Cortical Labs: How Biological Computers Use Living Neurons

Cortical Labs is a biotechnology and neurotechnology company developing a different approach to computing: instead of relying only on silicon processors, it combines living neurons with electronic hardware.

The idea sounds unusual, but the underlying concept is easier to understand than it first appears. Cortical Labs grows neural cells on specialized silicon chips and uses electrical signals to communicate with those cells. Software creates an environment for the neurons, sends information to them, and reads their activity back through the same system.

The company became widely known for DishBrain, an experimental system in which approximately 800,000 human and mouse-derived neurons were connected to a computer and used to play the simple video game Pong. Research published in Neuron found that the neural cultures showed changes in performance over time when they received closed-loop feedback.

Cortical Labs has since moved beyond the original Pong experiment with its CL1 biological computer and Cortical Cloud, making biological neural systems available as research and computing platforms.

What Is Cortical Labs?

Cortical Labs is an Australian biotechnology company focused on biological computing, a field that explores how living biological systems can participate in information processing.

Traditional computers use electronic components such as transistors to process information. Artificial intelligence systems use mathematical models running on digital hardware. Cortical Labs takes a different route by using living neural networks as part of the computing system.

The company's approach combines three important components:

1. Living neurons
2. Silicon-based electronic hardware
3. Software that creates a closed feedback loop

This combination allows researchers to study how biological neural networks respond, adapt, and process information in controlled digital environments.

Cortical Labs describes this broader approach as biological or synthetic biological intelligence. Its current platform includes the CL1 biological computer, Cortical Cloud, and software tools that allow researchers to interact with biological neural networks.

Who Founded Cortical Labs?

Cortical Labs was founded in 2019, and its development has been closely associated with Dr. Hon Weng Chong and other researchers working in neuroscience, biotechnology, and computing.

The company's current team includes Brett Kagan, CEO and Chief Scientist, along with technology and hardware leadership.

The original idea came from a simple but important question: if biological brains can process information and demonstrate learning, could living neurons be connected directly to computers and used as an active part of a computing system?

That question became the foundation for the company's research.

How Does Cortical Labs' Technology Work?

The easiest way to understand the technology is to think of it as a communication system between living neural tissue and a computer.

Neurons are placed on a specialized multi-electrode array. The electronic interface can stimulate the neural network and record its electrical activity.

Software then creates a digital environment around the neurons.

Cortical Labs connects living neurons with electronic hardware and software. The system sends electrical signals to the neurons and records how they respond. That response can then be used by the computer to update the digital environment.

This creates what is known as a closed-loop system. Instead of simply recording neural activity, the system allows neurons to receive information, respond to it, and then influence what happens next.

The CL1 biological computer is designed for this type of real-time interaction. Cortical Labs says it provides programmable two-way communication with biological neural networks while also providing the life-support conditions needed to keep the cells functioning.

What Was DishBrain?

Before CL1, one of the most important milestones associated with Cortical Labs was DishBrain.

DishBrain was an experimental biological neural network connected to a computer through a high-density multi-electrode array. Researchers placed approximately 800,000 human and rodent cortical neurons into the system and connected their activity to a simulated version of Pong.

The experiment was deliberately simple.

The neural network received electrical information representing the position of the ball. Its electrical activity was then interpreted as an action controlling the paddle.

When the system performed poorly, the feedback changed. When the paddle successfully interacted with the ball, the network received different feedback.

Researchers observed improvements in performance over time under the closed-loop setup. The scientific paper reported apparent learning within minutes and found that comparable stimulus without the feedback structure did not produce the same learning pattern.

This experiment helped demonstrate why biological neural networks could be interesting for computing research.

How Did Neurons Learn to Play Pong?

The Pong experiment is often simplified into the phrase “brain cells learned to play Pong.”

The reality is more specific.

The neurons were not given a traditional software program containing instructions for playing Pong. Instead, the researchers created an environment in which electrical signals represented information about the game.

The neural activity produced responses that affected the game.

Because the system operated as a feedback loop, the neurons were repeatedly exposed to the consequences of their activity. Over time, the researchers observed changes in the network's behavior.

This is important because it shows the difference between conventional programming and biological learning.

A conventional program follows instructions written by a developer. A biological neural network can change its activity based on the signals and feedback it receives.

Is DishBrain a Human Brain?

No.

This is one of the most important points to understand about Cortical Labs.

DishBrain and similar systems use cultured neural cells, not a complete human brain. They do not have a human body, normal brain anatomy, sensory organs, blood supply, developmental history, or the complex organization of an intact brain.

Neurofounders also highlights the terminology problem surrounding words such as “brain,” “intelligence,” cognition, and sentience. A neural culture can process information and respond to feedback without being equivalent to a complete human brain.

Therefore, terms such as biological computer or neural culture provide a more useful description than imagining a miniature human brain inside a machine.

What Is the CL1 Biological Computer?

The CL1 is Cortical Labs' commercial biological computing platform.

Unlike the earlier DishBrain research setup, CL1 is designed as a self-contained system that integrates the biological neural network, electronic interface, software environment, and life-support components into one platform.

Cortical Labs describes CL1 as a code-deployable biological computer. The company says neurons are maintained in a nutrient-rich environment on a silicon chip, while its Biological Intelligence Operating System, or biOS, creates the digital environment in which the neural network operates.

The system is designed to support:

• Real-time neural recording
• Electrical stimulation
• Closed-loop experiments
• Software-controlled experiments
• External device connections
• Biological neural network research
• Brain and cognition research
• Drug and disease-related studies

The company says the CL1's internal environment can maintain neurons for up to six months under its designed conditions.

What Is biOS?

biOS, short for Biological Intelligence Operating System, is an important part of the Cortical Labs platform.

It should not be thought of as a conventional operating system like Windows or Linux.

Instead, biOS provides the digital environment that connects software with the biological neural network.

It can create a simulated environment, deliver information to the neurons through electrical stimulation, and interpret the resulting neural activity.

This makes it possible to create a continuous feedback loop between biological cells and digital systems.

In simple terms, biOS acts as a bridge between code and living neural tissue.

What Is Cortical Cloud?

Not every researcher has access to a specialized biological computing laboratory.

This is where Cortical Cloud becomes important.

Cortical Labs describes Cortical Cloud as a biological cloud computing platform that allows users to work with biological neural networks remotely without maintaining their own specialized lab or CL1 device.

The platform is designed around remote access, software tools, and the company's Python SDK.

Researchers can execute code from a browser-based environment and interact with biological computing resources remotely.

That changes the accessibility of the technology.

Instead of every research group needing to purchase and maintain biological hardware, a cloud-based model can allow more people to experiment with the underlying platform.

Cortical Labs vs Traditional AI

Cortical Labs is not simply trying to build another version of a conventional AI model.

Traditional AI systems generally operate using mathematical algorithms running on electronic hardware. Training can require substantial datasets and computing resources.

Biological computing starts from a different substrate: living neural networks.

This does not mean biological computing automatically replaces conventional AI.

Instead, it represents a different research direction. Cortical Labs is investigating whether biological systems can offer useful properties for learning, adaptability, energy efficiency, and scientific research.

Why Use Living Neurons for Computing?

The main reason is that biological neurons are naturally designed to process information.

Brains evolved over an extremely long period to operate with limited energy while handling complicated environments.

Cortical Labs argues that biological neural networks may offer interesting characteristics that conventional systems attempt to reproduce computationally, including adaptability and learning efficiency.

The important scientific question is not simply whether neurons can perform a task.

It is whether researchers can control, measure, understand, and reproduce useful biological information processing in a reliable computing platform.

That is a much broader challenge.

Potential Applications of Cortical Labs Technology

The technology has potential applications beyond video games.

Drug Discovery

Living neural networks can potentially be exposed to compounds while researchers monitor changes in neural activity.

This could provide another way to study how substances affect biological neural systems.

Disease Research

Researchers may use biological neural networks to investigate mechanisms related to neurological conditions and changes in neural function.

AI Research

Biological systems can provide researchers with another model for studying learning, adaptation, and information processing.

Robotics

A biological neural network could potentially be connected to sensors or robotic systems, allowing researchers to study how living neural networks respond to real-world information.

The CL1 is designed with external connectivity that can support experimental setups involving additional devices.

Neuroscience

The platform can also be used to investigate how neural networks respond to stimulation and how their activity changes during learning.

Cortical Labs maintains a research program covering biological neural networks, NeuroAI, neural plasticity, drug-related research, and other areas.

Can Cortical Labs Technology Reduce Energy Use?

Energy efficiency is one of the major ideas behind biological computing.

Cortical Labs says biological neural systems can perform certain learning processes with substantially lower energy requirements than conventional approaches and positions biological computing as a potential route toward more efficient computing.

However, this should not be interpreted as meaning that biological computers are already more efficient than every conventional computer for every task.

The technology is still an emerging research field.

Keeping living cells alive also requires environmental control, nutrients, temperature management, and supporting hardware.

The interesting question is therefore not simply “Are biological computers lower power?”

It is:

Can biological neural networks perform useful computing or research tasks efficiently enough to justify the additional biological infrastructure?

That is one of the questions researchers are still exploring.

What Makes CL1 Different From a Normal Computer?

A normal computer can be switched on, run software, and perform calculations using electronic components.

CL1 has another layer of complexity because it contains living biological material.

The neurons need an appropriate environment to remain viable. The system therefore has to manage biological conditions while simultaneously handling electronic signals and software.

Cortical Labs says CL1 integrates life-support functions, neural recording, stimulation, software, and computing into a self-contained platform.

This is one reason biological computing is considerably different from simply putting a neural network algorithm onto a faster processor.

What Are the Limitations of Biological Computing?

Biological computing is promising, but it also has significant challenges.

Living Systems Are More Complex

Electronic components can be manufactured to extremely consistent specifications. Biological cells naturally vary.

That variability can make biological experiments harder to standardize and reproduce.

Neurons Need Care

A conventional processor does not need nutrients or a controlled biological environment.

Living neural networks do.

The supporting infrastructure therefore becomes part of the computing platform.

Scaling Is Difficult

A human brain contains an enormous number of interconnected neurons. A laboratory neural culture is much smaller and less structurally complex.

Scaling biological computing while maintaining stable and useful neural networks remains an important challenge.

Biological Behavior Is Not Fully Predictable

Neural networks are adaptive systems. Their behavior can change over time.

That can be scientifically valuable, but it also creates challenges when researchers need highly predictable outputs.

Ethical Questions Around Biological Computers

The use of living neurons in computing also creates questions that ordinary computer engineering does not normally face.

For example:

• How should biological neural systems be described?
• What terminology should researchers use for cognition or intelligence?
• Are there ethical boundaries for increasingly complex neural cultures?
• How should researchers monitor biological systems as their capabilities increase?
• What standards should guide future experiments?

These questions become more important as biological computing develops.

Cortical Labs has participated in discussions around terminology and the need for broader consensus concerning concepts such as cognition, intelligence, and sentience.

The goal should be to distinguish scientific evidence from dramatic descriptions so that people understand what these systems actually demonstrate.

Why Cortical Labs Matters to the Future of Computing

The most interesting part of Cortical Labs is not simply that neurons played Pong.

The bigger idea is that living neural networks can be treated as an experimental computing substrate.

For decades, computers have become increasingly powerful by improving electronic hardware and algorithms.

Biological computing asks a different question:

What happens if we use biology itself as part of the computing system?

Cortical Labs is attempting to build the infrastructure needed to explore that question.

Its progression from DishBrain to CL1 and Cortical Cloud shows a shift from a scientific demonstration toward a platform that researchers and developers can access. Neurofounders describes this as a move from research demonstration toward commercial infrastructure.

Cortical Labs Timeline

2019 — Company Founded

Cortical Labs was established in Melbourne and began developing technology around biological neural networks.

2021 — Neurons on a Chip

The company reports that it grew real neurons on a chip and demonstrated their ability to interact with a computer-controlled environment.

2022 — DishBrain and Pong

Researchers published the DishBrain study in Neuron, demonstrating learning-related changes in cultured human and rodent neural networks interacting with a simulated Pong environment.

2025 — CL1

Cortical Labs introduced CL1 as a code-deployable biological computer designed for research and experimentation with living neural networks.

2026 — Cloud-Based Access

Cortical Cloud provides remote access to biological computing infrastructure, allowing users to work with neural systems without maintaining a specialized biological laboratory.

Conclusion

Cortical Labs represents an unusual intersection of biotechnology, neuroscience, artificial intelligence, and computer engineering.

Its work shows how living neurons can be connected to electronic systems and placed inside controlled digital environments. DishBrain demonstrated the concept through Pong, while CL1 takes the idea toward a dedicated biological computing platform. Cortical Cloud extends access by allowing researchers to interact with biological computing resources remotely.

The technology is still developing, and many of its biggest questions remain open. Researchers are still exploring where biological neural networks are genuinely useful, how reliably they can be controlled, how they can scale, and what ethical frameworks should guide their development.

That uncertainty is also what makes the field interesting. Rather than asking computers to imitate biology entirely through mathematics, Cortical Labs is exploring what happens when biology itself becomes part of the computer.

FAQs

What is Cortical Labs?

Cortical Labs is a biotechnology company developing biological computing systems that combine living neurons with silicon hardware and software.

What is CL1?

CL1 is Cortical Labs' self-contained biological computer designed to let researchers interact with living neural networks through software and electronic interfaces.

What is DishBrain?

DishBrain was an experimental neural computing system that connected cultured neurons to a computer-controlled Pong environment. The study involved approximately 800,000 human and rodent neurons.

Did Cortical Labs create a real human brain?

No. Its systems use cultured neural cells and are not complete human brains.

Can biological computers learn?

Research involving DishBrain showed changes in neural network performance under a closed-loop feedback environment.

What is Cortical Cloud?

Cortical Cloud is a remote biological computing platform that allows users to interact with Cortical Labs' neural computing infrastructure without maintaining their own specialized laboratory.

What is biOS?

biOS stands for Biological Intelligence Operating System. It provides the software environment that connects digital simulations and code with the biological neural network in Cortical Labs systems.

Is Cortical Labs replacing artificial intelligence?

No. Cortical Labs is developing an alternative biological computing approach that can complement research into AI, neuroscience, and computing rather than simply replacing conventional AI.