Australian Startup Merges Human Brain Cells with Silicon to Create Living Computer
Cortical Labs has developed the CL1, a computing system that combines lab-grown human brain cells with silicon hardware. The system has been demonstrated playing the 1993 video game Doom.
Technical Specifications
The CL1 system operates by growing approximately 200,000 neurons derived from stem cells and placing them on a specialized chip equipped with microelectrodes. These electrodes send and receive electrical signals to and from the living neurons, enabling interaction. The stem cells used to grow the neurons can be generated from blood or skin samples, according to Cortical Labs Chief Scientific Officer Brett J. Kagan.
The system is approximately the size of a shoebox and sustains the living cell cultures using a nutrient-rich liquid medium, a technique sometimes referred to as "wetware." Approximately 120 CL1 units are currently operational in a data center in Melbourne, Australia. Cortical Labs is also establishing biological computing facilities in Melbourne and Singapore.
Gaming Demonstration
Researchers converted the digital environment of Doom into electrical signals the neurons could interpret. Specific electrode patterns correspond to game actions such as moving or firing. Researchers monitor neuronal activity and adjust inputs to train the cells.
Alon Loeffler, Cortical Labs' senior application scientist, stated that the neurons initially performed like a beginner, walking into walls and shooting incorrectly, but over time targeted enemies more regularly.
Applications and Capabilities
Cortical Labs reports that the CL1 system enables real-time adaptive learning and goal-directed behavior. Potential applications cited by the company include:
- Robotics
- Real-time learning tasks comparable to artificial intelligence
- Drug screening
- Disease modeling
- Personalized medicine
- Healthcare applications
Efficiency and Limitations
Brett Kagan stated that human biology requires significantly less data for learning compared to machine learning models and can manage uncertain or noisy information effectively. The human brain operates on approximately 20 watts of power, which the company highlights as more energy-efficient than current silicon computing.
Traditional silicon-based computers maintain advantages in precise, fast mathematical calculations. Kagan suggested future computing systems would likely combine biological and silicon-based approaches.
Limitations of the current system include:
- The neurons have a six-month lifespan
- The system does not yet produce consistent, programmable results
Expert Commentary
William Keating, CEO of semiconductor research company Ingenuity, stated the project represents "real science making real progress."
Ethical Considerations
Alysson R. Muotri, Director of Sanford Stem Cell Education and Integrated Space Stem Cell Orbital Research (ISSCOR) Center, stated that simpler networks of human neurons, such as those used by Cortical Labs, do not present significant ethical issues.
However, he noted that more complex, three-dimensional brain-like structures (organoids) could potentially generate a form of consciousness, which might necessitate new regulatory frameworks. Cortical Labs' Kagan indicated their approach could offer ethical advantages, including a reduction in the need for animal testing.