How Ai Is Accelerating Scientific Discovery

2026-07-08 · Source: hai.stanford.edu · Field: Science & Research — Life Sciences & Biology, Health & Medical Research, Space Science & Astronomy · Depth: Advanced, medium

Summary

AI is significantly accelerating scientific discovery across diverse fields by overcoming traditional human limitations in time, resources, and information processing. At Stanford, researchers are leveraging AI to push boundaries, such as with Evo 2, a DNA language model introduced in February 2025. Evo 2, trained on 9 trillion base pairs and 40 billion parameters, predicts gene mutations and designs new sequences. Another initiative involves building a human-centered foundation model to simulate human cells for drug discovery and personalized medicine. Furthermore, Biomni, an AI research assistant, has been used by 15,000 scientists to automate 100,000 biomedical workflows, acting as a "co-scientist." While AI excels at generating novel ideas, a study found human experts still produce more practical proposals, highlighting the need for AI to learn viability. In astrophysics, AI is crucial for decoding the universe, especially with the Rubin Observatory's upcoming massive data streams.

Key takeaway

For research scientists and ML engineers developing scientific AI, you should prioritize integrating AI tools as collaborative assistants rather than fully autonomous agents. While AI can generate novel ideas and analyze vast datasets, human expertise is essential for ensuring practicality and viability in experimental design and hypothesis validation. Focus on building systems that augment your analytical capabilities and streamline workflows, like Biomni, to maximize scientific value and accelerate discovery.

Key insights

AI is transforming scientific discovery by augmenting human capabilities in data analysis, hypothesis generation, and experimental design.

Principles

• AI excels at pattern recognition in massive datasets.
• Human oversight remains crucial for practical application.
• Open models democratize access to complex scientific tools.

Method

Evo 2 operates like a chatbot, prompting with DNA strings to autocomplete gene sequences for analysis. Biomni unifies hundreds of tools for automated biomedical workflows.

In practice

• Use DNA language models to predict gene mutations.
• Simulate human cells for drug efficacy testing.
• Employ AI agents for automated literature review.

Topics

• AI in Science
• DNA Language Models
• Biomedical AI Agents
• Scientific Discovery Acceleration
• Virtual Cell Modeling
• Astrophysics Data Analysis

Best for: AI Scientist, Research Scientist, Machine Learning Engineer

Related on AIssential

• AI ‘scientists’ promise to accelerate research — how do they work? — AI agents accelerate research, yet human involvement is essential for sound scientific practice.
• Advancing Scientific Discovery in the Agentic AI Era — Accelerated computing and AI agents are dramatically speeding up scientific discovery beyond human processing limits.
• What I’ve learned from 25 years of automated science, and what the future holds: an interview with Ross King — AI-driven automated science and DNA computing offer pathways to dramatically accelerate scientific discovery and address global challenges.
• Stanford Scientists Build An Ai Lab Partner — Biomni is an AI agent integrating LLMs and tools to accelerate biomedical discovery by automating complex research tasks.
• AI systems devise hypotheses and ways to test them — Multi-agent AI systems can extend human scientific discovery by generating and testing hypotheses.
• Four ways Google Research scientists have been using Empirical Research Assistance — ERA accelerates scientific discovery by generating expert-level empirical software for diverse, complex real-world problems.

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 hai.stanford.edu.