It sounds like a social network that kind of punishes participation...why would you recommend anyone if it could damage your reputation. The social contract that's supposed to underly social networks does this anyway; someone who retweets / toots bots gets a bad reputation adn loses followers. It's just the bots got really good at spoofing the social graph too...an llm can gain reputation by recommending humans just like humans can.

Hate to say it but when I talk w/ Microsoft Copilot I hear a lot of talk like

   ... being human isn't a policy, it's structural