It’s not novel for people to try and make computers more like human brains. Although there are some obstacles to overcome before we get there, a team of researchers from Johns Hopkins University claims that there may be many advantages to applying this idea more literally by using real neurons.
The researchers recently published a roadmap outlining the steps necessary to develop biocomputers powered by human brain cells (not taken from human brains, though). One of the researchers also asserts that the suggested “organoid intelligence” would have some glaring advantages over current computers.
“We have always attempted to make our computers more brain-like,” Thomas Hartung, a researcher at Johns Hopkins University’s Environmental Health and Engineering department and one of the paper’s authors, told Ars. The brain is virtually unmatched as a computer, at least in theory.
It’s alive (kind of?)
Organoids are tiny bits of tissue that are developed in the lab to mimic complete organs; they include many of the specialized cell types found in an organ as well as some internal structures. They make it possible for researchers to conduct investigations without using any human or animal subjects. Organoids made from human brain cells have been the subject of research by Hartung and several other collaborators. In 2012, he began cultivating them using samples of human epidermis that had been reprogrammed to resemble stem cells. Humans still perform admirably at Go despite the fact that it would take an extremely long period of time for them to play these games. The capacity of the human brain to retain data is also very high; it is thought to be 2.5 million gigabytes. The ExoMars project, which has been delayed, will finally take off in 2028 and land its Rosalind Franklin rover in the clay-rich plain of Oxia Planum.
“This European rover will contain a drill with the unprecedented potential of reaching down to a depth of 2 meters (6.5 feet) to analyze sediments better protected against the harsh Martian surface conditions,” Fairén said. If biosignatures are better maintained at depth, as we anticipate, deep samples will contain a greater variety and abundance of biosignatures.
Theoretically, organoid computers would also be smaller in size. Instead of just being something on a Petri dish, they might eventually have a three-dimensional structure, which would cause their cell density to significantly increase and more connections to develop between the neurons. Finally, human brains are superior at reasoning logical choices, such as being able to rapidly identify an animal, despite completely non-human computers being better at processing massive amounts of data.
The promise of this type of computing is also recognized by Sukhpal Singh Gill, an assistant professor of electronic engineering and computer science at the Queen Mary University of London. He pointed out that a significant barrier to computing, AI, and machine learning is electricity consumption. In contrast, he said, “Human brain cells integrate such tasks with ease and also have extremely low energy requirements, requiring only a relatively small volume of a nutrient-rich solution to operating.
A long road ahead
Before these biocomputers are a fact, however, there is still a lot of work to be done. Size is one of the most urgent problems. According to him, the organoids would need to grow from 50,000 to more than 10 million cells. But because the oxygen and nutrients fed to them can’t easily reach the center, it’s difficult to maintain these cells at a size larger than half a millimeter, meaning you could have a “rotten core” on your hands. Perfusion, or directing liquid into the interior of the organoid, might be one answer, according to Hartung.
In order to exchange information with organoids, much like a computer does, researchers must also develop novel methods of communication with them. Additionally important is a cognitive improvement. Brain organoids can only currently store short-term recollections. You could teach them to play Ping, for example, but Hartung warned that they would forget everything the following day. According to him, a portion of this is probably due to the organoids’ absence of microglial cells. These immune cells also show up in the brain and carry out synaptic pruning, or the removal of unnecessary synapses, which permits healthy brains to operate properly.
Additionally, there are moral concerns regarding the development of new brain cells as well as whether or not these organoids would be considered alive and capable of feeling pain or experiencing some form of awareness. In order to assess and ensure that everything is done in a manner that is not morally terrifying, Hartung notes that his team has been working in collaboration with bioethicists.
