Which ‘AI scientist’ suits your lab? A guide for the perplexed

2026-07-10 · Source: Machine learning : nature.com subject feeds · Field: Science & Research — Life Sciences & Biology, Research Methodology & Innovation · Depth: Intermediate, quick

Summary

Anthropic launched Claude Science in June, joining a growing suite of "AI scientist" tools for researchers, including offerings from OpenAI, Google DeepMind's Co-Scientist, and the open-source Biomni. These agentic AI tools, built on large language models, streamline scientific workflows by breaking down complex requests and integrating external software. For instance, geneticist Euan Ashley used Claude to analyze his genome in 30 minutes, identifying an Alzheimer's risk allele and drug metabolism variants, a process that took 31 scientists nine months in 2010. Researchers are leveraging these tools for tasks like literature reviews, data analysis, figure generation, and manuscript preparation. Examples include Boltz using Claude to design antibodies and immunologist Clare Bryant employing Co-Scientist to generate research hypotheses, potentially accelerating discovery by years.

Key takeaway

For research scientists evaluating new methodologies, you should actively trial various "AI scientist" tools like Claude Science or Co-Scientist. Begin with small, easily verifiable tasks to build confidence in their outputs. This hands-on approach will help you identify which agentic AI best supports specific stages of your research, from generating initial hypotheses to performing complex genomic data analysis, significantly accelerating your discovery process.

Key insights

Agentic AI tools, like Claude Science, dramatically accelerate scientific research tasks from genomic analysis to hypothesis generation.

Principles

• "AI scientists" function as agentic AI, decomposing requests into steps.
• These tools integrate with specialized external software systems.
• Early adoption of AI agents can yield significant time savings in research.

Method

Researchers should trial multiple AI tools to determine suitability for specific tasks and begin with small, verifiable outputs.

In practice

• Apply hypothesis-generating AIs for initial project stages.
• Utilize task-specific AIs for genomic data analysis.
• Design antibodies or identify drug candidates with AI agents.

Topics

• Claude Science
• "AI Scientists"
• Agentic AI
• Genomic Analysis
• Hypothesis Generation
• Drug Discovery

Best for: AI Scientist, Research Scientist, Domain Expert

Related on AIssential

• Claude Science, an AI workbench for scientists, is now available — Claude Science unifies scientific tools and compute, accelerating research with auditable, reproducible AI-driven workflows.
• 4 ways researchers are collaborating with Co-Scientist to solve big problems — Co-Scientist is a multi-agent AI system accelerating scientific hypothesis generation and refinement across life sciences.
• How to build an AI Scientist: first peer-reviewed paper spills the secrets — AI Scientist, an autonomous research tool, successfully generated a peer-reviewed paper, demonstrating AI's capacity for scientific discovery.
• Advancing Scientific Discovery in the Agentic AI Era — Accelerated computing and AI agents are dramatically speeding up scientific discovery beyond human processing limits.
• The Download: Anthropic launches Claude Science, and California’s carbon manure math — AI is rapidly expanding into specialized scientific domains and drug discovery, while climate policies face scrutiny for efficacy.
• 17 AI Models Tested on REAL Scientific Research — Advanced AI models currently demonstrate limited capability in independent scientific discovery, even with curated data and extensive tools.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine learning : nature.com subject feeds.