In a laboratory in Vevey, Switzerland, a startup called FinalSpark is operating computers whose processors are alive. The company grows small clusters of human brain cells, known as organoids, connects them to electrodes and rents remote access to researchers who want to test whether living neurons can process information. The approach is called wetware computing, a term coined by analogy with hardware and software to describe biological neural tissue used as a computing substrate.
How the system works
FinalSpark’s co-founder Fred Jordan and his team produce the organoids from human stem cells, which are derived from donated skin cells and coaxed into becoming neurons. Within each organoid, the cells form connections and communicate through electrical signals, as neurons do in a brain. The organoids sit on arrays of electrodes that can both stimulate the tissue and record its responses, allowing software to send inputs and read outputs.
In a 2024 paper in the journal Frontiers in Artificial Intelligence, the company described its Neuroplatform, an online system through which scientists anywhere can connect to the organoids and run experiments using Python code. According to the paper, the organoids survive for roughly 100 days, about four months, during which researchers can study how they adapt to stimulation. FinalSpark says this is the first open, remotely accessible platform of its kind.
The energy argument
The central promise of biocomputing is efficiency. Training and running large AI models consumes enormous amounts of electricity, and the companies in this field argue that biological neurons perform computation at a tiny fraction of the energy cost of silicon. FinalSpark has claimed its living processors could use orders of magnitude less power than conventional chips for some tasks. That remains a projection rather than a demonstrated result: current organoid systems are far from matching the reliability or scale of digital computers.
A small but active field
FinalSpark is not alone. Cortical Labs in Melbourne, Australia, drew attention in 2022 when its DishBrain system, a layer of neurons on a silicon chip, learned to play a simplified version of the video game Pong, and in 2025 it began shipping a commercial biological computer called the CL1 built around roughly 200,000 living human neurons. At Johns Hopkins University, researchers in the emerging discipline of organoid intelligence are using similar brain organoids to study conditions such as Alzheimer’s disease and autism, as well as to explore their computational potential.
Ethical questions
The work raises questions that go beyond engineering. Brain organoids are far simpler than a human brain and show no evidence of consciousness, but scientists in the field acknowledge that as the tissue grows more complex, the question of whether it could experience anything will become harder to dismiss. Some researchers have reported bursts of electrical activity in organoids shortly before they die, and whether such patterns carry any significance is debated. Ethicists have called for oversight frameworks to be developed before the technology advances further.
Wetware computing is at an early stage, and its practical uses are so far limited to research. But it marks a point at which the boundary between biological and artificial systems is becoming less distinct, and the field’s progress will be watched closely by both AI developers seeking more efficient hardware and by ethicists concerned about what living processors might one day become.
