Human Brain Cells Learn to Play Doom in Cortical Labs Experiment
Decrypt
by Jason NelsonMarch 3, 2026
AI-Generated Deep Dive Summary
Cortical Labs has achieved a groundbreaking milestone by training 200,000 human neurons to play Doom, marking a significant leap in neuromorphic computing. Using electrical stimulation and software that converts gameplay into neural signals, the neurons were able to navigate, react to enemies, and fire weapons, albeit at a beginner level. This experiment builds on the lab's earlier success with Pong and highlights the potential of biological systems to perform complex tasks.
The "Can it run Doom?" challenge has historically served as a benchmark for testing new computing platforms. By porting the game to living neurons, Cortical Labs has pushed this concept into uncharted territory, demonstrating that human brain cells can adapt to process real-time feedback and learn through rewards. The neurons received small rewards for accurate targeting and larger ones for successful eliminations, reinforcing desired behaviors over time.
The researchers employed artificial intelligence to refine how game information was translated into electrical signals, ensuring the neurons could effectively respond to inputs. Though the system doesn’t "understand" Doom like humans do, it exhibits adaptability and learning through trial and error. This approach differs from traditional programming, requiring a new mindset to work with biological systems.
While the project initially served as a public-facing demo, its implications extend into practical applications in medicine and beyond. Despite using human-derived neurons, the system doesn’t replicate human cognition; however, it does showcase neural adaptability outside the brain, offering insights into how cells can respond to engineered environments.
For those interested in crypto and Web3, this breakthrough underscores the intersection of decentralized computing and biological systems. As blockchain continues to evolve, integrating with technologies like neuromorphic computing could unlock new possibilities for processing power and AI development. This experiment not only advances gaming but also opens doors to future innovations in medical research and artificial intelligence.
Verticals
cryptoweb3
Originally published on Decrypt on 3/3/2026