Prof. de Garis gets Adjunct Professorship in Theoretical Physics at Utah State University's Physics Dept
I will be teaching a course on "Frontiers of Computing" in the fall semester of 2002, at USU. Part of this course will be on quantum computing (QC). One of the physics PhD students who has been wanting to research in QC for some years asked if I would supervise him. I agreed, but for it to be legal I would need to be at least an adjunct professor of theoretical physics in the USU Physics Dept. So, that is what has happened. I now have a joint appointment at USU - an associate professorship in the Computer Science Dept, and an adjunct professorship in theoretical physics in the Physics Dept.
Since my basic degrees were in theoretical physics and applied math, the research field of quantum computing combines the two great intellectual loves on my life. Also, I believe that QC is the future of computer science, since Moore's law will be taking us into the era of molecular circuitry within a decade or so. At that scale, quantum rather than classical physics becomes the descriptive tool.
My research group is spending some of its time wondering how to evolve neural networks (the basis of brain building) when our electronic circuits are the size of molecules. We will need a quantum version of what we do, so we are dreaming up a quantum version of the evolutionary engineering of neural networks (to the extent that that is possible?)
We feel that the traditional field of evolutionary algorithms (EAs) is doomed, to be killed off by QC. EAs are a form of sampled search in a huge search space of some 2**N points (derived from an N bit bitstring). Since classical computing can only process one bitstring (chromosome) at a time, a sampled search technique becomes necessary when N is large, and hence 2**N is huge.
However the great feature of QC is that all 2**N chromosomes can be processed simultaneously (using the phenomenon of QC "superposition"), thus making the fundamental assumption of EAs (the need to search only a tiny fraction of the total search space) irrelevant. QC will kill EAs, therefore I advise EAers, to start getting into QC if they want to stay employed. If you cant beat them, join them.
My group now dreams of using quantum techniques to create (we cant say "evolve" any more) molecular scale neural circuits, and assemble them into "quantum brains", where the neural net circuit modules are generated and function according to quantum principles. It is a new perspective for us, and very challenging. Since there are no actual quantum computers yet, our thinking has to be purely theoretical.
Stay tuned.