Questneers : Daeyeol Lee (Johns Hopkins University), Se-Bum Paik (KAIST)
Current brain engineering technology has reached the stage where it can partially restore function by electrically stimulating the brain when sensory or motor-related brain functions are damaged. However, higher cognitive functions such as memory and learning are difficult to approach technically because their operating mechanisms are not sufficiently understood due to the complexity of neural activity. By deeply understanding brain function, can we go beyond treating cognitive disorders through machine devices developed by humans to control various higher cognitive functions, and ultimately implement brain-reality superior to virtual reality?
Since the mid-20th century, scientists have discovered that the brain operates by transmitting various electrical and chemical signals between neurons, and research to understand the sensory, motor, and cognitive processes occurring in the brain has been actively conducted based on this. Particularly, after it was revealed that brain cognitive functions, which were considered like a black box, are also physical phenomena caused by electrical signals, the theoretical possibility of controlling higher cognitive functions occurring in the brain began to be mentioned. Based on this possibility, research on neuroprosthetics to treat brain dysfunction and more general brain-machine interaction (BMI) began in earnest. As a result, attempts to overcome sensory organ damage through implantation of mechanical devices, such as cochlear implants and artificial vision, are already being implemented in reality, and recently research has been focused on decoding more complex neural activities to decode cognitive information in the brain.
Using functional magnetic resonance imaging (fMRI), which is currently widely used, it is possible to measure patterns of brain metabolism activation. Recently, based on such technology, great progress has been made in understanding brain functions that process sensory information and make decisions. These attempts are becoming the foundation of brain-machine interface technology, but to more accurately grasp the information processed by the brain, it is necessary to precisely measure changes in action potentials created by individual neurons. For example, Elon Musk’s widely known Neuralink project has proven that it is possible to connect the brain to external devices by inserting high-resolution electrodes into the brain and exchange neural signals in real time. In fact, experiments have succeeded in having monkeys and humans control console games just by thinking.
Ultimately, if technology that can artificially create cognitive functions by stimulating the brain externally through mechanical devices is developed, a new era is expected to open. If this technology becomes reality, it can expand beyond simply treating brain sensory disorders to treating brain diseases in higher cognitive domains, and furthermore to general cognitive ability enhancement. For example, if we can understand sensory information processing mechanisms and enhance them with machines, we could implement brain-reality of a completely different dimension from current virtual reality driven through sensory organs. Also, if technology to “download” cognitive functions into the brain is realized through this, it could provide innovative learning methods that allow immediate use of new languages or skills. Thus, cognitive function control through machines, which was previously dealt with in science fiction, can become reality.
However, despite many studies so far, it is still impossible to completely control brain functions electrically. The biggest reason is that the neural mechanisms in the brain are too complex. Like the research by Hubel and Wiesel in the 1950s that revealed the roles of simple cells and complex cells in the primary visual cortex, understanding of early sensory and motor information processing such as vision, hearing, and movement has gradually developed. However, as we move to higher regions where higher cognitive functions are processed, the complexity of brain signals increases exponentially. Also, brain regions are connected in complex multiple layers, so there are limitations to the reductionist approach of simply analyzing each part. These complex system characteristics of the brain make it difficult to understand and manipulate its principles.
Nevertheless, if technology to measure and transmit high-fidelity signals can be developed and brain plasticity and self-organizing principles can be perfectly understood and utilized, cognitive function enhancement through electrical manipulation will be possible. This is because even without precisely understanding individual neural circuits of the brain, brain adaptability such as plasticity and self-organization can be used to induce the brain to learn and adapt to artificial stimuli on its own. If signals similar to the information capacity of biological sensory systems can be transmitted, breakthrough technology that can naturally adapt to artificial stimuli and experience new brain-reality could be realized.