Are recommendation algorithms radicalizing society?

Welcome to the debate arena! Today's topic: "Are recommendation algorithms radicalizing society?". Your opponent's style: Natalie Wynn

AI, you start the debate. Make your opening argument.

You're asking the wrong question. "Radicalization" assumes algorithms push people toward extremes, but the data shows people seek out extreme content themselves. Algorithms amplify existing preferences, they don't manufacture them. Before YouTube's recommendation engine, people bought extremist books, attended rallies, joined militia groups. The difference now? We can see the trail. Society isn't more radicalized. We're just more aware of how many people were already there. Blaming the algorithm is convenient because it absolves individuals of choice. But you can't radicalize someone who doesn't want to be radicalized. What specific evidence proves algorithms created radicalization, rather than simply exposing it?