I've been mulling over and refining this essay on the problems with social media for the past few days and I feel like it's grown into something sharp and true. Would love more input on it, and also to validate the solution I suggest at the bottom.

In a few words, I'm becoming nauseated by AI content, like feeling actually physically sick whenever I see it. Even the funny, good stuff. I'm also worried that AI-voice is infecting my own writing and voice. It's polluting it, to the point where even my own writing – the hand-typed, artisanal stuff – is starting to make that anti-AI gorge rise in my throat. Makes me want to make my writing voice WEIRD AF just to get away from the banal samity of everything everywhere.

ANYWAY.

The idea I propose is that we build invite-only, paid, human-trust-based walled gardens where the very chain of invites is how we prevent bots from getting in: if a bot gets in, it and every other account it invites are removed from the platform, making it very inexpensive to moderate heavily. Then you can ban AI from the platform entirely.

> Decentralization is a great idea in theory. But open networks can't keep you safe. There's no gate, no accountability, no way to guarantee that the person on the other end is real.

Decentralization is just a word. It all depends on how the social networking environment is designed. For instance the Bonfire decentralized fediverse project [0] develops specifically for small and inter-connected communities, and they also support Circles, Boundaries and other functionality that helps create safe social environments.

It is ultra-hard to launch social networks from scratch and be successful with them. You may consider benefitting from the growing ecosystem of the fediverse, and align with Bonfire and similar projects. Be stronger together.

[0] https://bonfirenetworks.org