ISN research is organized into three Strategic Research Areas (SRAs), representing the most fundamental subject areas for scientific inquiry at the Institute. SRAs are designed to address broad strategic challenges facing the Soldier, and are subdivided into specific Projects.
The phrase “positive reinforcement,” is something you hear more often in an article about child rearing than one about artificial intelligence. But according to Alice Parker, Dean’s Professor of Electrical Engineering in the Ming Hsieh Department of Electrical and Computer Engineering, a little positive reinforcement is just what our AI machines need. Parker has been building electronic circuits for over a decade to reverse-engineer the human brain to better understand how it works and ultimately build artificial systems that mimic it. Her most recent paper, co-authored with Ph.D. student Kun Yue and colleagues from UC Riverside, was just published in the journal Science Advances and takes an important step towards that ultimate goal.
The AI we rely on and read about today is modeled on traditional computers; it sees the world through the lens of binary zeros and ones. This is fine for making complex calculations but, according to Parker and Yue, we’re quickly approaching the limits of the size and complexity of problems we can solve with the platforms our AI exists on. “Since the initial deep learning revolution, the goals and progress of deep-learning based AI as we know it has been very slow,” Yue says. To reach its full potential, AI can’t simply think better—it must react and learn on its own to events in real time. And for that to happen, a massive shift in how we build AI in the first place must be conceived.
To address this problem, Parker and her colleagues are looking to the most accomplished learning system nature has ever created: the human brain. This is where positive reinforcement comes into play. Brains, unlike computers, are analog learners and biological memory has persistence. Analog signals can have multiple states (much like humans). While a binary AI built with similar types of nanotechnologies to achieve long-lasting memory might be able to understand something as good or bad, an analog brain can understand more deeply that a situation might be “very good,” “just okay,” “bad” or “very bad.” This field is called neuromorphic computing and it may just represent the future of artificial intelligence.
This week’s podcast features an interview with Ray LaPierre, who heads up the department of engineering physics at McMaster University in Canada. Ray talks to fellow Canadian Hamish Johnston about his research in semiconductor nanowires, in particular for use in photonics and quantum computers, and also shares his experiences of working at JDS Uniphase during the telecoms boom.
Physics World’s Anna Demming also joins the podcast to describe a flurry of new results in the emerging field of twistronics – where two layers of graphene are stacked on top of each other but twisted at a slight angle to each other. The discovery last year that bilayer graphene can become a superconductor if the two graphene layers are twisted at the so-called magic angle of 1.1º won Physics World’s 2018 Breakthrough of the Year, and since then the race has been on to investigate other angle-dependent properties of twisted bilayer graphene. Anna describes how different research teams are now trying to work out what causes these intriguing effects.
We also talk to industry editor Margaret Harris about the importance of technology and engineering for scientific progress. Margaret shares her own “light-bulb” moment, when she realized that new laser technology could have saved hours of experimental time during her PhD, and also highlights several articles in the latest Physics World Focus on Instruments and Vacuum that highlight how breakthrough scientific discoveries rely on developments in the enabling technologies – including the first images of a black hole that were revealed in April.
Working with mouse and human tissue, Johns Hopkins Medicine researchers report new evidence that a protein pumped out of some—but not all—populations of “helper” cells in the brain, called astrocytes, plays a specific role in directing the formation of connections among neurons needed for learning and forming new memories.
Using mice genetically engineered and bred with fewer such connections, the researchers conducted proof-of-concept experiments that show they could deliver corrective proteins via nanoparticles to replace the missing protein needed for “road repairs” on the defective neural highway.
Since such connective networks are lost or damaged by neurodegenerative diseases such as Alzheimer’s or certain types of intellectual disability, such as Norrie disease, the researchers say their findings advance efforts to regrow and repair the networks and potentially restore normal brain function.
Imagine an amazing sports car that could repair itself from damage and automatically change color! Well that’s exactly what Spanish designer Daniel Garcia was thinking of when he created this awesome concept design for the Audi A9. He got some inspiration for this particular design from Santiago Calatrava’s buildings in his hometown of Valencia. Calatrava’s architectural style looks very modern and futuristic, which is exactly what we would describe this car to look like. The windscreen and roof are to be made from a nanotechnology material that can auto-repair and adjust colors. This material isn’t even something that exists right now, so will it be a long time before we actually get to see something like this in action? The idea of it sounds awesome though, right?!
A team of chemists built the first artificial assembler, which uses light as the energy source. These molecular machines are performing synthesis in a similar way as biological nanomachines. Advantages are fewer side products, enantioselectivity, and shorter synthetic pathways since the mechanosynthesis forces the molecules into a predefined reaction channel.
Chemists usually synthesize molecules using stochastic bond-forming collisions of the reactant molecules in solution. Nature follows a different strategy in biochemical synthesis. The majority of biochemical reactions are driven by machine-type protein complexes that bind and position the reactive molecules for selective transformations. Artificial “molecular assemblers” performing “mechanosynthesis” have been proposed as a new paradigm in chemistry and nanofabrication. A team of chemists at Kiel University (Germany) built the first artificial assembler, that performs synthesis and uses light as the energy source. The system combines selective binding of the reactants, accurate positioning, and active release of the product. The scientists published their findings in the journal Communications Chemistry.
The idea of molecular assemblers, that are able to build molecules, has already been proposed in 1986 by K. Eric Drexler, based on ideas of Richard Feynman, Nobel Laureate in Physics. In his book “Engines of Creation: The Coming Era of Nanotechnology” and follow-up publications Drexler proposes molecular machines capable of positioning reactive molecules with atomic precision and to build larger, more sophisticated structures via mechanosynthesis. If such a molecular nanobot could build any molecule, it could certainly build another copy of itself, i.e. it could self-replicate. These imaginative visions inspired a number of science fiction authors, but also started an intensive scientific controversy.
Conventional computers need orders of magnitude more energy than the “computer” in our heads. Neural networks made of superconducting nanowires might come much closer to the real thing.
The rules about what makes a good magnet may not be as rigid as scientists thought. Using a mixture containing magnetic nanoparticles, researchers have now created liquid droplets that behave like tiny bar magnets.
Magnets that generate persistent magnetic fields typically are composed of solids like iron, where the magnetic poles of densely packed atoms are all locked in the same direction (SN: 2/17/18, p. 18). While some liquids containing magnetic particles can become magnetized when placed in a magnetic field, the magnetic orientations of those free-floating particles tend to get jumbled when the field goes away — causing the liquid to lose its magnetism.
Continue reading “Permanent liquid magnets have now been created in the lab” »
A new wireless transceiver invented by electrical engineers at the University of California, Irvine boosts radio frequencies into 100-gigahertz territory, quadruple the speed of the upcoming 5G, or fifth-generation, wireless communications standard.
Labeled an “end-to-end transmitter-receiver” by its creators in UCI’s Nanoscale Communication Integrated Circuits Labs, the 4.4-millimeter-square silicon chip is capable of processing digital signals significantly faster and more energy-efficiently because of its unique digital-analog architecture. The team’s innovation is outlined in a paper published recently in the IEEE Journal of Solid-State Circuits.
“We call our chip ‘beyond 5G’ because the combined speed and data rate that we can achieve is two orders of magnitude higher than the capability of the new wireless standard,” said senior author Payam Heydari, NCIC Labs director and UCI professor of electrical engineering & computer science. “In addition, operating in a higher frequency means that you and I and everyone else can be given a bigger chunk of the bandwidth offered by carriers.”
One of the main speculations about future technology is uploading. This is where our minds are copied in exact detail from our biological physical bodies and then created in artificial bodies. Alexander Bolonkin has posited many kinds of technology over the decades. He has a recent work which is summarized here where he considers that future uploading will mean that we can then use super-technology (nanotechnology, nuclear fusion etc…) to make people into literal gods and supermen. We can use control of matter, energy and information to make what he calls the E-man. Bolonkin then indicates that uploading and creation of minds could be used for the resurrection of long-dead people. This would be where we create the very close approximation of dead people. This would be like using gene editing to turn an African Elephant into a Whooly Mammoth. The vast technological capability would let us actualize what would be a simulation into living entities.
Bolonkin’s Case for E-Man and Resurrection
Alexander Bolonkin looks at methods and possibilities for electronic resurrection of long-dead outstanding personalities. He also considers the principles and organization of the new E-society, its goals and conditions of existence.
