Biological Intelligence on Silicon

Cortical Labs is a biocomputing company that creates hardware and software platforms powered by living biological neurons. Based in Melbourne, Australia, the company is known for creating DishBrain, a system where approximately 800,000 lab-grown neurons were integrated into a silicon chip and taught to perform tasks such as playing the game Pong. This approach differs from traditional artificial intelligence, which uses silicon-based processors to simulate neural networks. Cortical Labs uses actual biological tissue as the computational substrate, merging it with traditional digital systems to create what they call a living computer.

The technology is built on the premise that biological brains possess inherent advantages over silicon for certain tasks. While modern Large Language Models (LLMs) require massive datasets and gigawatts of power to reach proficiency, biological neurons are capable of intuitive learning from much smaller datasets with a tiny fraction of the energy consumption. A human brain operates on roughly 20 watts of power; Cortical Labs aims to replicate this efficiency in a commercial computing format. By growing neurons directly onto custom chips, the company enables a form of intelligence that learns through feedback loops rather than the massive backpropagation cycles used in deep learning.

The CL1 and Cortical Cloud

To make this technology accessible to developers, Cortical Labs offers the CL1 hardware and the Cortical Cloud. The CL1 is a physical device that allows users to deploy and test technology directly on real neurons in a local environment. For those without the physical hardware, the Cortical Cloud provides a distributed platform to interact with these biological systems remotely. This setup allows researchers to experiment with biocomputing without needing to maintain their own wet-lab or handle biological materials directly. The company recently demonstrated the capabilities of this system by running the game Doom on their biological-digital hybrid, showing that these neurons can process complex information and respond in real-time.

Competitive Position and Energy Arbitrage

Cortical Labs sits in a unique position within the AI hardware stack. They are not merely building another AI accelerator or a neuromorphic chip that mimics brain architecture; they are using the brain’s own components. This places them in competition with traditional compute providers for specialized research use cases, particularly in drug discovery and neurological study. As the costs of scaling silicon-based AI continue to rise due to energy demands and data scarcity, Cortical Labs offers a hedge. Their value proposition rests on the idea that biological systems are the most sophisticated learning machines in existence. By providing the tools to harness these systems, the company seeks to move computing away from pure silicon and toward a hybrid future where biological wetware handles complex, low-power learning tasks.