Can living brain cells really replace silicon chips? Scientists in Switzerland are building mini human brains to test a radical new era in biocomputing and AI.
 |
| Swiss researchers are using lab-grown mini-brains to develop wet-ware computers that learn and respond like AI — but are made from living human cells. Image Courtesy: BBC/ CH |
LAUSANNE, Switzerland — October 4, 2025:
Could the future of computing be made from living brain cells? In a breakthrough that feels more science fiction than science lab, researchers in Switzerland are pushing the limits of what’s possible in biocomputing — using lab-grown mini human brains, or organoids, to create biological computers that think, learn, and adapt.
Inside the FinalSpark lab in Lausanne, teams of biologists and engineers are growing tiny clusters of neurons from human skin-derived stem cells. These living organoids are then connected to electrodes, allowing them to receive and respond to electrical signals — much like how artificial intelligence systems process input. This fusion of biology and computing has given rise to a new term in tech: wetware.
“We’re building computers out of human brain tissue,” said Dr. Fred Jordan, co-founder of FinalSpark. “It’s not just about energy efficiency. It’s about creating machines that learn more like we do.”
The potential energy savings are massive. Today’s AI models require vast data centers and enormous power consumption. Wetware computers, by contrast, could use only a fraction of the energy — and offer a far more organic learning process.
The process starts with reprogrammed skin cells, which are transformed into neural stem cells and grown into spherical organoids. These are not true brains — they lack structure, consciousness, and blood vessels — but they contain the building blocks of neural activity.
After several months of cultivation, the organoids are placed on electrode grids and tested with simple commands. In a live demo, Dr. Jordan showed how pressing a key sends a pulse through the system, prompting the organoid to react. The responses, displayed on screen, resemble an EEG graph.
Sometimes, the system doesn’t respond. Other times, it shows unexpected bursts of activity — including one dramatic flare after repeated commands. “Maybe I annoyed it,” Jordan joked. But the truth is, researchers don’t fully understand how or why these living systems behave the way they do.
The ultimate goal is to train the organoids to perform tasks — recognizing patterns, making decisions — just like AI does with image recognition or natural language processing.
Unlike traditional computers, biocomputers are fragile. “Organoids don’t have blood vessels,” explained Prof. Simon Schultz, Director of the Center for Neurotechnology at Imperial College London. “Keeping them alive for long periods is still one of the biggest technical challenges.”
Currently, FinalSpark’s organoids survive up to four months. Their deaths are often marked by a final burst of electrical activity — a strange echo of what’s observed in dying human brains. “We’ve recorded 1,000 to 2,000 of these moments,” Jordan said. “It’s always a setback — we stop, analyze, and start again.”
But despite the emotional undertones, scientists stress that these are not conscious beings. “They're not sentient. They’re just a different type of computer,” Schultz said.
The Swiss team is not alone in this work. In Australia, Cortical Labs taught artificial neurons to play Pong. In the United States, Johns Hopkins University is using mini-brains to model neurological disorders like Alzheimer’s and autism — helping researchers test new treatments without using animals.
Still, experts like Dr. Lena Smirnova of Johns Hopkins believe that biocomputing is not a replacement for silicon-based systems, but a complementary tool. “It’s exciting, but we’re still in the early stages,” she said.
Even its most ardent advocates agree. “Silicon will always have strengths,” Schultz added. “But wetware could find unique niches — especially in medical modeling or adaptive systems where biology has the edge.”
For Jordan, the fascination lies in the unknown — and in watching a sci-fi dream come to life. “I used to read books and wish life was like that. Now, I feel like I’m writing one,” he said.
As organoids grow, respond, and slowly learn, they’re not just forming the foundation of a new kind of computer — they’re challenging our assumptions about intelligence, life, and what it means for machines to think.
In the labs of Lausanne, the next revolution in computing may not be built with code — but cultured in a dish.