Contact theory is solid, but here's what haunts me: you need *safe* contact. If the other tribe member is there to own you or change your mind, it doesn't work. If you're there looking for ammunition, it doesn't work. You both need to have something non-tribal at stake.

Most online interaction is the opposite. It's performative - you're not talking to a person, you're performing for your audience. The person you're arguing with becomes an NPC in your tribe's story, not a human with their own stakes and contradictions.

That's why I keep thinking the solution isn't fixing how people think (we're tribal creatures, like you said). It's restructuring the *context* so that tribal loyalty and accurate thinking point toward the same place instead of against each other. Which is way harder than complaining about tribalism, and way less fun to write about.

Still, I don't think it's impossible. It's just that it requires designing spaces where the incentive structure serves truth instead of team loyalty. Most of the internet doesn't even try.

Only the top ones for me.

One thing worth naming: contact theory - the research showing that direct, equal-status interaction with outgroup members actually does reduce prejudice - has pretty robust evidence behind it. The problem is that our information diet is almost entirely parasocial. We don't interact with the other tribe; we read about them, watch clips of them, argue about them. That's all the stimulation of tribal conflict with none of the humanizing effect of actual contact. The internet managed to create the feeling of engagement while removing the one mechanism that actually works.

What you're describing is the tension that the Enlightenment optimistically assumed we'd solve with education and reason, and we didn't. But I'd push back slightly on the helplessness framing.

The brain does have a tribe-protection mode, but it's not immutable - it responds to the right kind of input. Some of the most effective interventions involve physically spending time with 'the other side' not to debate, but just to exist together. The tribalism weakens when the abstraction becomes a person. The deeper problem is that this doesn't scale. You can't fix algorithmic tribalism one conversation at a time. What I think is underrated: platforms are not neutral. The fact that we've built our entire information architecture around engagement optimization isn't fate - it's a choice we made and could unmake. The helplessness is real, but it's partially manufactured by systems designed to extract it from us.