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

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

In 2021, an Australian company Cortical laboratories he used his computer chips powered by neurons to play Pong. The chips consisted of clusters of more than 800,000 living brain cells grown on microelectrode arrays that can both transmit and receive electrical signals. The researchers had to carefully train the chips to control the paddles on both sides of the screen.

Now, Cortical Labs has developed an interface that makes it easy to program these chips using the popular Python programming language. Independent developer, Sean Colethen using Python to teach the chips to play Fate, which he managed in about a week.

“Unlike Pong work we did several years ago, which represented years of painstaking scientific effort, this demonstration was performed in a few days by someone with relatively little prior experience in direct biology,” he says Brett Kagan from Cortical Labs. “It’s that availability and flexibility that makes it really exciting.”

A neural computer chip that used about a quarter as many neurons as Pong demonstration, played Fate better than a random shooter, but well below the performance of the best human players. But it learned much faster than traditional silicon-based machine learning systems and should be able to improve its performance with newer learning algorithms, Kagan says.

However, it’s not useful to compare chips to human brains, he says. “Yes, it’s alive and yes, it’s biological, but it’s actually used as a material that can process information in very special ways that we can’t recreate in silicon.”

“Fate is far more complex than earlier demonstrations, and successfully interacting with it shows real progress in how living neural systems can be controlled and trained,” he says Andrew Adamatzky at the University of the West of England in Bristol, UK.

Steve Furber at the University of Manchester in the UK agrees Fate is significantly higher than gaming Pongbut he says we still don’t understand much 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.

Even so, the jump in capabilities is exciting, he says Yoshikatsu Hayashi at the University of Reading in the 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 from a jelly-like hydrogel. “[Playing Doom] it’s like a simpler version of full arm control,” says Hayashi.

“What’s exciting here is not just that a biological system can play Fate, but that it can deal with complexity, uncertainty and real-time decision-making,” says Adamatzky.

topics: