“Living Computers”

By Scott Hamilton

Last week I wrote about a new technology linking robotics and biology, and this week I figured I would follow it up with another merge of computers and biology. Cortical Labs introduced the world’s first code deployable biological computer. The company’s flagship product, the CL1, is currently available on their website for a cool price of $35.000. PC Gamer recently promoted the product as the first “body in a box” biological computer.

The CL1 consists of a product that combines human brain cells, which send and receive electrical signals in order to boost the cognitive abilities of the computing system. These biological cells live on the surface of the computer’s silicon based chip and the machine’s Biological Intelligence Operating System (biOS), which allows users to manipulate the biological neurons for a variety of computing tasks. For those that follow conventional computing architecture you will notice a cut play on words, as BIOS is the central control of a classic computer that describes the external hardware to the processor, so it can use things like USB drives, monitors, keyboards and mice.

You might think organic hardware is flashy and new, but biological computers have been active in labs for several years, and the CL1 is not the first in the labs, but it is the first made publicly available. FinalSpark began offering rentable “minibrains” on its Neuroplatform last year. The biggest hurdle to overcome for biological computing is the fact that the machine has living parts, as far as the definition of life goes. It contains lab-grown neurons in a nutrient rich solution, which are kept alive in a tightly temperature controlled environment. They are also kept alive by an internal life support system which can keep the system alive for up to six months. The project’s chief scientific office’s Brett Kagan began pitching that the fact that the system requires a tightly controlled environment to remain “alive” is “like a body in a box.”

My first thought was why, until I realized that it has always been a focus of the AI community to make computers more and more life-like, with many qualities of human likeness. This project is just another step toward creating artificial life. Cortical Labs unveiled the product two weeks ago at Mobile World Congress in Barcelona. Cortical Labs has been in the news before with their progress in biological computers, when they announced last year that they were able to coax brain cells in a petri dish to learn to play Pong faster than AI, proving that they could teach the cells to learn.

The thought process behind the CL1 is that while an AI can look like it is learning new things and give the appearance of thought, it lacks the true understanding of context. AI can not really understand the concept of life and death. These lab-grown neurons stand a much better chance of really learning and adapting, not to mention that they are much more energy efficient than their AI counterparts. As it turns out, these lab grown cells are showing the silicon technologies new tricks.

However, all this work with biological computers raises a big question of ethics. What are these brain cells experiencing? Do they experience pain? Are they suffering? Do they understand that they are alive? These are questions that cannot and should not be avoided. A paper published by EurekAlert in October 2022 made claims that the early versions of this device led the scientists to believe the cells experience sentience, which in simpler terms means that they understood they were alive. When they exposed the cells to a virtual environment they were able to manipulate the cells into working together and the cells gained the ability to accurately communicate with each other and the outside world via means of the electrical signals.

According to Professor Karl Friston, the cells were learned by interacting with their environment. They worked to improve their living conditions, which was used as a part of the training. He says, “Remarkably, the cultures learned how to make their world more predictable by acting upon it. This is remarkable because you cannot teach this kind of self-organization; simply because – unlike a pet – these mini brains have no sense of reward and punishment.” The basic principle behind their training is based on the “free energy principle” which states that cells at this level self-organize to interact with their environment on the physical level with the express goal of reducing unpredictability in their environment. The more predictable the environment, the less energy is lost by the environment.

While I find this type of technology fascinating, I am not sure where the line should be drawn. Should we be creating life to power our computing systems? Are there health risks related to having this biological material in your computer? Are there ethical issues behind manipulating these lab grown cells to behave in expected ways? And what happens to your $36,000 body in a box when it dies? Do you have to start all your training work over again with the new “body?” I guess eventually all these questions will be answered; in the meantime, we should just stay safe and learn something new.

Scott Hamilton is an Expert in Emerging Technologies at ATOS and can be reached with questions and comments via email to sh*******@**********rd.org or through his website at https://www.techshepherd.org.