The petri dishes shown in the image contain mini biological brains — and a Swiss company has connected them to electrodes to use them as computers.
---
Modern computers are based on the venerable von Neumann architecture, which separates the processor, working memory, and long-term storage.
Transferring data between these components consumes significant energy and is considered a major bottleneck.
Compared to von Neumann computers, the human brain easily outperforms every existing machine in terms of processing power per unit of energy — at least on certain tasks.
The brain stores and processes information using the same cells and neurons, and its learning process is based on trial and error, delivering far greater efficiency than any computer.
The brain's reward system releases dopamine when we do something right, reinforcing the connections between the neurons that were involved in executing the correct decision just moments before.
---
The company is called FinalSpark, and its goal is to build mini computers based on biological brain cells that learn to perform specific tasks in a manner similar to humans, while interfacing with classical computing systems.
The process begins with ordinary skin cells taken from a person, which are reprogrammed into stem cells and then placed in a mold and growth environment that causes them to proliferate into brain cells.
The brain cells receive a steady supply of nutrients, temperature control, and conditions suitable for sustaining life, and are connected to a set of electrodes that carry input and output signals to and from a classical computer system.
The nutrient medium contains dopamine stored in chemical microcapsules. When the mini-brain performs a correct action, it is illuminated with ultraviolet light, which breaks down the capsules and releases dopamine — reinforcing the neural connections.
---
The system is connected to the internet, allowing researchers from around the world to send Python commands to the system and examine the results.
At present, the system is most useful for studying the effects of new drugs on human brain cells, eliminating the need to conduct experiments on humans or animals.
In the longer term, if the experiments succeed, the technology could mature into a commercial product — enabling us, for example, to equip robots or autonomous vehicles with navigation and vision systems comparable to those of humans.
In the age of artificial intelligence, such systems could prove essential, since their energy consumption is roughly one-millionth that of today's power-hungry computing hardware.
---
From an ethical standpoint, it is important to remember that although we are dealing with a small cluster of biological brain cells, they are still far from resembling the complex structure of our brain and are entirely devoid of self-awareness.
Progress in this field toward a brain capable of understanding and feeling would raise profound ethical questions — but the prospect of reaching that point technologically still appears to be a long way off.
---
👋 Hi, I'm Shlomo Strauss — follow me for more fascinating content on science and technology.
Image: FinalSpark