For years humans have aimed at improving organs using machines. But what if it was the other way around - we replace technology with organs. Specifically, what if we replaced the silicone-based computing chips with lab-grown brain organoids, also known as bioprocessors?
Multi-electrode arrays used for Neuroplatform organized into four sets of eight electrodes.
FinalSpark, a Switzerland-based biotechnology company, has created a neuro processor where hardware, software, and biology come together. It is composed of 16 lab-grown brain organoids, which can communicate and exchange information with each other via neurons, and a setup called a multi-electrode array (MEA). In this setup, each set of four organoids interfaces with eight electrodes that stimulate and record the data processed by the brains. Data transfer is done using analog-digital converters and a microfluidic system provides life support to the brain organoids.
The organoids are trained using a dopamine reward system, similar to how human brains learn. Upon completion of a task, light is exposed to a specific region of the brain to trigger dopamine release. In order to complete a task, they use a measly 20 watts of electricity. To put this into perspective, the best computers in the world at the same speed use 21 megawatts (21 million watts), making it a million times more efficient! Additionally, they have 1,000 times more memory than modern high-end computers.
Not only is the neuro processor extremely energy efficient, but it also has other functionalities. When the brain learns, its hardware changes, creating new synaptic connections. With this neuro processor, FinalSpark hopes to create artificial intelligence capable of passing the Turing test, a test to determine whether a machine can think like a human.
Though this project is in its early stages, it shows significant promise in being a solution to climate change, given how much less energy it uses compared to traditional computers.
Bibliography
Comments