There is a remarkable “biological computer” in a laboratory outside of Cambridge. The 200,000 human brain cells bred in the laboratory lie on silicon circuit, which convey their synchronized electrical activity on a screen.
The CL1 device over the size of two shoe boxes was developed by Australian start-up cortical laboratories with the British bit.
“Like our brain, biological computers will consume many orders of magnitude less energy than conventional electronics if they process information. Future applications could include robotics, security and meta -verses,” said the managing director of the Cortical Labs, Hon Weng Chong, Financial Times.
The rapidly growing search for alternatives to energy -intensive conventional electronics has stimulated the new field of Biological computerWhat aims to type directly into the intelligence of brain cells instead of simulating them in silicon through “neuromorphs”.
In this movement, Cortical Labs is at the top of this movement, although academic groups and other start-ups such as the Swiss group finals and the biological black box make progress in the USA.
Early applications of CL1 are in neuroscience and pharmaceutical research and determine how different chemicals and drug candidates influence the information processing of the brain cells.
“The next stages of innovation will enable new and more advanced forms of calculation beyond conventional AI systems beyond the same processors – neurons – beyond intelligence in living organisms.
For Mark Kotter, professor of clinical neurosciences at the University of Cambridge and Bit.Bio, the meaning of CL1 is “that it is the first machine that can reliably evaluate the computing power of brain cells. This is a real paradigm shift.”
Experts found that CL1 was a “remarkable performance” that contributed to promoting the young biological computer field.
Karl Frriston, a neuroscientific professor at the University of College London, who also worked academically with a number of scientists from the cortical laboratories, said it could be considered the first commercially available biomimetic computer.
“However, the real gift of this technology is not computer science – at the moment. Rather, it is an activation technology that enables scientists to carry out experiments on a small brain.”
Professor Thomas Hartung from Johns Hopkins University in Baltimore, who is examining “Organoid intelligence” Using cerebral organoids or mini hammers from stem cells, the outstanding contribution of cortical laboratories said in developing virtual games as a benchmark for biological computing.
The predecessor of CL1 named Dishbrain learned the simple video game pong, in which he moved a virtual paddle up and down to distract a ball.
The training included the indication of the neurons a “reward stimulus” if they moved the paddle correctly by using the electrical activity in the form of a sine wave that like the cells. The “punishment” when they misunderstood it was unpleasant white noise.
Experiments with Dishbrain and CL1 show how different diseases affect the information processing of the neurons, measured by the way they play a pong. “We treated them with chemicals that have an impact on our brain,” said Bit.Bios Kotter. “This machine shows, for example, that alcohol breaks down its calculation.”
Another experiment compared the effect of three epilepsy treatments and found that one of them, carbamazepine, was superior to improving gameplay metrics.
“We think a lot about how we can program our biological computers,” said Chong. “A big question is how we present digital information for these neurons.” The scientists teach the neurons the forms of the digits, he added, “and they now begin to see that a nine out of four or five differ.”
Cortical laboratories and bit.Bio lay pure layers of two specific types of neurons on the silicon circuit of the CL1 biocomputer – one to excite electrical activity, and the other to moisten them down. “The balance between acceleration and brakes is really important,” said Chong. The neurons are bred from stem cells that were originally derived from human skin.
Others such as Switzerland’s final park examine biological computing with cerebral organoids. But bit.Bio and cortical laboratories believe that their layers of standardized neurons provide more reproducible results than organoids.
“Our neurons look very homogeneous,” said Tony Oosterveen, who heads bit. BIOS brain cells work. “If you look at other technologies, you will see great variations. Our strength is to make pure populations.”
Regardless of the long-term promise of the biocomputation, its supporters enter this assumption for more general applications and AI decades in the future. One problem is to work out an efficient programming system.
Another is that the neurons can only live in a CL1 for a few months, which is experienced by a constant liquid flow to supply nutrients and remove waste products.
“A disadvantage of such a system is that we have not yet worked out how to carry out memory transmission,” said Chong. “As soon as the system dies, you have to start all over again.”
Chong is aware of the ethical concerns that could occur in the future if biological computers and neuron cultures develop the basics of consciousness.
At the moment he said: “These systems are sensitive because they react to stimuli and learn from them, but are not aware. We will learn more about how the human brain works, but we do not intend to create a brain in a VAT.”
