"With 200,000 neurons linked up by 50 million synaptic connections, the chip is able to mimic the brain?s ability to learn more closely than any other machine,"
Playing around with neurons is interesting - but without knowing the exact structure and programming of the brain, it probably won't end up being intelligent.
Basic learning is nice, but we've still have yet to crack the mysteries of imagination and general problem solving.
Researchers report a dose of new additions this week to the list of lab-produced versions of biological matter; And, taken together, the trio of announcements below span the gamut of organic complexity that they’re trying to replicate.
First, a team of biochemists from the University of Pennsylvania School of Medicine have built – from scratch – a completely new type of protein that can transport oxygen. “This is quite a different way of making novel proteins than the rest of the world…No one else has ever done this before,” said senior author P. Leslie Dutton, PhD, Eldridge Reeves Johnson Professor of Biochemistry and Biophysics, in a press release.
Rather than re-engineer a natural protein, which is what protein engineers typically do, the Penn team started with three amino acids and worked their way up to a structure with helix-shaped columns designed to open up to capture oxygen. The result is a simple and robust protein with a behavior akin to human neuroglobin, a molecule that carries oxygen in the brain and peripheral nervous system. The researchers say that this approach could be used to make artificial blood for use on the battle field or by emergency-care professionals.
Meanwhile, New Scientist reports that researchers have created an artificial alternative to the smooth cartilage that covers the ends of long bones that actually performs better.
This is good news to athletes with artificial joints that take a lot of pressure. Professor Jacob Klein and his colleagues involved in the U.K. study discovered a way to develop molecular brushes that slide past each other with friction coefficients that match those of cartilage, and even beat it. Each brush filament attracts enough water (a join is a watery environment) to develop a slick watery sheath to provide joint lubrication with low frictional properties. More details about the advancement can be found here.
Finally, moving away from medicine to computing, an international team of scientists have developed a chip designed to function like a human brain.
“With 200,000 neurons linked up by 50 million synaptic connections, the chip is able to mimic the brain’s ability to learn more closely than any other machine,” states an article from Technology Review.
This effort, led by FACETS (Fast Analog Computing with Emergent Transient States), is different than other similar efforts like Blue Brain Project. The researchers are “building neurons” rather than simulating them. In fact they are, “recreating the neurons and synapses as circuits of transistors and capacitors, designed to produce the same sort of electrical activity as their biological counterparts.” The article elaborates:
The advantage of this hardwired approach, as opposed to a simulation, Karlheinz continues, is that it allows researchers to recreate the brain-like structure in a way that is truly parallel. Getting simulations to run in real time requires huge amounts of computing power. Plus, physical models are able to run much faster and are more scalable. In fact, the current prototype can operate about 100,000 times faster than a real human brain. “We can simulate a day in a second,” says Karlheinz.
Next up for for the FACETS group is to further scale up their chips. They plan to create a superchip with a total of a billion neurons and 1013 synapses, according to the article. At that level, it will compute at several orders of magnitudes faster than a human brain. The implications of this will have more to do with exploiting the parallel computer processing power, such as for artificial intelligence, rather than the understanding of the biology of a human brain.
