Researchers at Cortical Labs in Australia have achieved a significant milestone in biocomputing: a computer powered by lab-grown human brain cells can now play the classic video game Doom. While not yet a pro-gamer, this represents a major step forward in developing hybrid organic technologies that blend biological and silicon-based systems.
From Pong to First-Person Shooters
The breakthrough builds on previous work with “DishBrain,” an earlier biocomputer using roughly 800,000 human neurons. DishBrain demonstrated the potential for these biological circuits by successfully learning to play Pong in 2021. However, Doom, with its dynamic visuals and real-time demands, posed a far greater challenge.
The key innovation lies in Cortical Labs’ new “CL1,” which they claim is the world’s first deployable biological computer. CL1’s open interface, programmable via Python, allowed independent developer Sean Cole to adapt the biocomputer to interpret visual data from Doom as electrical stimulation patterns for the neurons.
Why This Matters: Beyond Gaming
The ability to run Doom is more than just a tech flex. It demonstrates the biocomputer’s capacity for adaptive, real-time goal-directed learning, a fundamental requirement for more complex applications. Traditional machine learning often requires massive datasets and computational power; this biological approach suggests a potential alternative that could be more efficient in certain tasks.
The long-term implications extend far beyond gaming. Cortical Labs envisions biocomputers powering robotic limbs, running digital programs, or even handling specialized computational tasks that strain conventional silicon-based systems.
The Road Ahead
The current biocomputer still loses frequently in Doom, but performs better than random gameplay. Researchers anticipate rapid improvements as algorithms evolve. The CL1’s speed to reach this level also surpassed typical silicon-based machine learning systems.
“This was a major milestone, because it demonstrated adaptive, real-time goal directed learning,” said Brett Kagan, Cortical Labs Chief Scientific and Chief Operations Officer.
This achievement highlights a growing trend in bio-hybrid computing, where living cells are integrated with artificial systems. The future of this field hinges on further refining the interface between neurons and digital inputs, as well as scaling up the neuron networks for greater processing power.
