@berryxia: Great news for research dogs! If you're in research, you must watch this. At the just-concluded Google I/O, DeepMind CEO Demis Hassabis dropped a chilling () line: "Scientific progress is becoming compu…

X AI KOLs Timeline News

Summary

At Google I/O, DeepMind CEO Demis Hassabis announced that scientific progress is becoming computable, launching the Gemini for Science system to help researchers read papers, write code, and generate hypotheses, enabling science to scale and iterate like software.

Great news for research dogs! If you're in research, you must watch this. At the just-concluded Google I/O, DeepMind CEO Demis Hassabis dropped a chilling () line: "Scientific progress is becoming computable." "Scientific progress is becoming quantifiable." He's not just using AI as a productivity tool; he's directly positioning it as the infrastructure layer of science. A new system, Gemini for Science, is now live: Helping researchers read papers, write code, and quickly generate hypotheses. The real turning point: scientific research itself is starting to scale and iterate like software. Previously, scientific breakthroughs relied on flashes of genius and lengthy trial-and-error experiments. Now AI is making "discovery" engineering, programmable, and accelerated. What AI is truly changing isn't just how we work, but how we do science itself.
Original Article
View Cached Full Text

Cached at: 05/21/26, 08:24 AM

Great news for research folks! If you’re in academia, you must read this.

Just now at the Google I/O conference, DeepMind CEO Demis Hassabis dropped a line that sent a shiver down my spine:

“Scientific progress is becoming computable.”

He’s not just treating AI as a productivity tool — he’s directly defining it as the infrastructure layer of science.

A new system called Gemini for Science has just launched:

Helping researchers read papers, write code, and quickly generate hypotheses.

The real turning point is this: scientific research itself is starting to iterate at scale, just like software.

In the past, science relied on sudden flashes of genius and long experimental trial-and-error. Now AI is making “discovery” itself engineerable, programmable, and accelerated.

What AI is truly going to change isn’t just how we work — it’s how we do science itself.

Similar Articles

@AYi_AInotes: Damn, after watching this DeepMind documentary, I got goosebumps. There were no thrilling speeches or grand launch events—just an ordinary meeting room where a few engineers sat around a table and told Demis: We can now predict all the known 1.7 to 2 billion protein sequences in just one month. Dem…

X AI KOLs Timeline

The article reviews DeepMind's decision-making process behind open-sourcing AlphaFold in 2021, praising Demis Hassabis's leadership style for taking risks and freely making basic research available. It also notes that although this initiative did not generate direct profits, it successfully led to the creation of Isomorphic Labs, valued at $2 billion.

@seclink: https://x.com/seclink/status/2057291288655847533

X AI KOLs Following

Google announced at its developer conference that search has undergone its biggest upgrade in 25 years, officially entering the era of AI agents. It also unveiled nearly thirty updates, including the multimodal model Gemini Omni, the desktop agent application anti-gravity 2.0, the personal AI agent Gemini Spark, and its first AI audio glasses. AI capabilities have shifted from chat-based Q&A to autonomously executing tasks in the background, with a focus on scientific research and new drug development.

AI x Society | I/O 2026 Keynote

YouTube AI Channels

At the I/O 2026 keynote, Google announced the latest AI achievements including Gemini 3.5, Omni, Code Mender, Gemini for Science, WeatherNext, and emphasized that AGI is imminent, and demonstrated AI applications in cybersecurity, scientific simulation, and health.