Full Breakdown
Human Brain Cells Learn to Play Doom: A Leap for Biological Computing
2/28/2026, 6:03:01 AM
Breakthrough in Biological Computing
Cortical Labs, an Australian company, has achieved a significant milestone in biological computing by programming human brain cells to play the classic video game Doom. This advancement follows their previous success in 2021, where they demonstrated that neuron-powered computer chips could play Pong. The latest development utilizes a clump of over 800,000 living brain cells grown on microelectrode arrays, which can both send and receive electrical signals. The new interface allows for easier programming using Python, enabling independent developer Sean Cole to teach the chips to play Doom in approximately one week.
Performance and Learning Capabilities
While the performance of the neuron-powered chip in Doom does not match that of human players, it surpassed the capabilities of a randomly firing player. Brett Kagan from Cortical Labs noted that the biological chip learned much faster than traditional silicon-based machine learning systems and has the potential to improve with advanced learning algorithms. Kagan emphasized that the biological nature of the chip allows for unique information processing capabilities that silicon cannot replicate.
Implications for Future Applications
Experts believe this breakthrough brings biological computers closer to practical applications, such as controlling robotic arms. Yoshikatsu Hayashi from the University of Reading highlighted that playing Doom represents a simpler version of managing a robotic arm, showcasing the chip's ability to handle complexity, uncertainty, and real-time decision-making. Andrew Adamatzky from the University of the West of England remarked on the significance of this advancement, stating that successfully interacting with a complex game like Doom demonstrates real progress in controlling and training living neural systems.
Challenges and Unanswered Questions
Despite the excitement surrounding this development, researchers acknowledge that many questions remain unanswered. Steve Furber from the University of Manchester pointed out the need for further understanding of how the neurons interpret their environment and execute tasks without traditional sensory input. This gap in knowledge highlights the complexities involved in biological computing and the challenges that lie ahead.
Conclusion
The ability of human brain cells to learn and play Doom marks a notable step forward in the field of biological computing. As researchers continue to explore the capabilities and limitations of these neuron-powered chips, the potential for real-world applications grows, paving the way for innovative technologies that could revolutionize fields such as robotics and artificial intelligence.
