Living human brain cells successfully play Doom in new biotech experiment

In a laboratory experiment that sounds like something out of science fiction, researchers have managed to get living human brain cells to play the classic video game Doom. The unusual demonstration is part of a growing area of research known as biological computing, where scientists combine living neurons with electronic hardware to create hybrid computing systems. The experiment used roughly 200,000 human neurons that were grown in a laboratory and placed on a specialized microchip. These cells were kept alive in a controlled environment that provided nutrients, stable temperature, and the right chemical conditions for the neurons to function. Electrodes embedded in the chip allowed the neurons to communicate with a computer system. To connect the neurons to the video game, the researchers translated elements of the game into electrical signals. When something happened in Doom, such as an enemy appearing or the player moving through the environment, the system converted that information into patterns of stimulation delivered to the neurons through the electrodes. The neurons responded by producing their own electrical signals, which the computer interpreted as commands like moving, turning, or firing a weapon in the game.

Unlike traditional computer programs, the neurons were not explicitly coded to play the game. Instead, they learned through feedback. When the system made a useful action in the game, such as avoiding danger or reacting correctly to a situation, the neurons received signals that reinforced that pattern of activity. Over time, the neural network began to adjust its responses, gradually improving how it reacted to the game environment. The performance of the neurons is still very basic. Researchers say the gameplay resembles that of a beginner who has never used a computer before. The system can sometimes respond to threats or move through the game world, but it also makes frequent mistakes and struggles with complex situations. Despite these limitations, the ability of living neurons to learn and interact with a digital environment is a significant achievement. This experiment builds on earlier work where researchers trained a similar network of neurons to play the much simpler arcade game Pong. Moving from Pong to Doom represents a major increase in complexity. While Pong involves only a paddle and a ball in a two-dimensional space, Doom requires navigation through a three-dimensional environment, recognizing enemies, and reacting to multiple events happening at once.

Scientists are interested in biological computing because the human brain is incredibly efficient. A brain can perform vast amounts of processing while consuming only about 20 watts of power, far less than most modern computing systems used for artificial intelligence. By combining biological neurons with traditional hardware, researchers hope to develop new types of computing systems that are more adaptable and energy efficient. The research could also help scientists better understand how learning works in neural networks and may contribute to advances in brain-machine interfaces, neuroscience research, and future AI technologies. At the same time, the work raises important ethical questions. As biological computing systems grow larger and more advanced, researchers will need to carefully consider issues surrounding the use of living neural tissue and the possibility—however distant—of more complex forms of neural activity emerging. For now, the experiment remains an early demonstration rather than a practical computing solution. Still, watching a dish of neurons slowly learn to play Doom offers a glimpse into a future where the boundary between biology and technology becomes increasingly blurred.

Spread the love

MORE AMIGA & RETRO NEWS