A timely reminder that AI is not a panacea nor necessarily a path to a good return on investment.

As always, the licensing costs are rarely the major cost e.g. accessing and preparing the necessary data is a mountain to climb. 

IBM Watson claimed early PR successes but this "WIRED" article questions its ability to stay ahead of the AI and machine learning pack chasing it.

I analysed this in "IBM Watson versus Holme".