[Lastodon]

"human knowledge and perspectives not represented in this set could be systematically wiped from public discourse."

That's a feature, not a bug. With this technology "Dangerous misinformation" can be cleanly and easily removed from platforms with minimal human intervention. If you don't think that that's exactly what it'll be used for, you haven't been paying attention.

[redblacktree]

Sounds a lot like 1984. Who gets to decide the allowable discourse?

[amelius]

For now, AI is generating more misinformation than it removes, though.