Artificial Intelligence systems are gradually beginning to penetrate into all areas of life around us; elements of such systems are used in software, in computer games, they are already writing blog posts, translating from one language to another, writing program source code themselves, and even developing computer chips. It looks like nothing can stop the "triumphal march" of artificial intelligence. But researchers at Johns Hopkins University think otherwise. In fact, they have already created one thing that may be a good alternative to computer artificial intelligence in the long run.
That thing is so-called "organoid intelligence" (OI), which works on living biological neurons organized as artificially grown brains. These clusters of nerve tissue, organoids, are grown from stem cells, and their structure is optimized to actuate algorithmic systems. By creating such an incredible thing, researchers from Johns Hopkins University aim to simplify deep learning technologies and improve the quality of feedback, which should have a better effect on the decisions made by intelligence.
At present, scientists have already created the first prototype of a biocomputer, the core of which is a laboratory-grown organoid. This organoid is shaped like a ball, which allows it to form a greater number of connections, synapses, than would be possible with a flat section of nerve tissue.
And now there is intense work underway, the ultimate goal of which is the technology to "communicate" with the organoid, allowing it to transmit information to the artificial brain, to see and recognize what it is "thinking. To realize this goal, scientists use a neurocomputer interface they developed back in 2022. The flexible shell of this interface has a matrix of tiny electrodes that pick up signals from the organoid and transmit control signals to it.
The artificial brains scientists now have, the organoids, are too small to realize full-fledged "organic intelligence. These organoids contain approximately 50 thousand neurons each, and to realize intelligence requires at least 10 million neurons, between which 125 trillion synapses can be organized, each of which can store 4.7 bits of information.
However, the use of living human brain cells as parts of biocomputers also raises a number of obvious moral and ethical issues. One such issue is the protection of the privacy of human stem cell donors, and another, and no less important, is that organic intelligence may someday gain self-awareness and begin to feel pain and suffering. And, before starting to develop real organic intelligence, people will need to develop a set of norms and rules governing how to relate to all of this.
