Tech
Artificial Brain Blueprint Outlined by Scientists: Constructed Synapses May be Key
Catherine Griffin
First Posted: Feb 26, 2013 11:28 AM EST
Artificial intelligence in computers is closer than ever. Using electronic microcomponents that imitate natural nerves, scientists have constructed a memristor that is capable of learning, and may pave the way for an artificial brain.
The findings, which will be published in the beginning of March in the print edition of the Journal of Physics, examined the use of mesristors as components for a larger, artificial brain. Memristors are nanocomponents that are made of fine nanolayers and can be used to connect electric circuits; they're essentially the artificial version of synapses in the brain. Synapses are the bridges that nerve cells (neurons) use in order to contact each other. Their connections increase in strength the more often they're used and are able to process more quickly. Like synapses, memristors learn from earlier impulses. Currently, these impulses come from the circuits that they're connected to and are electrical in nature.
It's not surprising that the scientists involved in this work believe that these artificial synapses can be used to create an artificial brain. The study is the first of its kind to summarize exactly what principles from nature need to be transferred to technological systems in order to make the "brain" function.
In particular, the memristors could help in that goal. They store information more precisely than the bits on which previous computer processors have been based. If an artificial brain were created, it could make for a more efficient computer that could process data more quickly.
Andy Thomas, the lead researcher involved in the study, said in a press release, "They allow us to construct extremely energy-efficient and robust processors that are able to learn by themselves."
This artificial brain hasn't been created yet, but Thomas and his colleagues are hopeful for the future. Their paper outlines exactly what needs to happen before it's made, and could allow scientists to design a more efficient computer that we could see in future years.
See Now:
NASA's Juno Spacecraft's Rendezvous With Jupiter's Mammoth Cyclone
©2024 ScienceWorldReport.com All rights reserved. Do not reproduce without permission. The window to the world of science news.
More on SCIENCEwr
First Posted: Feb 26, 2013 11:28 AM EST
Artificial intelligence in computers is closer than ever. Using electronic microcomponents that imitate natural nerves, scientists have constructed a memristor that is capable of learning, and may pave the way for an artificial brain.
The findings, which will be published in the beginning of March in the print edition of the Journal of Physics, examined the use of mesristors as components for a larger, artificial brain. Memristors are nanocomponents that are made of fine nanolayers and can be used to connect electric circuits; they're essentially the artificial version of synapses in the brain. Synapses are the bridges that nerve cells (neurons) use in order to contact each other. Their connections increase in strength the more often they're used and are able to process more quickly. Like synapses, memristors learn from earlier impulses. Currently, these impulses come from the circuits that they're connected to and are electrical in nature.
It's not surprising that the scientists involved in this work believe that these artificial synapses can be used to create an artificial brain. The study is the first of its kind to summarize exactly what principles from nature need to be transferred to technological systems in order to make the "brain" function.
In particular, the memristors could help in that goal. They store information more precisely than the bits on which previous computer processors have been based. If an artificial brain were created, it could make for a more efficient computer that could process data more quickly.
Andy Thomas, the lead researcher involved in the study, said in a press release, "They allow us to construct extremely energy-efficient and robust processors that are able to learn by themselves."
This artificial brain hasn't been created yet, but Thomas and his colleagues are hopeful for the future. Their paper outlines exactly what needs to happen before it's made, and could allow scientists to design a more efficient computer that we could see in future years.
See Now: NASA's Juno Spacecraft's Rendezvous With Jupiter's Mammoth Cyclone