Biological AI and Real Neurons on Silicon

Have you heard about biological AI where real human neurons grown from stem cells are used instead of artificial neurons?

This is no longer science fiction. These cells are grown on a silicon chip, information is sent to them through electrical impulses, their response is read, and developers can even access them through an API.

When I first heard about it, I thought: what if this is a replacement for ordinary AI? I started digging and realized that everything is much more interesting and at the same time much more complicated.

Biological neurons will not replace GPT, graphics cards, or data centers. They transmit signals more slowly, and you cannot simply copy them with Ctrl-C and Ctrl-V. They must be grown, fed, kept at the right temperature, and given time to learn. And they die.

But biology has one fundamental advantage. For millions of years it learned to adapt to an unfamiliar environment and survive. A cell enters new conditions and starts looking for a way to adjust. A digital model often needs huge datasets and many examples. A biological system can learn from a much poorer signal and live feedback.

And this is where it becomes truly interesting.

It seems to me that biological AI is not needed for calculations. Its potential place is in the physical world: robotics, vision, movement, touch, and orientation in an unfamiliar environment. Digital AI will handle memory, calculations, and planning. The biological part may handle adaptation to a reality that cannot be completely described by a dataset in advance.

And this can already be touched with your hands. Cortical Labs created the CL1 and a cloud platform through which a developer can send stimuli to living neurons, read their activity, and build experiments with ordinary code.

After that begins a territory where there are still more questions than answers. At what point does a network of cells become something more than a biological computer? Can it experience discomfort? Where is the boundary between an experiment and consciousness? What rights might such systems have?

And the most striking thing here is not even possible cyborgs in ten years. The most striking thing is that the basic infrastructure for such experiments already exists today.

The longer you live, the more surprised you become. We are already inside science fiction. We simply do not always have time to notice it.