Every AI answer lately sounds a little too nice, and educator Mohit Rajhans argues that isn’t a glitch. It’s the design working exactly as intended, and it says less about the machine than it does about what humans actually respond to.
Mohit traces that shift back to the same engagement-driven incentives that got a major social media platform fined for keeping people hooked, and pokes a hole in the idea that a paid AI subscription buys you out of being the product. There’s a real line between crowdsourcing data on purpose and having it quietly collected anyway, and most people never bother to look for it.
Then there’s the loop nobody’s watching for: a thought gets filtered through AI, posted, read by another AI, filtered again, and reposted, drifting a little further from anything a person actually wrote each time it happens.
Topics: AI agreeableness, AI validation, data privacy AI, AI misinformation, Mohit Rajhans
GUEST: Mohit Rajhans | http://thinkstart.ca
Originally aired on2026-08-28
