The media is full of AI success stories which beat humans at the tasks we consider ourselves uniquely suited for—IBM’s Watson winning Jeopardy, Google’s AlphaGo beating the Go grandmasters, Deepfake image and video manipulations. Furthermore, conferences and AI commercials are quick to point out how AI is being used to cure cancer, improve food production, and basically usher in a global utopia.
Yet having been in many of the S&P500 headquarters, I actually observed that few if any Machine Learning projects have succeeded—usually following a trajectory of an announced implementation by the customer to much fanfare, a year-long implementation phase that stretches to two, some minor internally announced “wins” that ...
Get How to Talk to Data Scientists now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
