It sounds like an episode of science fiction from the ’70s, but scientists have already created ‘living computers‘ that run by what looks like a tiny Meatwad of brain neurons (0.5-millimeter-wide). Meatwad is a living meatball from an early 2000s animated show. But you get the idea: a ball of living tissue that raises ethical questions about whether such brain tissue, called organoids, could one day even become sentient. Thus far, they don’t think so, but we don’t exactly understand consciousness to begin with.
Meatwad via YouTube/Adult Swim
It’s not like these masses are actual human brains kept alive in a jar like an episode of Wonder Woman called Gault’s Brain.
But there are some interesting similarities. It all brings to mind another Sci-Fi epic, the Matrix, although these are merely cells and not entire organisms or humans.
Video via Wonder Woman TV HD:
Brain Organoid Computers
While the comparison with a brain in a jar is a bit of a stretch, the scientists are using living human neurons derived from Stem cells. Even though they might superficially look like Meatwad from Aqua Teen Hunger Force, the proper term is organoid.
Inside a human being, brain neurons can live an entire lifetime. However, they “die” much faster inside their computer processing unit. With improved technology, they can live for 100 days. While it doesn’t sound like much, they only stayed alive for a few hours in the earliest attempts.
Dopamine Rewards for Good Organoids
To encourage these living computers to ‘behave” as needed for computations, the scientists “reward” the organoids. If they are good little organoids, scientists give them a hit of dopamine derived by exposing them to a flash of light. With rewards, these cells “learn,” creating new connections and pathways between neurons just as we do.
“The neurons are selectively exposed to the feel-good neurotransmitter dopamine to mimic the human brain’s natural reward system. These twin setups—positive dopamine rewards and electrical stimulation—train the organoids’ neurons, prompting them to form new pathways and connections much in the same way a living human brain appears to learn,” writes LiveScience.
They also stimulate the cells with electrodes, which also record data from the cells. According to MSN, the organoids store information in “biobits.”
As many as four tiny spherical organoids are kept in one oblong “processing unit” that keeps them protected and nourished. Through the attached electrodes, the organoids can communicate with conventional computer networks.
Video by b/60:
Why Create Organoid Computers?
At first glance, one wonders what benefit such living computers would have, especially given that new technologies like quantum computers are rapidly advancing. For instance, scientists are mimicking natural molecules to vastly improve quantum computers.
For the time being, organoid computers are also short-lived and take at least a month to cultivate in a lab. Yet even so, the scientists believe that one day, they can potentially get them to “live” much longer and grow much larger (see more below).
The ultimate advantage is that these cells need only a tiny fraction of the energy of silicon-based computing. Nature has made us extremely efficient in this regard over hundreds of thousands of year. And in a world where AI data centers are threatening to consume 29.3 terawatt-hours per year all by themselves, lower energy consumption is highly attractive. Instead of AI, it’s called BI, or Bioorganoid intelligence, but the differences are much more than mere energy consumption.
Besides extremely low energy needs, the cells can be grown in a lab to replace those that “die.” Another advantage is that computations take place in rapidly complex ways similar to our organic brain’s neural networks.
Wetware Computers
A pioneer in this new Biocomputing field is Swiss company FinalSpark, with their “Neuroplatform.” Right now, researchers and universities around the world are renting their platform for $500 a month over the Internet. In all, over 3,000 organoids are currently stored in their facility.
You can see an example of the FinalSpark system working live on their website.
Other scientists have created similar “wetware” computers using mushrooms or fungus. As it turns out, the same thin, threadlike fungal mycelium highways that enable communication between trees can also work in computers without the same ethical concerns as with human organoids.
Related: Could Biocomputers of Human Brain Cells Develop Consciousness?
Video by Space Science Guy about the ethical concerns of these new “living computers”:
Will Organoid Computers Become Sentient?
Of course, creating computers with human cells raises all kinds of ethical considerations. As they inevitably grow more complex, will they achieve a rudimentary consciousness? And as they grow in size, where does that all lead to?
Related: Could Synthetic Human Embryos Have the Spark of Life and Consciousness?
For insight, neurophysiologist Christof Koch argues (see video below) that brain organoids are conscious while AI is not, based on Integrated Information Theory (IIT). According to his view, AI is like a “vampire” of human knowledge.
He hilariously described AI using LLMs (long language models) as “gigantic vampires that feast on humanity’s collective creativity over the last few thousand years and just suck up everything and then somehow regurgitate that in very interesting, very surprisingly adept patterns.”
Although AI is good a mimicking humans, if you look “under the hood,” he notes it’s radically different from a human brain. While AI can outperform us in many areas, AI can’t feel anything, he says. Contrast this view with Google engineer and futurist Ray Kurzweil, who predicts that AI will pass the Turing Test (pass for a human being) by 2029.
Future Organoids the Size of Whales?
So far, brain organoids must be small since tissue engineers aren’t able to deliver a blood supply to the larger ones. However, Koch suggests scientists will eventually figure out how to supply them by building larger vascular structures. At that point, the organoid could be the size of “a peanut butter cookie” or, eventually, the size of a monkey’s brain or larger.
“Once you can build them, you can build them to any size, more like a whale, even bigger than a whale,” he said (see video below).
As the organoids become larger and more complex, he argues that they will be capable of “experiencing something,” a rudimentary level of consciousness.
“If it has any sort of of visual input or tactile input, it will certainly begin to form conscious experience of space and time.”
If the organoid is given a way to move around, then it is able to “act in the world” with some kind of intelligence and simple emotions. It sounds far-fetched, but in 2018, scientists from the University of California San Diego School Of Medicine were conducting a project intended to give crab-like robots a “minibrain” made with Neanderthal DNA. Yes, really. They were also reportedly pitting them against human-brained robots.
Will AI one day find itself pitted against BI, the organic versus the artificial? It seems we are well on our way to finding out, like it or not.
Video about the subject of consciousness and brain organoids by Ihm Curious:
Featured image via YouTube/b/60