Moving through circuits of zeros and ones, millions of calculations per second, artificial intelligence algorithms operating at full speed and huge databases, all connected to… human brain cells. In the near future, such solutions may be possible, and biocomputers will greatly facilitate the development of modern information technologies.
In Frank Herbert’s ‘Dune’, set in the distant future, modified humans served as navigators on interstellar ships. Thanks to a substance called spice, they increased their brain’s capabilities so that it could find a lightning-fast path through space-time and guide the ship between systems faster than light.
Although completely fictional, Herbert’s vision was based on a premise that is true: the ‘computing power’ of the human brain is truly enormous. We may not be able to plot a faster-than-light journey in seconds, but we are very good at constructing meaningful, grammatically correct sentences and drawing logical conclusions.
These skills may seem trivial, but the truth is that computers are still learning them. AI chatbots, although impressive in their abilities, can still make mistakes when performing trivial commands. No wonder, then, that scientists are turning to biocomputers to gain new computing power for the ever-advancing development of computer technology. Where did this technology come from and what stage is it at today?
Biocomputers made of neurons. Is this the future of computer science?
The last several months have been marked by two sensational announcements. In both cases, there was talk of ‘the world’s first’ commercially available biocomputers based on human-like neurons. First, in mid-2024, Swiss start-up FinalSpark announced the launch of such a service. Last year, a similar announcement was made by Australian company Cortical Labs.
The combination of biology and computers is nothing new in science. Back in the 1990s, American biologist and computer scientist Leonard Adleman proposed solving mathematical problems using DNA strands. However, the technology used by the latest biocomputers is based on something completely different. It is no longer about creating a DNA sequence that, once read and decoded, can be used to solve certain problems. Instead, scientists are using stem cells to create something that resembles a miniature human brain. Although, as we will see later in this article, the term ‘brain’ is somewhat of an exaggeration.
Stem cells are cells in a ‘raw’ state, not specialised for any particular purpose. Scientists can then ‘cultivate’ them to create different types of cells, including the neurons that make up our brains. Currently, it is not difficult for a biotechnology company to purchase stem cells – it only requires the funds and appropriate conditions for their storage and further research. The real innovation comes in what specialists in specific laboratories can do with such cells. As it turns out, it is also possible to build residual tissue from neurons cultivated in this way and put it to work.
Does the biocomputer exist? A Polish researcher responds
In an interview with our editorial team, Dr Ewelina Kurtys, a Polish woman who has been working at FinalSpark for over four years, clearly emphasises that biocomputers do not yet exist. ‘For now, we only have prototypes,’ says the researcher. Dr Kurtys has been involved in implementing modern technologies in the business world for years. In 2022, she founded Ekai.io, which helps companies in this area. At FinalSpark, the Polish researcher serves as a strategic advisor.
If bio-computers do not yet exist, what are we striving for? What is this device supposed to be? ‘It is a computer built from living neurons,’ Dr Kurtys tells us. ‘Instead of silicon and logic gates, as in a normal computer, we use living neurons. They act as the building blocks of the computer.’
Dr Kurtys emphasises that building a biocomputer using this method is not ‘creating a brain’. Such a device would simply use the same ‘building blocks’ that make up the human organ.
Organoids in the service of humans. Artificial neurons are thinking better and better
Scientists call these artificially created clusters of neurons organoids. In the case of the technology developed by FinalSpark, these organoids are three-dimensional balls with a diameter of half a millimetre. One contains about 10,000 neurons. How do researchers manage to harness such a cluster of cells to work?
We place the organoids on electrodes. This allows us to send them electrical signals and measure their response. Neurons change the voltage in their membrane, and this can be measured as current, explains Dr Ewelina Kurtys, FinalSpark.
A computer with a written programme is connected to these electrodes. Scientists use it to send actions to the organoid and check how the neurons respond. For now, these are simple, almost primitive tests. ‘At this stage, engineers are trying to achieve anything, any effect,’ says Dr Ewelina Kurtys. ‘We try to do the simplest experiments because an organoid is a very unstable environment, unlike digital systems.’
Biocomputers, AI and the energy revolution
For now, the organoid from FinalSpark is capable of storing one bit of memory. Scientists call it a biobit because it is difficult to compare such a unit to concepts known from computers. Due to their nature, neural organoids will process information in a very different way than digital environments.
What will we gain from this? First and foremost, energy savings. In recent years, the rapid development of artificial intelligence has increased the world’s energy demand. According to a report by the International Energy Agency, by the end of this decade, the entire sector is expected to consume as much electricity as Japan. Capacities are constantly growing, but in reality, modern AI algorithms are still quite limited compared to our brains.
To enable traditional computers to generate speech that resembles human speech, we need large data centres and enormous amounts of energy. Meanwhile, each of us has a supercomputer in our heads that has much greater capabilities, and all we need to power it is a little oxygen, food and drink. If the energy efficiency of biocomputers made from human neurons approaches that of our brains, we will be able to talk about an energy revolution.
The goal is to build a computer, and the advantage of this computer will be that it will consume a million times less energy than digital devices – Dr Ewelina Kurtys, FinalSpark.
The first biocomputer in just 10 years?
FinalSpark is working on a three-dimensional organoid, while the neurons in Cortical Labs’ devices are arranged on a two-dimensional surface. The first method involves more unknowns because, as Dr Kurtys says, researchers at the Swiss company do not really know exactly what processes are taking place inside the three-dimensional structures. So why is the company using it? ‘We believe that this method will give us greater computing power,’ says Ewelina Kurtys.
According to the researcher, the first true biocomputers will be created within the next 10 years. Due to the unstable nature of neural networks, they will function differently than traditional digital computers. Today, this technology has various names: wetware, biocomputing, organoid intelligence. One of these terms will most likely stick as the technology develops.
Although the technology is well on its way, we still do not really know how to control organoids built from neurons. ‘No one knows how to program neurons. We do not fully understand how they encode information and what signals to send them to achieve the desired effect,’ admits Dr Ewelina Kurtys. As the researcher tells us, a neural organoid may one day provide a completely different solution to a task than it did the day before. This complexity may give us an idea of how complicated the brain is and how little we still know about how the human thought process works.
Consciousness from a test tube? Biocomputers and moral dilemmas
Ethics is also an important issue in the context of biocomputers. Will scientists creating organoids from neurons lead to the emergence of consciousness? Slogans such as ‘playing God’ or ‘machine rebellion’ come to mind. Is there a risk that robots powered by biocomputers will one day rebel against us for years of slave labour? According to Ewelina Kurtys, this is not a possibility.
As Dr Ewelina Kurtys, FinalSpark, explains:
“It is technically impossible. Promoting such ideas is detrimental to this research because it creates a false image of what we do in society. We are not building a brain. The brain is too complex; it is the result of millions of years of evolution, and we are not able to build it in a laboratory. We only use building blocks, i.e. neurons. There can be no talk of a brain or a person or consciousness here. Consciousness is a term from the field of philosophy, not science; we are not able to measure or record someone’s consciousness.”
At the moment, both FinalSpark and Cortical Labs are making the computing power of their prototype devices available for remote rental. This allows companies and scientific institutions from around the world to conduct experiments on neurons connected to electrodes. This open approach to the entire project may lead to further breakthroughs and a transition from single-bit prototypes to fully functional biocomputers. For now, however, research on these devices shows us above all how little we understand the workings of our own brains and consciousness.
