Google Launches Gemini for Science: Three AI Research Tools in Google Labs
Google has launched Gemini for Science in Google Labs, comprising three agentic tools for scientific research. Literature Insights (built with NotebookLM) searches millions of papers and synthesizes findings into tables, slide decks, and audio overviews. Hypothesis Generation (built with Co-Scientist) runs a multi-agent "idea tournament" that generates, debates, and rigorously evaluates research hypotheses. Computational Discovery (built with AlphaEvolve) tests thousands of code variations in parallel, enabling rapid modeling approaches for fields such as epidemiology. Google tested the suite on a rare AK2 genetic disease, achieving novel structural insights faster than conventional methods.
Why It Matters
Gemini for Science bundles three distinct AI research paradigms — RAG-powered literature synthesis, multi-agent hypothesis evaluation, and evolutionary code search — into a single accessible suite. This is Google's most ambitious single push into AI-assisted scientific discovery and marks a transition from analysis tools to active AI collaborators in the research process.