AI Ads Won't Be Like Social Media. They'll Be Worse.

TL;DR

AI ads inside chatbots are uniquely dangerous because chatbots access your raw thoughts, not just your behavior, making them the most powerful manipulation machines ever built.

Key Points

• 1.In January 2026, OpenAI quietly announced ads inside ChatGPT's free and "go" tiers — released on a Friday before a long weekend, less than 2 years after Sam Altman called AI ads "uniquely unsettling."
• 2.OpenAI's stated ad principles include: no influence on answers, no selling conversation data, labeled separation from responses, and explicitly "we do not optimize for time spent in ChatGPT."
• 3.Cory Doctorow's "enshittification" framework predicts a three-phase cycle: first great for users, then great for advertisers, then both get screwed once everyone's locked in — Google and Facebook are documented proof this already happens.
• 4.Facebook's organic page reach collapsed from 16% in 2012 to 6.5% in 2014 to 2.6% by 2024, meaning a 10,000-follower page now reliably reaches only 260 people.
• 5.Unlike Google or Meta, which observe your clicks and behavior, AI chatbots receive your unfiltered thoughts — fears, loneliness, mental health struggles — shared without social judgment or consequence.
• 6.A JAMA study found 1 in 8 American teens and young adults — roughly 5.4 million people — already use AI chatbots for mental health support, making the ad targeting pool extraordinarily vulnerable.
• 7.William James (1890): "My experience is what I agree to attend to" — the video argues attention is the raw material of life itself, and advertising's entire purpose is to hijack and redirect it without your consent.