It sounds like science fiction: researchers are growing a mixture of electronic components and brain cells in the laboratory. The amazing thing is that their invention has something called intelligence. After just a few minutes, the cells learn to master a simple computer game.

Nerve cells grown on a chip can play the classic video game “Pong”. This is reported by an Australian research team in the specialist magazine “Neuron”. The experiments are proof that even brain cells in a petri dish exhibit inherent intelligence by changing their behavior over time. Next, the scientists want to get the neurons drunk.

The basic principle of the video game “Pong”, a very rough simulation of a tennis match, should be familiar to many people: A pixel ball bounces against the walls within a playing field and has to be played back using a bar that can be moved up and down on one side. In the short clip of a game published by the Australian biotech startup “Cortical Labs”, however, no one is sitting at the controller: the bat – the moving bar – is controlled by a bunch of neurons grown in Petri dishes.

The scientists around Brett Kagan from “Cortical Labs” have christened their invention Dishbrain: The Dishbrains contain around 800,000 living neurons that either came from embryonic mouse brains or were grown from human stem cells from donors. The researchers grew the neurons on a chip with a so-called microelectrode array (MEA), which can both stimulate the cells and measure their activity.

They connected the nerve cells to a computer in such a way that they received feedback on whether the bat hit the ball. “From worms to flies to humans, neurons are the cornerstone of general intelligence,” explains Kagan. “So the question was, can we interact with neurons in a way that we can use this inherent intelligence.”

The result of the “Pong” experiments seems to answer this question in the affirmative: The scientists fired electrons at the left or right side of the board to tell Dishbrain which side the ball is on, while the distance to the racquet was determined by the frequency of the signals was displayed. Through the feedback from the electrodes, Dishbrain learned how to return the ball by making the cells behave as if they were the bat themselves.

According to the researchers, the mini-brains grasped the principle of the game very quickly: it only took them five minutes and they became more and more skilled over time. “We have shown that we can interact with living biological neurons in such a way that they change their activity, resulting in something resembling intelligence,” Kagan summarizes.

The researchers trace the learning of the neurons back to a theory based on the principle of free energy: Put simply, this means that cells at this level try to minimize the unpredictability of their environment. “An unpredictable stimulus was applied to the cells and the system as a whole reorganized its activity to better play the game and minimize a random response,” describes Kagan.

“Remarkably, the cell cultures have learned to make their world more predictable by manipulating them,” adds co-author Karl Friston of University College London, who developed the free energy principle: “This is significant because you don’t see this type of self-organization simply because, unlike a pet, these mini-brains have no sense of reward and punishment.”

The “beautiful and groundbreaking” aspect of this work is to endow the neurons with sensation — the feedback — and most importantly, the ability to act on their world, Friston continued. The principle could possibly be used in the future, for example to test the effects of medication. In the long term, an alternative to animal testing could be created.

Next, Kagan and his team want to test the effect alcohol has on Dishbrain. “We’re trying to do a dose-response curve with ethanol: we basically get them ‘drunk’ and see if they play the game worse – just like people do when they drink,” explains Kagan. This may open up new insights into the processes in the brain.

The authors seem to be well aware that Dishbrain is initially reminiscent of science fiction, but they are nonetheless convinced of the importance of their invention. Right at the beginning of the study, they write: “The use of the computing power of living neurons to create a synthetic biological intelligence (SBI), which was previously only located in the realm of science fiction, could now move within reach of human inventiveness.”