In a light-filled lab in Mission Bay, a few blocks from the water, two neurosurgeons are betting on something that sounds almost heretical in the age of GPUs: that living brain cells may one day outcompute silicon.
Dr. Alexander Ksendzovsky and Dr. Jon Pomeraniec founded The Biological Computing Company with a bold premise. Instead of etching ever-smaller transistors into wafers, they are cultivating networks of human neurons and asking them to compute.
The idea is not science fiction. It is an extension of what biology already does. The human brain runs on about 20 watts of power. It learns continuously. It adapts. It rewires. Silicon systems, by contrast, consume vast energy and must be trained on massive datasets before they generalize. The founders’ argument is simple and provocative: evolution has already solved many of the problems we struggle with in artificial intelligence.
In practice, the company grows living neuronal cultures derived from stem cells. These cells self-organize into networks. They form synapses. They exhibit spontaneous electrical activity. Placed on multi-electrode arrays, the networks can both receive input and generate measurable output. In other words, they can be stimulated and their responses recorded in real time.
The ambition is to build a hybrid system where biology and electronics speak to each other. Electrical signals from a computer feed into the neuronal network. The network processes that information according to its own intrinsic dynamics. The output is read back into software. Training, in this context, means shaping patterns of stimulation so that the living network gradually modifies its connectivity and improves performance on specific tasks.
What makes this compelling is not raw speed. It is efficiency and plasticity. Neurons operate through electrochemical gradients, not clock cycles. They encode information in timing, frequency, and synchrony. They adapt structurally. A silicon chip does exactly what it is etched to do. A neuronal network can reorganize itself.
There are scientific and ethical complexities, of course. These are not conscious brains. They are cellular networks, closer to organoids than to minds. But they display properties that traditional hardware does not. They are noisy. They are dynamic. They forget and relearn. Those traits, once seen as liabilities, may prove advantageous in certain forms of pattern recognition and adaptive control.
The company has raised significant early funding to explore this space, positioning itself at the intersection of neuroscience, synthetic biology, and computing. It sits alongside a small but growing field sometimes referred to as “organoid intelligence” or wetware computing. The premise is that computation does not have to be purely digital. It can be biological.
Photographing Ksendzovsky and Pomeraniec in that Mission Bay lab, what struck me was not hype but clinical focus. Both are practicing neurosurgeons. They have spent years operating on the brain, seeing firsthand its fragility and resilience. Their startup feels less like a moonshot and more like an extrapolation. If the brain is the most efficient computing system we know, why not study it not just as an organ, but as hardware.
There are open questions. Scalability. Reproducibility. Standardization of living systems. Long-term stability of neuronal cultures. Integration with conventional machine learning pipelines. And the deeper philosophical question: if we harness biological substrates for computation, what does that say about the boundary between machine and organism?
For now, The Biological Computing Company is exploring first principles. Can living neural networks be trained to perform useful computational tasks. Can they do so with lower energy cost than GPUs. Can they adapt in ways that silicon cannot.
It is early. The lab benches are still crowded with incubators, microscopes, culture plates. But the wager is clear. The next revolution in computing may not be smaller transistors. It may be cells in a dish, firing, learning, and responding in ways that feel strangely familiar.
And in Mission Bay, under fluorescent lights and the quiet hum of incubators, that future is already pulsing.
