Human Brain Cells Play Doom: A Leap for Biological Computing

In a groundbreaking experiment, a biological computer powered by approximately 200,000 living human neurons has successfully played the classic first-person shooter Doom. This achievement, spearheaded by Australian biotech company Cortical Labs, marks a significant step forward in the field of biological computing, demonstrating the potential of using living neural cultures for complex tasks.

From Pong to Doom: The Evolution of Biological Computing

Cortical Labs first gained attention in 2022 when they demonstrated a culture of lab-grown human brain cells playing Pong The Verge. The latest experiment builds upon this success, showcasing a more sophisticated application of their CL1 platform. The CL1 houses live neurons, kept viable by a system providing continuous nutrients and oxygen, and utilizes micro-scale electrodes to both stimulate and read the neurons’ responses, known as spikes.

How it Works: Bridging the Digital and Biological

The Doom experiment involved translating the game’s video feed into patterns of electrical stimulation. For example, a demon appearing on the left side of the screen would activate a specific cluster of electrodes. The neurons’ responses were then decoded as control inputs – move right, turn, shoot – creating a closed-loop system of perception, processing, and action Abit. Independent developer Sean Cole leveraged Cortical Labs’ open API to connect Doom to the neural culture in under a week, highlighting the increasing accessibility of the technology.

Beginner Level Performance, Significant Potential

Currently, the neural colony plays Doom at a beginner level – chaotic and inefficient, but self-directed. Researchers have observed signs of adaptive behavior and early-stage learning Abit. Whereas far from matching the skill of human players, the system’s performance exceeds that of random input. Cortical Labs notes that meaningful progress will require more sophisticated feedback algorithms and improved methods for encoding and decoding neural activity Recent Scientist.

The CL1 Platform and the Future of Neurocomputing

The CL1 platform represents a significant advancement in biological computing. The experiment demonstrates that a digital-to-biological interface is viable, opening the path to biological neurocomputing. A single desktop unit containing living neurons costs $35,000 Abit. Cortical Labs believes this technology could eventually be applied to real-world applications, such as controlling robotic arms New Scientist.

Key Takeaways

• A biological computer powered by 200,000 human neurons has played Doom.
• The experiment demonstrates the feasibility of a digital-to-biological interface.
• The CL1 platform offers a new approach to neurocomputing with potential applications beyond gaming.
• The technology is rapidly evolving, with accessibility increasing through open APIs.

This experiment signifies a pivotal moment in the development of biological computing, suggesting a future where living neurons could play a crucial role in processing information and controlling complex systems.