By Muon Ray
Want to Learn More about Quantum Computers? Check out: http://muonray.blogspot.ie/2014/11/qu…
Lockheed Martin, NASA and Google all have one thing in common: they all own quantum computers. These quantum computers are superconducting chips which were designed by D-wave systems and were fabricated at the NASA Jet Propulsion Laboratories.
NASA and Google share a quantum computer for use at the Quantum Artificial Intelligence Lab using a 512 qubit D-Wave Two that would be used for research into machine learning to assist in using artificial neural networks to search through large astronomical data sets for extrasolar planets and to increase the efficiency of internet searchs by using A.I. metaheuristics on heuristical search engines.
Such A.I. metaheuristics may resemble global optimization problems similar to classical problems such as the travelling salesman, ant colony or swarm optimization, which can navigate through maze-like databases. Using entangled particles as qubits, these algorithms can be navigated far faster than conventional computers and with much more variables.
By using a decentralized, quantum swarm AI, it may be possible to simulate emergent behavior aswell, such as Langton’s ant, which could see the rise of quantum-based simulated intelligence which could go so far as to create realistic cellular automatons on a computer.
The use of sophisticated metaheuristics on lower heuristical functions may see computer simulations which can select specific sub routines on computers by themselves to solve problems in a truly intelligent way. In this way the machines would be far more adaptable to changes in sensory data and would be able to function with far more automation than would be possible with normal computers.
Moreover, it may be possible to use metaheuristics to perform error correction on software using neural networks by comparing the optimized problem solving in a quantum computer with the regular programming software of a regular computer. Since regular computers are not quantum mechanical, they must be programmed classically. However, by using quantum metaheuristics it may be possible to perform optimization problems using Artificial intelligence on a quantum computer and then compare to the command line architecture in a piece of conventional software on a classical computer, which may be too complex to modify or to check for errors using human software engineers.
It may become routine that to design large software, rather than use a team of engineers to design a single piece of large software and to do multiple tests, revisions and redistribution, quantum optimization problems will be run though a machine instructed to do a particular task and the software will be designed around such optimizations. Computer chips could even be designed to be superconducting and quantum compatible in a factory or laboratory and classically operating in conventional computer equipment. In this way, optimization and programming could be done on the chips in a factory or lab to search for any possible errors that could arise in designing software. After which the chips would be used in conventional computers for their particular tasks. Quantum computers may well find their place in industry more than anywhere else, designing AI optimized software around computer chips.