Google I/O showed how the path for AI-driven science is shifting

2026-05-22 · Source: MIT Technology Review · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, Robotics & Autonomous Systems · Depth: Intermediate, medium

Summary

Google is reorienting its AI for science strategy, shifting emphasis from highly specialized tools like AlphaFold and WeatherNext towards more general-purpose, agentic AI systems. While specialized tools such as WeatherNext, which provided an advance alert for Hurricane Melissa, and AlphaFold, used by over three million researchers, remain valuable, Google's I/O keynote highlighted a move towards LLM-based systems. This includes the new Gemini for Science package, featuring hypothesis-generating AI Co-Scientist and algorithm-optimizing AlphaEvolve. This realignment is evident in personnel changes, like Nobel laureate John Jumper now focusing on AI coding, and industry trends, such as OpenAI's general-purpose model disproving a mathematics conjecture. Google envisions AI evolving from a scientific tool to a collaborative "co-scientist," potentially leading to autonomous research.

Key takeaway

For AI Scientists and Research Directors evaluating future R&D investments, recognize Google's strategic pivot towards agentic AI systems like Gemini for Science. This shift suggests prioritizing general-purpose LLM capabilities for autonomous research over solely developing highly specialized tools. You should explore integrating agentic AI for hypothesis generation and algorithm optimization, preparing for AI as a peer collaborator in scientific discovery.

Key insights

Google is shifting its AI for science focus from specialized tools to agentic, LLM-based systems for autonomous research.

Principles

• Agentic AI systems can make independent research contributions.
• General-purpose AI models show potential in scientific discovery.
• AI's role in science is evolving towards collaboration.

In practice

• Explore Gemini for Science for hypothesis generation.
• Investigate agentic systems for research project execution.
• Consider general-purpose LLMs for mathematical conjectures.

Topics

• AI for Science
• Agentic AI
• Large Language Models
• Google DeepMind
• Gemini for Science
• AlphaFold
• Scientific Discovery

Best for: AI Scientist, Research Scientist, Director of AI/ML

Related on AIssential

• Gemini for Science: AI experiments and tools for a new era of discovery — Gemini for Science leverages AI agents to automate complex research tasks, significantly accelerating scientific discovery and exploration.
• Google’s new Deep Research and Deep Research Max agents can search the web and your private data — Google's new Deep Research agents fuse web and private data, generating visual reports for enterprise workflows.
• OpenAI cracks an 80-year math belief — A general-purpose AI model achieved novel mathematical discovery, signaling advanced AI capabilities for scientific breakthroughs.
• Meet the AI "Co-Scientist" Changing Everything 🤖🧪 #ai — Co-scientist is a fine-tuned Gemini with tools for scientific hypothesis generation, data analysis, and literature summarization.
• Towards Generalist Agents for Accelerating Scientific Discovery171 — LLMs can accelerate scientific discovery by generating and refining hypotheses within agentic, iterative workflows.
• Gemini for Science: AI experiments and tools for a new era of discovery — Gemini for Science uses general AI agents and specialized tools to accelerate scientific discovery by automating complex research tasks.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by MIT Technology Review.