An example is that non-convex algorithms can be run more efficiently to remove misclassified images in image recognition training datasets. This will enable classification and searches of images to be improved using regular classical algorithms.
If there are 1 million images that were classified by humans and 1% were wrongly labelled and the quantum computer could effectively remove the ones that were wrong then the image searches would be improved. Google has been cooperating with Dwave on research to prove out these methods.
So a combination of more qubits, faster machines, better algorithms and experience operating the machines and learning what they are best at should lead to success.
If you liked this article, please give it a quick review on ycombinator or StumbleUpon. Thanks
