A clump of human brain cells can play the classic computer game Doom. While its performance is not up to par with humans, experts say it brings biological computers a step closer to useful real-world applications, like controlling robot arms.
In 2021, the Australian company Cortical Labs used its neuron-powered computer chips to play 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 paddles on either side of the screen.
Now, Cortical Labs has developed an interface that makes it easier to program these chips using the popular programming language Python. An independent developer, Sean Cole, then used Python to teach the chips to play Doom, which he did in around a week.
"Unlike the Pong 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 with biology," says Brett Kagan of Cortical Labs. "It’s this accessibility and this flexibility that makes it truly exciting."
The neuronal computer chip, which used about a quarter as many neurons as the Pong demonstration, played Doom better than a randomly firing player, but far below the performance of the best human players. However, it learnt much faster than traditional, silicon-based machine learning systems and should be able to improve its performance with newer learning algorithms, says Kagan.
However, it's not useful to compare the chips with human brains, he says. "Yes, it's alive, and yes, it’s biological, but really what it is being used as is a material that can process information in very special ways that we can’t recreate in silicon."
[...] Even so, the jump in capability is exciting, says Yoshikatsu Hayashi at 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 which Hayashi and his colleagues are attempting with a similar computer made from jelly-like hydrogel. "[Playing Doom] is like a simpler version of controlling a whole arm," says Hayashi.
"What's exciting here is not just that a biological system can play Doom, but that it can cope with complexity, uncertainty, and real-time decision-making," says Adamatzky. "That's much closer to the kinds of challenges future biological or hybrid computers will need to handle."
(Score: 3, Touché) by looorg on Sunday March 01, @11:27AM (1 child)
Clearly defective. My brain cells learned to play DOOM in a few minutes at most.
(Score: 3, Insightful) by lars_stefan_axelsson on Monday March 02, @10:20AM
No. It actually took you several years to learn how to understand how your eyes 2D representation of the outside world mapped onto a 3D reality, and how to navigate and interact with the real world.
If these cells could learn to do that just after "birth", as it were, then they were so much faster than you that there's no real comparison.
Stefan Axelsson
(Score: 5, Interesting) by KritonK on Sunday March 01, @11:40AM (4 children)
The distance between navigating a virtual player in virtual corridors shooting virtual monsters, and navigating a vehicle equipped with a gun and shooting people in a city seems very small to non-existent. If that chip can do the former, and perform "better than a randomly firing player", it can also do the latter. Just give it a joystick control of the vehicle and a low-res front camera view.
(Score: 0) by Anonymous Coward on Sunday March 01, @03:44PM
That's how this was used in Venezuela. Except that it was remote control.
(Score: 3, Touché) by bussdriver on Sunday March 01, @04:42PM
And the elder said the terminators were made to serve man but retained the training program used to evolve them which someday was activated and the end of the world began. Something called Doom... a game for children that some nostalgic scientist picked to raise his AI on. I think this unrealistic hallucination likely planted by AI; no serious adult would take a kids game called "Doom" of all things and build an AI upon it ! It's clearly a fanatic who wrote a logic bomb, hacked the AI and inserted it to doom the world.
(Score: 2) by Spamalope on Monday March 02, @02:01PM (1 child)
Doom only requires turning left and right. The space is 2D despite looking 3D.
Playing Quake would take several orders of magnitude more understanding. That's closer to being a threat generally.
Doom is excellent marketing given it's limited ability.
I wonder what it'd be useful for at this stage, that machine learning doesn't do better? Or could this do it cheaper?
Or a hybrid; 'trained' via billions spend on Nvidia AI-ish system (possibly a very distilled model) with this adding the last layer of learning?
(Score: 2) by KritonK on Tuesday March 03, @09:41AM
So is a road network. Buildings may be three-dimensional, but roads are two-dimensional, and cars don't have "up" and "down" controls. Sure, you might be able to escape being hunted by a car, by finding some stairs to climb, but the car would still be a considerable threat. Imagine loosing such a vehicle upon a crowd. Not only would it mow down a lot of people, but it would then be able to chase those that managed to escape, and manage to shoot some of them.
(Score: 3, Touché) by SomeGuy on Sunday March 01, @04:37PM
I assume that by "Play Doom" they taught it to play as one of the evil hell monsters, since that is what it will eventually be used for.
I'm gonna need a BFG 9000 and a LOT of ammo.
(Score: 0) by Anonymous Coward on Sunday March 01, @06:16PM
Caffeine and sugar?
(Score: 2) by negrace on Sunday March 01, @07:13PM
Come back when it beats E1M1.
(Score: 2) by krishnoid on Monday March 02, @04:35AM
Shouldn't there be something in the Bible about this? Maybe it's a loophole [wiley.com].