ISTANBUL\u00a0<\/p>\n
Swiss scientists are attempting to produce a biocomputer from living cells, an idea that dates to science fiction, the BBC reported Saturday.<\/p>\n
According to the report, scientists are working toward creating data centers filled with living servers that replicate artificial intelligence while consuming far less energy. The efforts are paving the way for what researchers call \u201cwetware,\u201d a new component in the computing world alongside software and hardware.<\/p>\n
One of the pioneers of the vision is Fred Jordan, co-founder of the FinalSpark laboratory, which leads research in the sector.<\/p>\n
\u201cWhen you start to say, \u2018I am going to use a neuron like a little machine,\u2019 it gives you a different view of our own brain and makes you question what we are,\u201d said Jordan.<\/p>\n
He acknowledges that the idea of \u201cbiocomputers\u201d may seem strange to many, stressing that at its core it remains a concept born of science fiction.\u00a0\u00a0<\/p>\n
Turning human cells into organoids for biocomputing<\/p>\n
Obtaining stem cells from human skin cells represents a first step toward FinalSpark\u2019s larger goal of triggering learning in the biocomputer\u2019s neurons so they can eventually carry out tasks, said the report.<\/p>\n
Skin cells sourced from certified providers are cultured into living stem cells, which are then developed into tiny lab-grown brain-like spheres known as \u201corganoids.\u201d Researchers emphasize that while organoids are far less complex than the human brain, they share the same fundamental building blocks.<\/p>\n
After a process that can take several months, the organoids are connected to electrodes and asked to respond to simple keyboard commands. When an organoid receives an electrical signal, a small spike of activity appears on a moving graph, resembling an EEG trace on a computer screen. In that way, the transmission and reception of electrical signals can be verified, with results recorded on the computer.<\/p>\n
By tracking electrical stimulation and addressing unanswered questions, the aim is to advance the learning capacity and command-response systems of the living cell-based computers.<\/p>\n
\u201cFor AI, it\u2019s always the same thing,\u201d Jordan explained. \u201cYou give some input, you want some output that is used. For instance, you give a picture of a cat, you want the output to say if it\u2019s a cat.\u201d\u00a0\u00a0<\/p>\n
Survival problem of biocomputers<\/p>\n
As stated in the report, one of the biggest unanswered questions for biocomputers is what their source of energy will be. While an ordinary computer can simply be kept running with a power supply, the case is entirely different for biocomputers, which lack even the blood vessels needed to nourish themselves.<\/p>\n
\u201cWe don\u2019t yet know how to make them properly. So this is the biggest ongoing challenge,\u201d said Simon Schultz, professor of neurotechnology and director of the Center for Neurotechnology at Imperial College London.<\/p>\n
But thanks to FinalSpark\u2019s work over the past four years, organoids can now be sustained for up to four months, said the report.\u00a0\u00a0<\/p>\n
Similar efforts underway worldwide<\/p>\n
The report highlights that alongside FinalSpark, several other laboratories are pursuing similar efforts in this field.<\/p>\n
For example, in 2022, the Australian company, Cortical Labs, announced that it produced artificial neurons capable of playing a computer game.<\/p>\n
Meanwhile, researchers at Johns Hopkins University in the US are working to build \u201cmini brains\u201d to shed light on how information is processed, to develop drugs for conditions such as Alzheimer\u2019s and autism.<\/p>\n
Lena Smirnova, who leads the research, said that \u201cwetware\u201d is still in its early stages and the likelihood of it replacing the main material used in computer chips is currently very low.<\/p>\n
\u201cBiocomputing should complement \u2014 not replace \u2014 silicon AI, while also advancing disease modeling and reducing animal use,\u201d she explained.<\/p>\n
The report notes that artificial intelligence may soon play a role in accelerating such efforts.<\/p>\n