Hey {{first_name}}
Data quality is the new moat in AI. Volume was never the hard part.
This week on The AI Why, Liam sits down with Enzo Blindow, VP of Data & AI at Prolific, to unpack what actually separates good training data from garbage… and why "good taste" might be the one thing AI can't fake its way into.
They get into why models default to stereotypes, how a single bad translation can quietly poison a dataset, and the research Prolific ran showing how easily AI can be pushed toward commercially convenient (and sometimes harmful) answers.
Also available on Spotify & Apple Podcasts.