Scientists Merge Lab-Grown Human Brain Tissue with Electronics for AI Algorithm Studies
In a study published in the journal Nature Electronics, a group of scientists extensively describes a "hybrid biocomputer" that combines lab-grown human brain tissue with conventional circuits and artificial intelligence algorithms. The system, named Brainoware, has been trained to recognize voices with an accuracy of 78%, as reported by Engadget.
Brainoware integrates brain organoids – clusters of human cells derived from stem cells and transformed into a "mini-brain" filled with neurons – with conventional electronic circuits. Researchers placed one organoid on a plate with thousands of electrodes to establish a connection between the brain and electrical circuits. Electronic circuits, interacting with the organoid, convert the information they want to convey into electrical impulses.
The brain tissue then learns and "communicates" with the electronics. A sensor in the electronic array captures the reaction of the mini-brain, deciphered by a trained machine learning algorithm. In essence, through artificial intelligence, neurons and electronics merge into a unified (albeit currently simple) biocomputer.
Scientists trained Brainoware on 240 recordings of conversations involving eight individuals, translating sound into an electrical signal to transmit to the organoid. The organic part reacted differently to each voice, generating a neural activity pattern that AI learned to understand, achieving a 78% accuracy in voice recognition.
Researchers view this work more as a proof of concept, allowing the integration of electronic and biological elements for computer processing. They express doubt about the immediate practical application of this technology, although it holds potential. Previous similar studies demonstrated how two-dimensional neuron cultures could perform comparable functions, but Brainoware represents the first experimental prototype using a three-dimensional cluster of human brain cells.
The research suggests the possibility of future biological computing, where the "speed and efficiency of the human brain" could serve as a foundation for creating powerful artificial intelligence. Additionally, this technology enables a better understanding of real human brain reactions, potentially advancing the treatment of conditions such as Alzheimer's disease.
In related news, The Gaze reported earlier that the artificial intelligence system MIA will be integrated into the curriculum of French schools. The system will be available as an app for second-year high school students starting September 2024.