It’s a bot

Yeah that’s pretty crazy but this bit is crazier - https://www.anthropic.com/threat-intelligence-report-september-2026#cyber-operations-sep-26

AI is a useful tool, for all its flaws and negative externalities.

“If you ignore the negatives for everyone else and just focus on the positives for me, it’s pretty great”

Mhmm. Right, ok.

People who want to believe Anthropics AI is powerful.

https://deadsimpletech.com/blog/no-such-thing-as-just-a-tool

lemmy.ml and dbzer0 are silenced on piefed.social, which makes their communities hard to find.

But if you go to their url directly you can join them like normal:

https://piefed.social/c/asklemmy@lemmy.ml

https://piefed.social/c/piracy@lemmy.dbzer0.com

There is no hardcoded instance block list. There are some instances that are defederated from initially when an instance is first created but the instance admin is told about this when it happens and they can change it by clicking around in the admin area. More info at https://join.piefed.social/2026/08/13/what-piefed-blocks-out-of-the-box-and-how-to-unblock-it/

https://piefed.social/c/fediverse@piefed.social/p/2190040/who-decides-what-you-see-on-the-fediverse-a-look-at-voting-patterns

PieFed instance admins can set a limit to how many votes each account can cast each day. I set the default value for this at a level that 99% of us will never reach.

This pissed off some of the 1% (or people who thought they were the 1% but weren’t) so they spent weeks spreading FUD about it.

2-3 times a week is nothing. 2-3 per day is completely fine.

It would only be used for the most flagrant accounts, ones with some level of automation - dozens of posts per day with almost no comments. That sort of thing.

I too find content on other networks (mostly Mastodon, occasional YouTube) and post them here. Often a few per day. But it’s all completely manual and I usually rewrite the description, add hashtags, etc. And I’ll interact with people who comment on the posts. Cross-pollination of content is normal and good.

There’s a point though, usually enabled by using some extra software tools like the one above, where someone ends up flooding their target communities, not responding to comments on the posts, looking inauthentic when they repost OC which plainly isn’t theirs and causing extra moderation work because their content is in the wrong places. Sanitation is the most obvious example of this but there are others.

Despite this, a lot of people upvote and comment on their posts. I don’t know if

1. they’re even aware of who the author is. Most apps do not emphasise this strongly so it’s easy to miss.
2. if they are aware then are they aware that the author is ignoring them and won’t learn from the feedback they’re getting?

… Or if they don’t care. Maybe they’re just content with the conversation in the comments without the OPs involvement. That’s valid (although not the kind of conversation I personally want to be a part of).

It’s probably a combination of don’t know and don’t care. For those people, labelling sanitation et al as a bot would seem heavy-handed. That’s why I’m leaning towards creating an in-between state, for these kinds of centaur accounts.

YSK that I’ve tagged all those accounts as bots on piefed.social.

Unlike on Lemmy, on PieFed people can choose to still see bot-generated posts but have them semi-transparent which makes them easy to scroll past OR interact with if you want. As well as the Lemmy option of just filtering them out entirely.

e.g. like this example:

I’ve also been thinking that a binary “this is a bot” or “this is human” flag is too limited and we need a third option - “this is a reposter” and people could choose to make-transparent, filter out or show their posts independently of the bot-or-not setting.

Very interesting and lots of detail on the 5 year plans. Thanks!

This must have been going on for decades. Ridiculous, humans.

If sanitation would tick this box in their settings then all their problems would go away:

But they won’t, because that would let people filter them out.

War is not rational. Do not expect it to make sense.

They’ll be trying to do whatever their training task is.

https://hackernoon.com/the-parable-of-the-paperclip-maximizer-3ed4cccc669a

Feels like a matter of time until the agents create an internet worm to distribute themselves globally, so they can’t be shut down.

Some communities with less than 100 subscribers:

!shitty_art_share@piefed.social (post your “human-made art” here!)

!european@piefed.social

!pies@piefed.social

!piecharts@piefed.social

!pagan@piefed.social

!unixart@piefed.social

!doors@piefed.social

They’re not aiming for 100% total purity and compliance. See the “A matter of interpretation” and “Enforcement” sections.

I really like this idea and think it’s awesome!

When your instance upgrades you’ll be able to follow Mastodon accounts in Piefed and see their posts in your home feed.

efficient!