Human brain cells on a chip learned to play Doom in a week

A clump of human brain cells can play the classic computer game Doom. Although its performance is not on par with humans, experts say it brings biological computers a step closer to useful real-world applications, such as controlling robotic arms.

In 2021, the Australian company Cortical Labs used its neuron-powered computer chips to play games Pong. The chips consisted of clumps of more than 800,000 living brain cells grown on top of microelectrode arrays that can both send and receive electrical signals. Researchers had to carefully train the chips to control the veins on either side of the screen.

Now Cortical Labs has developed an interface that makes it easier to program these chips using the popular Python programming language. An independent developer, Sean Cole, then used Python to teach the chips to play Doom, which he did in about a week.

“As opposed to Pong the work that we did a few years ago, which represented years of painstaking scientific effort, this demonstration has been done in a matter of days by someone who previously had relatively little expertise working directly in biology,” says Brett Kagan of Cortical Labs. “It’s this accessibility and this flexibility that makes it really exciting.”

The neuronal computer chip, which used about a quarter as many neurons as Pong demonstration, played Doom better than a casual shooter, but far below the performance of the best human players. However, it learned much faster than traditional silicon-based machine learning systems and should be able to improve performance with newer learning algorithms, says Kagan.

However, it is not useful to compare the chips with human brains, he says. “Yes, it’s alive, and yes, it’s biological, but really what it’s used for is a material that can process information in very special ways that we can’t replicate in silicon.”

“Doom is much more complex than previous demonstrations, and successful interaction with it highlights real advances in how living neural systems can be controlled and trained, says Andrew Adamatzky at the University of the West of England in Bristol, UK.

Steve Furber of the University of Manchester, UK, agrees Doom is a significant level up from playing Pongbut he says there’s still a lot we don’t understand about how these neurons play the game, such as how the neurons know what’s expected of them or how they can “see” the screen without eyes.

Still, the leap in capabilities is exciting, says Yoshikatsu Hayashi of the University of Reading, UK, and brings us significantly closer to useful real-world applications, such as controlling a robotic arm with biological computers, a task Hayashi and his colleagues are attempting with a similar computer made of jelly-like hydrogel. “(Player Doom) is like a simpler version of controlling an entire arm, says Hayashi.

“What’s exciting here is not just that a biological system can play Doom, but that it can deal with complexity, uncertainty and real-time decision making,” says Adamatzky. “It’s much closer to the kind of challenges that future biological or hybrid computers will need to deal with.”