# AI agents for science model the iterative research process, not just apply powerful techniques

> This piece makes a sharp distinction between two ways AI can do science. The famous example, AlphaFold, worked because protein folding came with an enormous ready-made dataset, roughly 170,000 validated structures built from an estimated $21 billion of prior experimental work. That approach is powerful but narrow, and most scientific questions do not arrive with such a corpus waiting. The argument is that AI agents acting as generalist researchers could reach further. Instead of applying one massive technique to one data-rich problem, an agent can replicate the iterative, contingent process a human researcher follows, forming a hypothesis, testing it, and adjusting, all digitally. That makes it potentially useful in fields that lack comprehensive experimental datasets. Drawing on voices like Eric Schmidt's Schmidt Sciences and its AI-for-science lead Suhas Mahesh, the article frames agents less as oracles and more as tireless collaborators. The promise is real, but so is the caveat, an agent that reasons like a scientist also inherits a scientist's capacity to be confidently wrong.

_Section: [Interesting AI Articles](https://www.wortins.com/articles) · Source: MIT Technology Review · Published Wednesday, August 12, 2026_

## Wortins' read

This piece makes a sharp distinction between two ways AI can do science. The famous example, AlphaFold, worked because protein folding came with an enormous ready-made dataset, roughly 170,000 validated structures built from an estimated $21 billion of prior experimental work. That approach is powerful but narrow, and most scientific questions do not arrive with such a corpus waiting. The argument is that AI agents acting as generalist researchers could reach further. Instead of applying one massive technique to one data-rich problem, an agent can replicate the iterative, contingent process a human researcher follows, forming a hypothesis, testing it, and adjusting, all digitally. That makes it potentially useful in fields that lack comprehensive experimental datasets. Drawing on voices like Eric Schmidt's Schmidt Sciences and its AI-for-science lead Suhas Mahesh, the article frames agents less as oracles and more as tireless collaborators. The promise is real, but so is the caveat, an agent that reasons like a scientist also inherits a scientist's capacity to be confidently wrong.

## Source

[Read the full story at MIT Technology Review](https://www.technologyreview.com/2026/08/10/1141526/the-download-ai-agents-science-censorship-industrial-complex/)

## Related coverage

- [Anthropic Signs $45 Billion Compute Deal with British Infrastructure Firm Nscale](https://www.wortins.com/story/anthropic-signs-45-billion-compute-deal-with-british-infrast-9d89144c) — [TechCrunch](https://techcrunch.com/2026/08/26/anthropic-continues-compute-gobbling-streak-in-45-billion-deal-with-nscale/)
- [DARPA and US Air Force Successfully Fly F-16 Fighter Jet Under Full AI Control](https://www.wortins.com/story/darpa-and-us-air-force-successfully-fly-f-16-fighter-jet-und-b659da37) — [DARPA](https://www.darpa.mil/news/2026/darpa-us-air-force-fly-ai-controlled-f-16)
- [Checking In on AI and the Big Five](https://www.wortins.com/story/checking-in-on-ai-and-the-big-five-8f54332c) — [Stratechery](https://stratechery.com/2025/checking-in-on-ai-and-the-big-five/)
- [OpenAI Announces Astra Model Solves 10 Previously Unsolved Math Problems](https://www.wortins.com/story/openai-announces-astra-model-solves-10-previously-unsolved-m-cc3a3dfc) — [OpenAI](https://openai.com/index/introducing-astra/)
- [Generalist AI's GEN-1.5 Learns New Robot Tasks From Single 3-12 Second Demo](https://www.wortins.com/story/generalist-ai-s-gen-1-5-learns-new-robot-tasks-from-single-3-dd77da49) — [Generalist AI](https://generalistai.com/blog/gen-1.5)
- [Hugging Face Summer 2026: Open-Weight Models Now Majority of Production Usage](https://www.wortins.com/story/hugging-face-summer-2026-open-weight-models-now-majority-of--ca0dd8b8) — [Hugging Face](https://huggingface.co/blog/state-of-open-models-summer-2026)

---

_Curated and written by [Wortins](https://www.wortins.com) — The daily AI briefing. Every story links to its original source; the "Wortins read" on each is our own original analysis. [About Wortins & our editorial approach](https://www.wortins.com/about)._