Do This, NOT That: Marketing Tips with Jay Schwedelson l Presented By Marigold Does Bullying AI Work? + Jay's TV Recos!🤖 Jay’s SCOOP | Ep. 391
12 snips
Aug 8, 2025 Jay Schwedelson reveals his surprising tactic of yelling at AI, thinking it might boost performance. He discusses Sergey Brin's controversial claim that threats improve AI results, but new research disagrees. Instead, frustration may lead users to provide more useful information. Jay's analogy likening AI to a doctor transforms how we should treat it for better outcomes. On the entertainment front, he shares thoughts on TV shows, questioning the future of the Bachelor franchise while recommending the UK version of Love is Blind.
AI Snips
Chapters
Transcript
Episode notes
Jay Bullies AI for Better Results
- Jay Schwedelson admits he bullies AI by being mean to get better responses.
- He shares that he yells at AI when answers disappoint him, expecting improvement.
Tone Does Not Affect AI Answers
- Wharton School research shows AI performance is unaffected by the tone of questions.
- Being mean or nice yields the same answer quality and validity from AI.
More Info, Not Mean Tone Helps
- Frustration with AI leads users to add more context and detail in follow-up queries.
- Additional information, not aggression, improves AI responses.