Fukoku Mutual Life Insurance believes it will increase productivity by 30% and see a return on its investment in less than two years. The firm said it would save about 140m yen (£1m) a year after the 200m yen (£1.4m) AI system is installed this month. Maintaining it will cost about 15m yen (£100k) a year.
Human workers being replaced by machines? Hasn't that been happening for at least a few hundred years now?
Cynicism aside, the IBM Watson system would be a really interesting piece of technology if I knew how it actually worked. I was allowed to play with a cut-down version for a university project two years ago which allowed you to feed it a large number of text articles and then ask it questions in natural language. You had to prime it with set questions and set answers, and it would then use its machine learning technology to detect the kinds of relations you were interested in, in order to anticipate new unforeseen questions.
From the articles I was able to read, it didn't use any particularly fancy models, the genius was more in how it could collate the results of all of these methods and estimate how confident it was with each hypothetical answer to each question (a metaphor: anyone can how to prepare sushi rice, but each restaurant will closely guard its particular ratio of sugar, salt, rice vinegar and mirin). Unfortunately this wasn't of much use to me because while I was able to test out a particular subset of its abilities I wasn't able to figure out how it worked internally, which as a computer scientist interested me more than whether it could be use to parse medical articles (it definitely can). Oh well. I can understand why IBM does not want to show its hand. At least they managed to contribute to the academic discourse by publishing papers regarding some of the individual components of the project.