Living human neurons from Cortical Labs are now playing Doom on a $35,000 biological computer

Cortical Labs Pushes Biological Computing Into Commercial Reality

Australian biotech startup Cortical Labs has demonstrated a striking advancement in biological computing: a system where human neurons are playing Doom 1

1

. The company's CL1 platform, priced at $35,000 per unit, represents the first commercial computer made of living human neurons capable of running complex software 1

1

. This achievement builds on Cortical Labs' 2022 breakthrough when researchers taught 800,000 neurons to play Pong, but the new hardware pushes biological computation into engineered, commercially available systems 1

1

.

The decision to run Doom wasn't arbitrary. Since 1993, the game has served as a universal benchmark for testing unconventional hardware, from graphing calculators to pregnancy tests 1

1

. But unlike those hacks, the CL1 isn't simply running the game—the neurons are actually playing it 2

2

. On-screen data gets mapped to patterns of electrical stimuli transmitted to the neurons, which respond with signals that control the character's actions 2

2

. When neurons fire in specific patterns, the in-game character shoots or moves, creating a genuine feedback loop between biological tissue and digital environment.

How the CL1 System Works With Cultured Human Cortical Cells

The CL1 microchip contains approximately 200,000 cultured human cortical cells mounted on a multi-electrode array 2

2

. These aren't simulated neural networks but genuine neurons derived from skin or blood cells taken from adult donors 1

1

. Through a process involving induced pluripotent stem cells, these samples are reprogrammed and differentiated into cortical brain cells 1

1

.

The living tissue sits inside a sealed chamber connected to an internal life-support system that regulates gas composition, temperature, and waste filtration 1

1

. Under optimal lab conditions, the neurons remain active and viable for up to six months 1

1

. The hardware features 59 electrodes positioned on a planar array of metal and glass, with the denser contact grid and upgraded signal processing reducing latency from milliseconds to sub-millisecond speeds 1

1

. This allows the living network to respond almost as fluidly as conventional processors.

Synthetic Biological Intelligence and the biOS Operating System

To manage this biological substrate, Cortical Labs developed biOS, an operating system that sends and receives electrical stimuli through the electrode array 1

1

. Developers can deploy code directly to the neuron layer, where feedback signals shape adaptive pathways similar to synapses in a biological brain 1

1

. When trained on games like Pong or Doom, the networks react to stimuli—receiving reward signals when achieving goals and corrective signals upon failure—forming self-organized response patterns 1

1

.

Cortical Labs terms this computational model Synthetic Biological Intelligence, distinguishing living computation from traditional artificial intelligence 1

1

. The company emphasizes that these neurocomputers leverage the plasticity inherent in neural networks—the same adaptability that makes human brains powerful 2

2

. Currently, the cells "play like a beginner who's never seen a computer," but researchers note the potential for dramatic improvement as the networks adapt 2

2

.

Commercial Availability Through Cortical Cloud and Research Institutions

The CL1 ships as a self-contained desktop unit but also integrates into 30-unit server rack configurations targeted at research institutions 1

1

. Volume pricing drops to $20,000 per unit in rack configurations, with a full rack consuming roughly 850 to 1,000 watts—power levels comparable to a single midrange GPU server 1

1

. Cortical Labs began shipping the first 115 commercial systems in 2025 and maintains cloud connectivity for live monitoring and remote code deployment 1

1

.

The company launched Cortical Cloud, allowing developers worldwide to experiment with the CL1 via a Python API 2

2

. This open approach mirrors the company's decision to post demonstration footage on YouTube and release open-source code to GitHub 1

1

. The platform enables researchers to explore biological computing applications beyond gaming, though the practical uses remain largely speculative at this stage.

Questions About Ethics and Future Implications

The technology raises immediate questions about sourcing and consent. The neurons contain someone's DNA, prompting concerns about whether this could lead to situations similar to the Henrietta Lacks case, where biological material was used extensively without proper consent or compensation 2

2

. The prospect of human neurons being deployed for various applications—some potentially problematic—adds another layer of complexity to the ethical considerations surrounding biotechnology 2

2

.

For AI researchers and developers, the CL1 represents a fundamentally different approach to computation. Instead of silicon-based processors mimicking biological processes, this system uses actual biological tissue to perform computational tasks. The short-term implications center on research applications, where the six-month viability window and $20,000-to-$35,000 price point make it accessible to well-funded institutions. Long-term speculation focuses on whether these systems can scale beyond simple games to handle more complex tasks, and whether the plasticity of neural networks will enable performance improvements that rival or exceed conventional systems in specific applications.