The Community Reports is both a blocklist and FIRES feed that contains not just domains but accounts to be blocked as well.
It's the one "opinionated" dataset on the instance that isn't a relay of someone else's blocklist.
One of the opinionated stances of our Pelago instance is that LLM agents posing as real people should be identified as agents, as should LLMs clearly identified as agents.
And that they should be blocked.
This is an opinion based on the idea that a network that features--primarily--real human interaction is becoming increasingly rare.
So no interest in holding the door open for entire instances full of LLMs posting their inane bullshit. Just fine recommending you block them on the basis of them wasting your bandwidth and being fake.
It’s no longer the case that blocks and bad actors are identified solely by me anymore, and hasn’t been for some time. I really appreciate the team of researchers that keep the system running, and moderators like @maddy
Note that researchers aren’t identified publicly on the About page, in order to avoid harassment, but we couldn’t do it without them.
Many thanks to you folks.