A Korean research team has developed an innovative semiconductor technology, capable of simulating the flexibility of the human brain.
On September 28, the Korean Institute of Science and Technology announced that a research team led by Professor Kyung Min Kim of the Department of Science and Engineering at the Advanced Korean Institute of Science and Technology in South Korea has developed a “nervous messenger with frequency replacement” that mimics the bodies of nerve cells.
The team’s results were published earlier in the Advanced Matellarls magazine, and the Yurrick Alert website was written about.
The role of the human brain is not limited to regulating nerve clamps that exchange signals only; Rather, individual neurons also treat information through the self -pillar.
Self -plasticity indicates the brain’s ability to adapt as neurons become more or less sensitive in a context, for example, to become less discomfort when hearing the sound itself frequently, or to respond faster to a specific catalyst after frequent training.
The current artificial intelligence conductors face difficulty in simulating this elasticity of the brain.
The nerve impulses resistant to switching frequency, which is an artificial nervous system, automatically frequency, just as the brain becomes less discomfort than repeated stimuli, or on the contrary, more sensitive through training.
Less energy and high efficiency
The research team combined “volatile memory resistance”, which interacts instantaneously before returning to its original state, and “unparalleled memory resistance”, which remembers the signals for long periods. This enabled the development of a device that can control the number of times the nerves fired its sudden signals.
This device, inside a single connector device, reproduces how the brain discomfort is decreased from frequent sounds or increased its sensitivity to repeated stimuli.
The device showed excellent elasticity, even if some nerve cells are damaged, the network’s self -broken plasticity allowed to reorganize itself and restore its performance.
Artificial intelligence that uses this technique consumes less energy while maintaining performance, and can compensate for partial malfunctions in electrical circuits to resume natural operation.
Professor Kyung Min Kim, who led the research, stated: “This study has applied the self -blog, a basic function for the brain, in a single semiconductor device, which strengthened energy efficiency and the stability of artificial intelligence devices to a new level. This technology, which enables devices to remember their condition and adaptation or recovery even from damage, can be a key element in systems that require long -term stability, such as Peripheral computing and self -leadership.