How Algorithms Manipulate You (and How to Fight Back) | Idea Knock Down | TED

0:00 – 3:42

The scam hidden inside convenience

The tower starts wobbling almost immediately, and so does the polite fiction that algorithms are just being helpful. Asked whether they have ever felt manipulated by an algorithm, both answer without hesitation: "Wholeheartedly, yes" and "all the time." The example is almost funny until you notice what it reveals. Shaolin jokes about a pair of yellow shoes she did not need that "followed me everywhere," and Jen says the worst part is when the system is right, because then the feeling is not just annoyance, it is exposure. "I do like that thing. Stop spying on me." That is the first unexpected angle of the video: manipulation works best when it feels useful. From there, Jen draws a line around what AI should never be allowed to do. Her rule is simple and memorable: AI should never make decisions that remove opportunity from a person. She uses mortgages as the concrete case. Maybe an AI could help approve someone, she says, but it should never be the thing that says no. The reason is not that algorithms are always wrong. It is that when a black-box system denies housing, credit, or some other life chance, it turns human judgment into something unchallengeable. That leads into what she calls the "core legal problem" in the AI ecosystem. In the US, once people hand data to a third party, that company can often "do whatever they want with it." Jen and Shaolin both insist personal data should belong to us, not because privacy is a luxury, but because so much of the current dystopia flows from giving away intimate information and losing control over how it is bought, sold, and repurposed. When they get to what ordinary people can do, the answer is frustratingly practical: stop agreeing to everything, reject cookies, install tracker blockers, accept a little inconvenience. Jen says the trade is real. If you protect your privacy, life gets less frictionless. "You're not going to be able to order pizza in that browser. But like you'll be fine." Shaolin sharpens the point with a challenge that hangs over the whole video: convenience for whom? She points to self-checkout and phone systems that make it impossible to reach a human. What gets marketed as convenience for users often looks more like efficiency for companies, with the human part quietly stripped away.