Travis Dirks, PhD, University of Illinois @ Urbana Champaign, CTO at XLabs
Building AI for difficult and impactful problems - Like most things worth doing it is terribly difficult. I'll attempt to pass one the key lessons learned, and mistakes to be avoided when one is operating at just the right amount of lost, such that something new and interesting might hope to be found. Key pillars include:
What is AI - Fastest delivery of Ah-ha I know; The Biggest Mistake - Overfitting, the AI Triumvirate and AI's Dirty secret; The Second Big Mistake - Validation and Confusion of Business Decisions and Tech Decision; The Common AI Tragedy - Spending huge effort to determine is the AI is correct, but failing utterly to ask if it is right. People People - 2 AI Super powers you bring that must be at the table.