As firms collect more data to enhance how AI identifies photos, learns languages, and does other difficult jobs, a recent research outlines a method for computer chips to dynamically reorganize themselves to take in new data, similar to how the brain does, allowing AI to learn over time.
The human brain adjusts as it learns anything new. When artificial intelligence learns anything new, though, it has a tendency to forget what it already knows.
As companies collect more data to improve how AI recognizes images, learns languages, and performs other complex tasks, a paper published this week in Science demonstrates how computer chips could dynamically rewire themselves to take in new data, similar to how the brain does, allowing AI to learn over time.
“Living beings’ brains may continue to learn throughout their lives. We’ve now built an artificial learning platform for robots to use throughout their lives “Shriram Ramanathan, a professor at Purdue University’s School of Materials Engineering who studies how materials might imitate the brain to better computers, agreed.
The circuits on a computer chip do not change, unlike the brain, which continually establishes new connections between neurons to facilitate learning. A circuit that has been used by a machine for years is identical to the circuit that was created for the machine in the factory.
This is a difficulty when it comes to making AI more portable, such as for autonomous cars or space robots that must make choices on their own in remote locations. These devices would be more efficient if AI could be incorporated directly into hardware rather than operating on software, as it now does.
Ramanathan and his colleagues developed a novel piece of hardware that can be reprogrammed on demand using electrical pulses in this investigation. Ramanathan thinks that the device’s versatility will enable it to do all of the activities required to create a brain-inspired computer.
“If we want to develop a computer or a system that is inspired by the brain, we need to be able to program, reprogram, and modify the chip on a constant basis,” Ramanathan said.
Towards the creation of a chip-based brain
The hardware is a tiny rectangular device built of perovskite nickelate, a material that is very sensitive to hydrogen. The gadget can shuffle a concentration of hydrogen ions in a matter of nanoseconds by applying electrical pulses at varying voltages, generating states that the researchers discovered could be mapped out to corresponding activities in the brain.
When there is more hydrogen towards the device’s core, it may function as a neuron, or a single nerve cell. The gadget acts as a synapse, a link between neurons, which the brain utilizes to store memories in complicated neural circuits, with less hydrogen at that place.
The Purdue team’s collaborators at Santa Clara University and Portland State University showed that the internal physics of this device creates a dynamic structure for an artificial neural network that can recognize electrocardiogram patterns and digits more efficiently than static networks through simulations of the experimental data. This neural network employs “reservoir computing,” which describes how various areas of the brain connect and exchange data.
In this study, researchers from The Pennsylvania State University revealed that when new issues arise, a dynamic network may “choose and choose” which circuits are most suited to solving them.
Ramanathan expects that the semiconductor industry would quickly embrace this technology since the team was able to manufacture the device using regular semiconductor-compatible manufacturing procedures and run it at ambient temperature.
Michael Park, a Purdue Ph.D. student in materials engineering, stated, “We proved that this gadget is quite resilient.” “After a million cycles of programming, the device’s reconfiguration of all functions is impressively repeatable.”
The researchers are working on large-scale test chips that will be utilized to create a brain-inspired computer to showcase these notions.
Experiments were carried out at Purdue’s Discovery Park’s FLEX Lab and Birck Nanotechnology Center. The device’s characteristics were measured by the team’s partners at Argonne National Laboratory, the University of Illinois, Brookhaven National Laboratory, and the University of Georgia.
The National Science Foundation, the US Department of Energy’s Office of Science, and the Air Force’s Office of Scientific Research also contributed to the study.