. They're not really obligated to translate stuff.

Because you wouldn’t let it slide… these are rolling out today for our most requested feature: Prompt-Based Revisions: Tweak, tailor, and tune your slides just by prompting the revisions you want PPTX Support: You can now export your Slide Decks (Google Slides coming next!)

What Google just showed with AI Co-Scientist and generative UI isn’t just better interfaces — it’s the beginning of a new format for doing science. For decades, research has been bottlenecked by one limitation: humans explore sequentially, while nature operates in parallel. What we’re seeing now is that AI systems can finally work the way reality works — multi-agent, iterative, hypothesis-driven, and self-correcting. Generative UI is the missing layer. It turns AI from “a text generator” into a dynamic research environment, where simulations, literature synthesis, experimental design, and visualization all update in real time as the agents coordinate. AI Co-Scientist essentially creates: • a planner that explores the hypothesis space • analysts that test alternative pathways • critics that challenge assumptions • and a visual layer that reveals interactions humans normally miss That’s how scientific insight actually emerges — not from a single linear thought, but from structured collaboration between multiple reasoning processes. When this becomes standard, the scientific cycle compresses dramatically: What took weeks becomes hours. What took labs becomes laptops. What took teams becomes tools. And the most interesting part? Science stops being static documents and becomes living, navigable systems — where a question isn’t answered by a paragraph, but by an explorable model. 2025 wasn’t just a year of AI upgrades. It was the year AI started behaving like an actual scientific partner. The next phase isn’t bigger models — it’s better coordination, better representations, and better interfaces that let humans finally see the hidden structure behind the problems they study. The future of research will be co-authored.

Today marks a moment I'll remember for the rest of my life. When we started Manus, few believed that general AI agents could work. We were told it was too early, too ambitious, too hard. But we kept building. Through the doubts, the setbacks, and the countless nights wondering if we were chasing the impossible. We weren't. This isn't just an acquisition. It's validation that the future we've been building toward is real, and it's arriving faster than anyone expected. But this is not the end. The era of AI that doesn't just talk, but acts, creates, and delivers, is only beginning. And now, we get to build it at a scale we never could have imagined. To everyone who believed in us before it was obvious: thank you. The best is yet to come.