
Scientists create world’s first living computer that thinks with human brain cells
Australian startup Cortical Labs introduced the first commercially available biological computer capable of running code. The system, called CL1, combines live human brain cells with silicon hardware, providing a new platform for neuroscience and biotechnology research.
Built with Living Neurons: A New Era of Brain-Powered Machines:
Launched in March, the CL1 uses around 800,000 lab-grown human neurons derived from adult skin or blood cells. Scientists culture these neurons on a chip and maintain them in a closed-loop environment that supplies nutrients, regulates temperature, filters waste, and stabilizes fluid levels. The neurons communicate through electrical pulses and adapt in real time to external stimuli.
Unlike traditional computing systems, CL1 processes information using sub-millisecond electrical feedback loops. This creates a living neural network that responds dynamically and performs biological computation based on natural intelligence.
Real-Time Computation:
The platform functions through a continuous cycle of input, adaptation, and output. It introduces small electrical signals—representing data—to the neurons, which then generate measurable responses. The system’s firmware and hardware capture and interpret these reactions, enabling real-time computation and learning behavior.
Brett Kagan, Chief Scientific Officer at Cortical Labs, explains that this setup mimics the brain’s communication method. It allows researchers to observe how neural circuits react to different types of input, including drug stimulation and simulated lesions.
Scientists Turn “Brain in a Vat” Idea into Reality:
Researchers primarily use the CL1 for scientific purposes. They leverage the system to study neurological processes, genetic differences, and responses to pharmaceutical interventions. The platform provides a closed-loop experimental environment where neural activity continuously interacts with simulated digital environments.
Karl Friston, a theoretical neuroscientist at University College London, describes CL1 as a significant tool for experimentalists. He notes that it brings to life the “brain in a vat” concept long discussed in philosophy and cognitive science.
Availability and Cost:
Cortical Labs will start shipping the first 115 CL1 units this summer. They will price individual units at \$35,000, with discounts lowering the cost to \$20,000 per unit for bulk purchases of 30-unit server racks. The company also offers a cloud-based service, enabling researchers to access and run experiments on in-house units for \$300 per week.
Each unit stays viable for up to six months. A full server rack consumes between 850 and 1,000 watts of power, which is considerably less than typical AI data center requirements.
A New Direction for Bio-Computing:
While CL1 does not aim to replace traditional silicon-based AI, the platform offers researchers a new way to study learning, adaptation, and cognition. Friston emphasizes that the technology may have its greatest impact outside computer science, serving as a transformative research tool to explore how biological systems process and respond to information.