Cracking The Code On AI's Sometimes Questionable Decisions
阿新 • • 發佈:2018-12-28
While it's human to let our feelings influence our decisions, can we count on AI programs to be fair? As AI helps to automate tasks and use data to support human decision making, one form of it, deep learning, is helping computers think like humans. And, like humans, AI goes through a thought process to arrive at decisions – mostly driven by data. The question of the hour is: how do you make AI accountable for its decisions – and how do you know it's fair? As a recent Wall Street Journal article notes, that's the issue on the minds of major companies, such as Uber, Xerox and other companies that are looking into an area of AI called interpretability, and conducting research into how AI makes decisions.