Living data centers:Why AI is turning to human neurons

Traders Union
All News
Editors' Picks
How Cortical Labs technology could change the future of computing

All news

Editors' Picks

Crypto News

Financial News

Market Voices

Billionaires

Companies

Broker News

Regulators News

Darina Pashchenko

13.03.2026

Living data centers:Why AI is turning to human neurons — Biological computers: How brain cells could change architecture of AI data centers

While technology giants continue building increasingly powerful data centers for artificial intelligence, some researchers are searching for a completely different computing architecture. Australian startup Cortical Labs has introduced a system in which computations are performed not by silicon chips but by human brain cells grown in a laboratory. This experiment may mark the beginning of the era of biological computers, where the boundary between biology and technology gradually becomes blurred.

Biological Computers as a New Approach to Computing

For several years now, the development of artificial intelligence has been running into the same barrier—computing power. The more complex models become, the more energy and infrastructure data centers require. Against this backdrop, Australian startup Cortical Labs presented an unusual solution: a prototype biological data center in which computations are performed not by processors but by human brain cells. According to Bloomberg, two sites where these biocomputers will operate are planned in Singapore and Melbourne.

Such developments may seem like a scientific experiment, but interest in them is growing precisely because of the limitations of traditional computing architecture. Modern GPU clusters consume hundreds of watts per chip and require massive data centers, while biological neural systems may be capable of solving learning and adaptation tasks with significantly lower energy consumption. If these technologies continue to develop, this could represent not just another startup but an attempt to rethink the very principles behind how computing systems operate.

How biological computers work

At the core of Cortical Labs’ development is a system in which living neural cells are connected to a silicon chip. The neurons are grown in a laboratory and placed on a special electrode array that allows researchers to both read their activity and send electrical signals to them. Essentially, this creates a two-way interface: electronics stimulate the cells, and the cells respond with electrical impulses that can be interpreted as computational output.

The key feature of such systems is the ability of neurons to learn and adapt. Unlike classical processors that strictly execute programmed instructions, living neural networks can change their behavior depending on incoming signals. Experiments with these systems have already shown that biological neural cultures are capable of basic learning. For example, in the DishBrain system, cultured neurons learned to interact with a simulation of the game Pong, adapting their activity to what was happening on the screen—the results of this experiment were published in the journal Neuron. In later demonstrations, researchers also showed that neural cultures could respond to elements of gameplay in Doom, forming simple models of learning and adaptive behavior.

Loading...

The tweet was deleted by the author.

But we saved everything 🙂.

In practice, this represents a hybrid system where biology works together with programmable electronics. Silicon chips provide the interface and signal processing, while neural cells perform part of the computational work using their learning mechanisms. This approach could connect two worlds—artificial intelligence and neurobiology—creating a new type of computing that currently exists somewhere between a laboratory experiment and a future technological platform.

Why industry is Looking for alternatives to traditional data centers

Modern AI models require ever larger GPU clusters, and training large neural networks can take weeks while consuming enormous resources. The world’s largest technology companies are now building data centers worth billions of dollars because computing infrastructure has become the main limiting factor in AI development.

At the same time, the energy burden is also growing. Modern data centers already consume about 1%-1.5% of the world’s electricity, and cooling them requires significant amounts of water.

Loading...

The tweet was deleted by the author.

But we saved everything 🙂.

A single high-performance GPU can consume between 400 W and 700 W, and large clusters contain thousands of such chips. As a result, AI infrastructure is becoming one of the most energy-intensive segments of the digital economy.

This is why researchers have recently started seeking alternative computing architectures. Biological systems could potentially be far more efficient. For example, a single CL1 module from Cortical Labs consumes around 30 W, an order of magnitude less than modern graphics processors. Although these technologies are still at an early stage, their emergence shows that the industry is beginning to search for solutions to the energy crisis in computing that accompanies the rapid growth of artificial intelligence.

Where biological computing could be applied

For now, biological computing systems remain an experimental technology, but researchers are already discussing several areas where they could prove especially useful. One of the most obvious is fundamental brain research. Neural cultures connected to electronic interfaces allow scientists to observe how signals form, how cells respond to stimuli, and how learning emerges within neural networks. For neuroscience, this offers an opportunity to study processes that remain almost impossible to observe directly inside the living brain.

Another important application involves disease modeling and drug development. Neural cultures can be grown from human cells and used as models for studying neurodegenerative diseases such as Alzheimer’s or Parkinson’s. In these systems, researchers can observe how neural activity changes under the influence of different substances and test potential treatments faster and more accurately than in traditional laboratory models.

Finally, such systems may also play a role in the future development of artificial intelligence. Living neural networks naturally possess the ability to learn and adapt, which makes them a potential platform for experimenting with new learning algorithms. Unlike conventional neural networks that require enormous computing resources for training, biological systems can demonstrate adaptive behavior through interactions between cells. For this reason, biological computing could prove particularly promising in areas where learning, self-organization, and adaptation to new data are essential.

New technology means new rules

The emergence of biological computing systems inevitably raises new questions—not only technological but also ethical. Unlike traditional computers, these platforms use living human cells and therefore sit at the intersection of several fields: biotechnology, neuroscience, and the digital industry. As a result, discussions about such projects are moving beyond laboratories and becoming a topic of debate among researchers, lawyers, and bioethics specialists.

One of the key questions concerns where the boundary lies between biological material and a system capable of demonstrating signs of sensitivity or complex behavior. Today’s neural cultures are relatively simple structures consisting of tens or hundreds of thousands of cells, and they do not possess consciousness. Nevertheless, researchers acknowledge that as the technology develops, new challenges may arise—from biosafety standards to rules governing the use of human cells and limitations on creating more complex neural systems.

This is why experts increasingly argue that the legal and ethical framework for the future biological computing industry should be discussed in advance. If such technologies eventually move beyond laboratories and become part of computing infrastructure, they will likely require separate regulation—much as happened with genetic research and artificial intelligence. The sooner this conversation begins, the greater the chance that the development of this new technological field will proceed not only rapidly but also responsibly.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.