The late and great Terry Pratchett encapsulated the difficulty of discussing aspects of Quantum Physics, (let alone quantum computing and machine learning) when he wrote:
‘It’s very hard to talk quantum using a language originally designed to tell other monkeys where the ripe fruit is.’
(Terry Pratchett, Night Watch, 2002)
There are regular news stories about the race to construct and develop the first truly quantum computers; ever since Feynman and Manin first posited the notion in the early 1980’s. The main players in this endeavour include IBM, Google, Microsoft and Intel, among others. Various different approaches are being followed, based either on analogue or digital implementation; with both systems utilising quantum bits or qubits.
Quantum Machine Learning
The implications of the development of such computers for the field of Machine Learning (ML) are significant. Due to the large amount of data utilised in ML, as well as the amount and nature of the complex processes and algorithms used, the ability to deploy quantum computing resources to these will enhance, streamline and make more efficient many ML operations. It is generally accepted that both traditional computation methods operating in tandem with or even directing / allocating tasks to quantum computers would provide the most efficient and successful results.
An additional facet would be the creation of complex quantum neural networks, with the same predicted exponential increase in processing / learning power. A more in-depth analysis of the progress in this area can be found in the National Science Review.
Finally, even as the predicted aspects of quantum machine learning are already being demonstrated and utilised, further future uses are being posited and developed, including researchers at the University of the Basque Country who have combined machine learning, quantum computing and artificial intelligence in an attempt to synthesise models of artificial life. Whilst at a nascent stage, the advances in quantum computing facilitate ever more complex models with random or disruptive variables to be constructed and analysed.