Skip to Content
Artificial intelligence

AI can invent new ways to create complex molecules

March 29, 2018

A new algorithm trained on millions of reactions can tell scientists the steps required to build organic compounds.

How it works: Researchers at the University of Manchester trained a neural network on the 12.4 million single-step organic-chemistry reactions currently known to science. That data set lets the network predict more complicated multi-step chains of reactions that produce increasingly complex molecules, without having any chemistry rules hard-coded into the algorithm.

Double checking: In a double-blind study, the researchers gave 45 chemists two potential synthesis routes for creating nine different molecules—one each from the algorithm and a human scientist. According to Nature, the chemists didn’t prefer one over the other, and eight of the algorithm’s suggestions worked properly when tested in the laboratory.

Why it matters: Currently, chemists create these kinds of chemical recipes with painstaking research into existing processes (and a little intuition). That can take days. This new tool could speed up basic science in fields like drug discovery, freeing up chemists to do other work.

Deep Dive

Artificial intelligence

Why Meta’s latest large language model survived only three days online

Galactica was supposed to help scientists. Instead, it mindlessly spat out biased and incorrect nonsense.

DeepMind’s game-playing AI has beaten a 50-year-old record in computer science

The new version of AlphaZero discovered a faster way to do matrix multiplication, a core problem in computing that affects thousands of everyday computer tasks.

A bot that watched 70,000 hours of Minecraft could unlock AI’s next big thing

Online videos are a vast and untapped source of training data—and OpenAI says it has a new way to use it.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.