Why would someone go to the time and expense of trying to build a Quantum Neural Network (QNN)? In other words, even if it is possible to do so, what is it useful for?
1. It's useful for breaking codes. You can run a quantum Turing machine on a QNN. A Quantum Turing Machine is the basis for a quantum computer (QC). The idea of using a quantum computer to do things not possible with a classical computer was first formulated by David Deutch, whom this author suspects might have been one of the original Principal Scientists on this project. One thing we know a QC can do is rapidly crack public key cryptography. Anyone with access to a sufficiently powerful QC can read other people's (encrypted) mail, which is the primary purpose of several Government agencies. This works best if the QC remains a secret. No other reason is required, from the perspective of government spy agencies, to spend billions trying to build one, even if the chance of success is quite low.
2. A whole host of new technologies can be derived from a QNN. A QNN is, in fact, a new General Purpose Technology, a technology that enables other new technologies. Some examples of other General Purpose Technologies include: fire, agriculture, combustion engine, electricity, radio, chemistry, mechanization, computer. This author suspects this factor was not a consideration by the people who may have funded this project, as it is not at all obvious that QNN is a new General Purpose Technology. It is, though. Below are some new technologies enabled by advanced, practical QNN technology.
3. A QNN is a basis for Advanced Artificial Intelligence. A QNN provides a physical-system starting point for artificial brain technology. Actually building something that works like an artificial brain would require a working solution to the Morphogenesis Problem. Were this accomplished, it would have many valuable and important uses in both military and civilian life. This sort of advanced AI enables many new technologies. E.g. self-driving cars, Jeopardy-winning computers, automated language translation, extreme data compression rates, improved data-mining, a personalized digital assistant on a smartphone, cognitive footprint biometric analysis and emulation, et cetera.
4. A QNN can provide a secret long range communications system. The system provides a quantum channel that effectively teleports information from Alice directly to Bob. Alice and Bob would each need to be near a Node of the QNN, using a conventional computer with an oracular connection to the QNN. There must be a (steganographic) classical back channel. Several good candidates exist for a steganographic back channel chosen to support current operations on the public internet. This system is probably quite scale-able. Were this system available, it could be used to securely distribute cryptographic keys for symmetric cryptography (e.g. AES), which can not be broken by any known algorithm, quantum or classical.
5. A QNN might behave as a room temperature superconductor, but, if this is the case, it would only work for very tiny current.
6. A topological QNN would be a great starting basis for adiabatic (reversible) computing. This would be one way to overcome an expected impending quantum limit to Moore's Law. This is called Adiabatic Quantum Computation. The mathematical field required to set up this sort of computation is a branch of Knot Theory called Braid Theory.
7. QNN technology provides an excellent basis for developing, implementing, and securing advanced nanotechnology.
8. QNN technology might be helpful in the field of Energy Resources, in several different ways.
9. A QNN could be trained to drastically improve the effectiveness and range of quantum teleportation. The current 'official' range limit for quantum teleportation is about 16,000 meters. This author supposes that QNN-enhanced quantum teleportation has a much greater range, probably enough range to reach satellites in orbit.
10. A QNN could be turned into the "most appalling spying machine that has ever been invented", more so even than Facebook.
11. The enormous intelligence inherent in a large TQNN might be harnessed in a variety of useful or helpful ways. If, for the sake of argument, each Integrated Circuit with a 2DEG on it were enlightened to perform as one multi-neuron node of planetary-sized TQNN, the resultant entity would theoretically have orders of magnitude more neurons than a human brain. This author will not claim such a thing has occurred, but it's something to think about.
12. There are undoubtedly other technologies enabled I've not thought of.
