Okay, so I just wrote out a really quick and rough wiki page here: https://fediverse.wiki/wiki/EUnomia

If you have any links, info, whatever to add then please just yell at me to add it and I'll hop right on it (or edit it yourself, if you're so inclined)

I want to provide a quick but comprehensive and referenced resource to share with people in the future to warn and educate them on the topic of EUnomia.

So far, it could probably use some hashing out on details, but so far the jist is there. Looking forward to any responses... I feel like it's bare but I need to take a small break.
@mangeurdenuage @coyote Are they illegally downloading and reproducing our copyrighted content, as an essential component of a proprietary algorith used and sold to some third party to the tune of 20 million dollars

I don't recall granting them a license to our content. And I know for a fact that many people explicitly brand their writing as CCNC. Are they doing the due diligence to make sure they have licensed all of the proprietary content that is visible on a public timeline API?
@fluffy @coyote
Depends on the default license applied on the server. The licenses that are an issue are either proprietary licensed content (because they have to ask each user for consent) or copyleft licensed content like the BY-SA creative commons which would requires for them to use the same license for the results of their work.
@mangeurdenuage @coyote hmm. it's not necessarily that simple. Wasn't there a big SCOTUS case recently where it was concluded that a neural network is not held by the copyright of the content that went into it?

Still, they do not have a license to reproduce posts from my server, even if they showed up on the "public timeline" (as if that means anything). Unless SPC assigns copyright to the server and then releases it, your posts should still be copyrighted you, and they are not permitted to reproduce it for scraping purposes.

Over under on if it's worth filing a lawsuit CC @p @graf
@fluffy @coyote @graf @p
>it's not necessarily that simple
The issue at this point is that we're the ones that have to prove that they're using data against our license/consent.

>Wasn't there a big SCOTUS case recently where it was concluded that a neural network is not held by the copyright of the content that went into it?
There's a similar topic in the GDPR, there's a clause in which you need to state specifically that you don't consent your content to be used in, scientific research, AI etc....
@mangeurdenuage @fluffy @coyote @p the issue is because of the size of poast people look up to us for guidance for stuff like this and I'll be honest I have no idea how we are going to approach it yet but there needs to be accountability and we will try to figure out a means to achieve that. perhaps through an email chain with several fedi admins personally petitioning to exclude us with the caveat we can escalate if need be
@graf @coyote @fluffy @p There's a need for real legal advice and the only people coming to my mind right now that maybe able to help are the EFF.
@icedquinn @mangeurdenuage @coyote @fluffy @p EFF doesn't care about us because Eugen is the flagship instance. poast is very quickly approaching half of their active users monthly but if we say anything the only people it matters to is some of you guys.

earnestly the only way this will have any kind of effect with pushback is if we agree to put differences aside and approach him personally via email with our concerns and a request to be removed and all data collected be disclosed.

poast has legal representation its users have paid for so we can generate it but I would need you guys all to be on board with that too.
@graf @icedquinn @fluffy @mangeurdenuage @p What do you need help with? What can I do as a nobody with no particular talents?
@coyote @graf @fluffy @icedquinn @mangeurdenuage @p Same. I'm just a single user instance, but I was unaware this type of stuff was happening.

I personally don't have issues with scraping public posts; that's kinda the nature of Fedi, the Internet and search engines in general. But I do not like the idea of using any type of algorithm or ranking system to editorialize a post, especially as true or false. That just sounds like algorithmic defamation.
@djsumdog @coyote @fluffy @graf @mangeurdenuage @p the scraping of public resources to train bots has been going on for quite a while now. people are kind of paying lip service to whether its okay to do this or not. i think it's kind of a pointless debate.

like everything else this generation, it's about doing an end-run to get the toys in place before people are able to do anything about it.
@icedquinn @djsumdog @coyote @fluffy @mangeurdenuage @p this isn't the issue. the issue is this is an EU funded attempt to train AI to tell people something they don't like is wrong. you need to focus on the end goal not the path to get there
@graf @icedquinn @coyote @djsumdog @mangeurdenuage @p If the guy just did it to get a payday for once in his life I don't expect him to back down because of a letter no matter what we put in it
@fluffy @coyote @djsumdog @icedquinn @mangeurdenuage @p but his entire flagship is behind it. it's built on mastodon. it's very easy for someone with a lot of push, like poast and it's very vocal users, to put pressure on that publically

@graf None of his instances even federate with Poast, do they? I would be surprised if you were not suspended on his instances. @fluffy @coyote @icedquinn @mangeurdenuage @p

Sign in to participate in the conversation
Game Liberty Mastodon

Mainly gaming/nerd instance for people who value free speech. Everyone is welcome.