Thanks to new technologies like artificial intelligence, scientists are increasingly freed from the constraints of the laboratory. It raises questions about how much humans should outsource to robots.

SCOTT DETROW, HOST:

Researchers may be able to start running experiments without ever setting foot in a lab. New technology, including artificial intelligence, is increasingly allowing them to delegate all kinds of tasks, from repetitive lab work to designing experience. NPR's Katia Riddle visits one company in Boston that is building something called an autonomous lab.

KATIA RIDDLE, BYLINE: The origin story for this company called Ginkgo Bioworks begins with several graduate students from MIT. It was nearly two decades ago they united around a shared idea.

JASON KELLY: We wanted to make biology easier to engineer, right? We believe that programming cells would ultimately be more important than programming computers.

RIDDLE: Jason Kelly was one of those students. Now he's one of the company founders. Things like gene editing or testing new molecules typically take many hours of painstaking labor in the laboratory. These scientists wanted to replace the humans doing these tasks with robots. Not everyone believed this idea would work.

KELLY: We were, you know, living on ramen, buying equipment on eBay, and we could not raise venture capital.

RIDDLE: Then came the AI boom. In 2014, Kelly remembers reading a blog post by someone named Sam Altman, who, of course, went on to found artificial intelligence company OpenAI. Altman saw a way to use AI to automate biotechnology. The two got in touch. Eventually, the Silicon Valley money started flowing. Today, Ginkgo Bioworks labs are in a building overlooking the Boston Harbor. These engineers and researchers believe they are building the science labs of the future.

KELLY: That's done on pipetting robots. I'll show you where we do that.

RIDDLE: Kelly walks around in his automated lab. Robots encased in glass are all working on separate projects. A big screen shows a color-coded schedule of experiments that will be carried out on this day. A circuit that looks like an oversized toy train track runs through the room, delivering materials from one robot to another.

KELLY: And it pulls plates up. That arm picks up the plate and puts it onto the device.

RIDDLE: Instructing robots with AI, the scientists do all kinds of work here, like research on new kinds of drugs...

KELLY: So that one there, with a deep well, that has, usually, actual live cells in it.

RIDDLE: ...Also, things like microbes for better fertilizer, creating proteins that will make snow or ice. Much of this work involves using AI to translate directions to the robots. Recently, the scientists here have been taking it a step further, empowering the AI to be the scientist. Reshma Shetty is another of the founders.

RESHMA SHETTY: The really, really wild moment was the first time I saw a lab notebook entry written by the model.

RIDDLE: Shetty recently worked on a collaboration with OpenAI, the company that created ChatGPT. They gave it a challenge.

SHETTY: So we just took off-the-shelf GPT‑5 that everybody had access to and just said, like, please design a cell-free protein synthesis reaction. And we had no idea if it would even be able to make protein, right?

RIDDLE: Usually, they give the AI agents very specific instructions, kind of like giving it a recipe. In this case, they asked ChatGPT to write its own recipe.

SHETTY: It was summarizing scientific data, analyzing it and then actually generating new hypotheses.

RIDDLE: Shetty says AI has already fundamentally changed the way she practices science.

SHETTY: Normally, I actually rush through designing my experiment because I need to get it done so that I can actually do all the pipetting in the lab and set it all up and then - and get it done before I go home.

RIDDLE: Now she spends more time designing her experiments so that the robot can do them for her overnight. Some people caution these new scientific freedoms come with risk. Drew Endy studies bioengineering at Stanford.

DREW ENDY: The meta risk is that we delegate science to AI and lose our understanding of how to do science, right?

RIDDLE: He warns these kinds of labs could someday become available to people with little to no scientific training.

ENDY: I'm not excited about that.

RIDDLE: Corrupt governments could produce viruses, he says, or create other biosecurity threats. Historically, says Endy, there just haven't been that many people who understand how to manipulate biology. That has made it difficult to weaponize until now.

ENDY: It's been hard for people to really gain control over it. AI could nudge it a little bit more towards concentration of power.

RIDDLE: Ginkgo Bioworks scientist Jason Kelly agrees. For better or worse, he foresees a day when anyone can work with these bots to run an experiment.

KELLY: And then I do think you will have, like, a culture clash coming of what happens when everyday people can ask scientific questions.

RIDDLE: Questions they will work to answer with assistance from science robots and increasingly, robot scientists. Katia Riddle, NPR News, Boston.

(SOUNDBITE OF THE DARE SONG, "GIRLS")

Copyright © 2026 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

Accuracy and availability of NPR transcripts may vary. Transcript text may be revised to correct errors or match updates to audio. Audio on npr.org may be edited after its original broadcast or publication. The authoritative record of NPR’s programming is the audio record.