A Dish of Brain Cells Figured Out How to Play Pong in 5 Minutes : ScienceAlert

How many brain cells does it take to play a video game?

Not really. This is no joke and there is no punch line. Instead, there’s a real actual answer, thanks to a neural network system called DishBrain.

If this game is pong, the number of brain cells is about 800,000.

While their slow, one-sided digital table tennis strategy won’t see them win esports championships anytime soon, it does reflect the potential of fusing living tissue with silicon technology.

frameborder=”0″ allow=”accelerometer; autoplay; write clipboard; encrypted media; gyroscope; picture-in-picture” allow full screen>

This is the first synthetic biological intelligence experiment to show that neurons can adapt their activity to perform a specific task – and, given feedback, learn to perform that task better. It’s pretty amazing stuff, with potential applications in computing, as well as the study of all sorts of brain issues, from the effect of drugs and medication on brain activity to the development of intelligence.

“We have shown that we can interact with living biological neurons in a way that forces them to change their activity, resulting in something resembling intelligence,” says neuroscientist Brett Kagan of biotech startup Cortical Labs in Australia.

DishBrain is an intoxicating mix of neurons extracted from embryonic mice and human neurons grown from stem cells. These cells were grown on arrays of microelectrodes that could be activated to stimulate the neurons to provide sensory input.

In a game of pong, microelectrodes on either side of the bowl indicated whether the ball was to the left or right of the racquet, while the frequency of the signals indicated the ball’s distance.

With just this setup, DishBrain is able to move the racquet to hit the ball, but overall it performs pretty poorly. In order to play the game well, the neurons need feedback.

The team developed software to provide criticism via electrodes when DishBrain missed the ball. This allowed the system to improve when playing pong, with learning observed by the researchers in just five minutes.

“The beautiful and groundbreaking aspect of this work lies in endowing the neurons with sensations – the feedback – and above all the ability to act on their world,” says theoretical neuroscientist Karl Friston of University College London in the UK.

“Remarkably, cultures have learned to make their world more predictable by reacting to it. This is remarkable because you cannot teach this type of self-organization; simply because, unlike a pet, these mini-brains have no sense of reward and punishment. “

A few years ago, Friston developed a theory called the free energy principle, which proposes that all biological systems behave in a way that narrows the gap between what is expected and what is experienced — in other words, to make the world more predictable.

By adjusting its actions to make the world more predictable, Friston says, DishBrain is simply doing what biology does best.

“We chose Pong for its simplicity and familiarity, but it was also one of the first games to use machine learning, so we wanted to acknowledge that,” says Kagan.

“An unpredictable stimulus was being applied to the cells, and the system as a whole would reorganize its activity to better play the game and minimize a random response. You can also think of simply playing the game, hitting the ball, and receiving predictable stimulation that inherently creates more predictable environments.”

This offers some really fascinating possibilities, especially in artificial intelligence and computing. The human brain, which contains around 80 to 100 billion neurons, is far more powerful than any computer, and our best computers struggle to replicate it. Our best effort to date required 82,944 processors, a petabyte of memory, and 40 minutes to replicate just one second of the activity of one percent of the human brain.

If the architecture is closer to that of an actual brain—perhaps even a synthetic biological system like the one developed by Kagan and colleagues—that goal might not be so far off.

But there are other, perhaps more immediate, implications.

For example, DishBrain could help chemists understand the effects of different drugs on the brain down to the cellular level. It could even one day help tailor drugs to a patient’s specific biology by using neurons grown from stem cells reconstructed from that patient’s skin.

“The translational potential of this work is really exciting: it means we don’t have to worry about creating ‘digital twins’ to test therapeutic interventions,” says Friston. “We now have essentially the ultimate biomimetic ‘sandbox’ in which to test the effects of drugs and genetic variants – a sandbox made up of the exact same (neural) computational elements found in your brain and mine.”

For now, the next step is to figure out how DishBrain’s ability to play Pong is affected by drugs and alcohol. “We’re trying to create a dose-response curve with ethanol — basically getting them ‘drunk’ and seeing if they play the game worse, just like when people drink,” says Kagan.

In other words, a bowl of brain cells rolls into a bar…

The team’s research was published in neuron.