We're not close to replicating the complexity and intricacy of the human brain with anything artificial, but scientists are making progress with certain dedicated devices.
A structure that mimics the human brain can be used to make up neural networks.
The device can process information a million times faster than the brain.
The artificial synapse is intended to be used in a way that improves speeds while reducing energy use, which is important for affordability as well as the demands on the planet's natural resources.
The use of a specially selected and efficient material is the key to the significant improvements. The team behind the project believes the gains in neural network learning speeds will be significant.
A computer scientist from the Massachusetts Institute of Technology says that once you have an analog processor, you will no longer be training networks other than your own.
You will be training networks that are so complex that no one else will be able to match them. This is not a fast car and it is not a spaceship.
The material in question is based on a type of glass called psg. Its nanoscale pores allow protons to pass through it at never-before-seen speeds when it's used as a solid electrolyte.
The same fabrication techniques that are used to make Silicon can be used to make PSG. It should be easier to integrate into existing production processes.
The flow of signals and other information in the brain can be controlled by weakened or strengthened synaptic connections. The same effect can be achieved if you control the movement of protons. It is fast, reliable, and can all operate at room temperature, making it more practical.
Onen said the speed was surprising.
We wouldn't apply such extreme fields across devices in order to not make them ash. The protons ended up moving at a million times faster than before.
The movement has a small size and low mass of protons. It's almost like going back in time.
There is a lot of potential for using less energy. In order to create a neural network, resistors would be stacked together in a chess board style array, which can be operated in parallel to improve speeds.
The researchers will have to adapt the prototype so it can be produced on a bigger scale. The team is confident that it can be accomplished.
The end result would be an artificial intelligence system that can do things such as identify what's in an image or speak into a microphone.
Artificial intelligence can learn by analyzing a lot of data. Self-driving cars, medical image analysis, and other related fields are all covered by that.
It will be possible for these resistors to be embedded into actual systems, and to overcome potential performance problems that limit the voltage that can be applied.
The path forward is going to be very challenging, but at the same time, it is very exciting.
The research has appeared in a journal.
All Rights Reserved © 2024
