Research · 2026-10-21

Thousands of AI agents search for new directions in biological data

OpenAI together with PerturbAI and Scripps Research used many AI agents working at the same time to find promising directions for scientific research in data on millions of brain cells.

✨ What happened

OpenAI, the research organization Scripps Research, and the startup PerturbAI showed how AI can be used to analyze large scientific datasets. The data contains information about millions of brain cells.

Instead of a single AI assistant, scientists used thousands of AI agents. The agents worked in parallel, each checking its own part of the data, and the results were gathered together.

📌 Why this is needed

There is so much cell data that it would be extremely difficult for people to review all of it manually. That is why such tasks are increasingly being handed over to AI.

The main goal is to find promising “threads,” that is, directions worth testing further. Such findings are not ready discoveries. Scientists still have to confirm them in laboratories, and only after that can their practical significance be discussed.

🔹 Why this matters for ordinary people

The news shows how AI can speed up scientific research. Fast analysis of large data sets helps scientists move more quickly to testing ideas.

At the same time, this is only the beginning. It usually takes a long time to get from a promising direction to a real result, so immediate changes should not be expected.

Sources: 🔗 OpenAI News