OpenAI's 2,000 AI Agents Analyze Mouse Brain Data in 32 Minutes — a Job That Would Take Humans Months

Visualization of an artificial neural network over a human brain

Artificial intelligence just took a giant leap in biological research. OpenAI, working with PerturbAI and Scripps Research, deployed nearly 2,000 AI agents running on GPT-Rosalind to analyze data from millions of mouse brain cells — completing in 32 minutes what human researchers would need months to a year to do.

The agents worked through a whole-brain CRISPR atlas covering about 8 million mouse brain cells and nearly 2,000 disease-linked genes. Each agent investigated a single gene disruption, analyzed how cells responded, searched the scientific literature, and proposed an explanation. The reports were combined into a publicly available 'Reasoning Atlas' — with the software framework released on GitHub and the underlying paper on bioRxiv.

In blind comparisons, human experts preferred the multi-agent outputs 75 percent of the time over a single agent's report. The distributed agents cited more than 700 references, compared to just 11 from one agent. The analysis surfaced new hypotheses, including a possible link between the DCTN1 gene and cholesterol regulation in neurons, relevant to Perry syndrome. But OpenAI is clear: these are leads to test in the lab, not proven discoveries — the real work of validation still lies ahead.

Source: read more

Comments

Popular posts from this blog

Fairbairn's Walk-Off Field Goal Seals Wembley Thriller as Texans Edge Jaguars 34-31

Anthropic IPO: Public S-1 Filing Expected Today Ahead of $2 Trillion Listing

Netflix Beats Earnings but Slashes Ad Forecast — Stock Tumbles After Hours