What Does The Future Hold For AI? Five Predictions

Deep learning has been immensely successful in solving perception problems, particularly when there are large supervised (that is to say, labeled) data sets available. This is the technique that has given us amazing image and handwriting recognition, language translation and much more. However, the need for massive amounts of labeled data can be a problem in the real world. For example, AI can work with industrial systems. In order to protect companies, AI can predict when a large industrial asset such as a turbine might fail in the real world.