Explainable AI
Getting more from AI doesn’t mean doing more AI. We need to consume the AI we’ve already got.
But how?
Michelle Lacy, head of data science at Bayer Crop Science, suggests we “put the algorithm behind the button.” Hers is the hide-the-medicine-inside-a-treat trick we use with our cat, Yogi. When you hide AI in a chart, it’s easier to consume.
Jeff McMillan at Morgan Stanley, says “We find that, with our data science, that the single most powerful driver of client satisfaction is a phone call.” Jeff’s approach is the pick-up-the-damn-phone-and-talk approach.
Mark Hallard from the Scottish Environmental Protection Agency augments charts with text to explain findings. They’ve reached an astounding level: over 80% of SEPA’s employees use data-driven insights.
We have data science; too much. We have tools; too many. The best outcomes come from making it easier to consume AI, not just making more of it.
Yup, that’s our cat, @yogi_peepers
