The use of quantum computing, as a potential means to slip past the limits experienced by the current design in microprocessors, has been capturing imaginations since the 1980s. Now, Google (News - Alert) has been spotted making some new advancements on that front, pushing up its research into the field and calling in John Martinis and his team from the University of California at Santa Barbara (UCSB) to help advance the development of quantum processors with a basis in superconducting electronics and an eye toward putting same to use in artificial intelligence (AI) development.
Martinis and company made a good choice on this front, as the group was part of the first such movements into this field back in 2008. Google, meanwhile, seems to be expanding on an interest in quantum computing, as it was one of the first buyers of a similar quantum computing system from D-Wave back in 2008. Though at the time, no one was sure how, or even if, the processor worked, it was reportedly enough for Google to take an interest. That interest carried on through work with NASA and the Universities Space Research Association in Google's Quantum (News - Alert) Artificial Intelligence Lab.
All that work, meanwhile, seems to have been just the groundwork necessary for a larger project, one in which Google uses its own hardware in the development of such projects. Word from Google's director of engineering, Hartmut Neven, notes that Google, “...will now be able to implement and test new designs for quantum optimization and inference processors based on recent theoretical progress and insights from the D-Wave quantum annealing architecture.” Neven further noted that the company will be working with D-Wave in the future looking to its Vesuvius machine and some further work with NASA.
Though the actual connection to AI here was somewhat downplayed, Martinis did reportedly mention a particular excitement in terms of dealing with Google's expertise as far as “...mapping machine learning applications to a quantum computer” goes, likely referring to Google's recent hire of deep neural networking pioneer Geoff Hinton.
It's still very early days as far as both quantum computing and artificial intelligence go, but it's likewise clear that the one is likely to prosper with the rise of the other. Quantum computing might well be able to put forth the kind of sheer power that will be necessary to run an AI system properly, and working AI might prove the reason to invest the time and resources necessary to generate quantum computing systems in the first place. Each side works to help the other, and the idea of a computer program that can do potentially as much as AI can be a sufficiently tempting target to make developing quantum computing systems seem more necessary to help close the gap between what we have now and what we would need to get to that point.
It's sort of a roundabout way of getting there, but it looks like the concept of the tech tree might well help quantum computing. We need that branch fully grown before the fruit of AI can take hold, and with all those minds and all those resources in the fray, we may well have both technologies in hand before too much longer has passed.