Salta al contenuto
Citiverse è uno spazio aperto a tutte le comunità. Se vuoi aprire un gruppo locale o una sezione per la tua organizzazione, puoi contattare gli amministratori: pagina dei contatti.

Can someone explain to me why was Pixelfed, Loops, and Dansup being singled out for AI use?

General Discussion
12 9 21
  • Can someone explain to me why was Pixelfed, Loops, and Dansup being singled out for AI use?

    There's no shortage of Fediverse software that are either built with AI assistance (including vibe coding), or welcomes AI contribution, but I'm not seeing the same level of outrage.

    Did Pixelfed, Loops, and Dansup specifically promise to be AI-free?

    scritto su ultima modifica di
    #3

    @manlycoffee A lot of people don't like him in the first place, so he attracts more attention on this. But also his software are among the most used in the fediverse.
    As far as i know Mastodon doesn't ban AI contributions, they do have an AI policy that limits it, and didn't get as much attention.

    IMO asking open source projects to entirely ban LLM use is too far of a reach. Having an AI policy to restrict it to what makes sense to the contributors is IMO sufficient.

  • Can someone explain to me why was Pixelfed, Loops, and Dansup being singled out for AI use?

    There's no shortage of Fediverse software that are either built with AI assistance (including vibe coding), or welcomes AI contribution, but I'm not seeing the same level of outrage.

    Did Pixelfed, Loops, and Dansup specifically promise to be AI-free?

    scritto su ultima modifica di
    #4

    @manlycoffee

    Some people feel that they are fighting insidious forces (capitalism, AI, racism, etc etc) and change can only come from purity. They also can’t or don’t make anything for themselves.

    Therefore, they yell at the people who actually do stuff to be perfectly pure.

    I actually saw a presentation at Fedicon by a guy who wanted to do neighborhood peer discussion and he rejected protocol after protocol for not being perfectly distributed enough. And then he had nothing to demo

  • @manlycoffee

    Some people feel that they are fighting insidious forces (capitalism, AI, racism, etc etc) and change can only come from purity. They also can’t or don’t make anything for themselves.

    Therefore, they yell at the people who actually do stuff to be perfectly pure.

    I actually saw a presentation at Fedicon by a guy who wanted to do neighborhood peer discussion and he rejected protocol after protocol for not being perfectly distributed enough. And then he had nothing to demo

    scritto su ultima modifica di
    #5

    @neilk The last paragraph of your post perfectly sums up a very loud (but probably not very productive) part of the Fediverse. Perfect is the enemy of good etc. @manlycoffee

  • Can someone explain to me why was Pixelfed, Loops, and Dansup being singled out for AI use?

    There's no shortage of Fediverse software that are either built with AI assistance (including vibe coding), or welcomes AI contribution, but I'm not seeing the same level of outrage.

    Did Pixelfed, Loops, and Dansup specifically promise to be AI-free?

    scritto su ultima modifica di
    #6

    @manlycoffee there two groups. The group who fundamentally against any use of AI those would be hypocritical if they only focus on pixelfed . The other group have only problem with ai development, those people have a point ai dev introduce bugs at very high level.to my knowledge pixelfed was the only platform on fediverse that was willing do it

  • @manlycoffee there two groups. The group who fundamentally against any use of AI those would be hypocritical if they only focus on pixelfed . The other group have only problem with ai development, those people have a point ai dev introduce bugs at very high level.to my knowledge pixelfed was the only platform on fediverse that was willing do it

    scritto su ultima modifica di
    #7

    @hiphopheaven

    I mean, Mastodon maintainers are promising not to use AI, but they don't ban AI-generated code, so Mastodon is not guaranteed to be devoid of AI-generated code.

  • @hiphopheaven

    I mean, Mastodon maintainers are promising not to use AI, but they don't ban AI-generated code, so Mastodon is not guaranteed to be devoid of AI-generated code.

    scritto su ultima modifica di
    #8

    @manlycoffee I wasn't aware

  • @manlycoffee

    Some people feel that they are fighting insidious forces (capitalism, AI, racism, etc etc) and change can only come from purity. They also can’t or don’t make anything for themselves.

    Therefore, they yell at the people who actually do stuff to be perfectly pure.

    I actually saw a presentation at Fedicon by a guy who wanted to do neighborhood peer discussion and he rejected protocol after protocol for not being perfectly distributed enough. And then he had nothing to demo

    scritto su ultima modifica di julian@activitypub.space
    #9

    @neilk@xoxo.zone this is a pretty succint explanation.

    I'm also pleasantly surprised that this thread hasn't devolved into anti-AI vitriol. Refreshing.

    Will you be attending FediForum this week?

  • Can someone explain to me why was Pixelfed, Loops, and Dansup being singled out for AI use?

    There's no shortage of Fediverse software that are either built with AI assistance (including vibe coding), or welcomes AI contribution, but I'm not seeing the same level of outrage.

    Did Pixelfed, Loops, and Dansup specifically promise to be AI-free?

    scritto su ultima modifica di
    #10

    @manlycoffee "There's no shortage of Fediverse software that are either built with AI assistance"

    They should stop. Or you should name names so we can stop using that software.

  • @hiphopheaven

    I mean, Mastodon maintainers are promising not to use AI, but they don't ban AI-generated code, so Mastodon is not guaranteed to be devoid of AI-generated code.

    scritto su ultima modifica di
    #11

    @manlycoffee @hiphopheaven Then they should ban AI-generated contributions

  • Can someone explain to me why was Pixelfed, Loops, and Dansup being singled out for AI use?

    There's no shortage of Fediverse software that are either built with AI assistance (including vibe coding), or welcomes AI contribution, but I'm not seeing the same level of outrage.

    Did Pixelfed, Loops, and Dansup specifically promise to be AI-free?

    scritto su ultima modifica di
    #12

    @manlycoffee the thing people need to know about Daniel is he often moves ahead with ideas impulsively without thinking it all through. Thankfully, unlike a lot of techbros, he's also quick to reverse bad decisions and listen sincerely to community feedback.

    In this case, the issue is already moot, which is awesome! 👏

    https://mastodon.social/@pixelfed/117350705262301318

  • Lemmy publication compatibility check

    Fediverse fediverse
    1
    1
    1 Votazioni
    1 Post
    0 Visualizzazioni
    Nessuno ha risposto
  • 0 Votazioni
    5 Post
    0 Visualizzazioni
    julian@activitypub.spaceJ
    @deadsuperhero@fedizen.net you can approximate something a little bit using duck-typing, but it is fairly unreliable. e.g. WordPress implemented postingRestrictedToMods, because it's a blog, and the public can't make new blog posts (<img class="not-responsive emoji" src="/emoji/ap/scream/activitypub.space" title="" />) so you could filter out group actors that way. But... as @skavish@welley.social says, we'd need something new to distinguish. Any particular reason you don't want the non-discussion-groups?
  • Just thinking out loud....

    General Discussion activitypub iso
    4
    0 Votazioni
    4 Post
    22 Visualizzazioni
    jschwa1@mastodonapp.ukJ
    @webmink Two thoughts:1. The PAS process, whilst easier, would need a body to fund it. The end result would, arguably, have less 'weight' than a full ISO2. I may be wrong, but I don't believe ISO can charge to use standards, only to purchase the standard. This may mean those developing, say, a standard implementation of the standard, would need to pay to read/ access it. Those wanting to use the resulting implementation would not need to purchase anything from ISO. Knowing it was developed to the ISO standard will add credibility
  • 0 Votazioni
    4 Post
    26 Visualizzazioni
    tom_s@chambr.esT
    Hab mir mal ein kurzes Script gebastelt, um festzustellen, wieviele Nachrichtenes betreffen würde. Allerhand.Nachgeladene Nachrichten pro Tag (distinct Nachrichten):2026-09-01 92026-09-02 02026-09-03 122026-09-04 172026-09-05 32026-09-06 122026-09-07 112026-09-08 62026-09-09 122026-09-10 102026-09-11 512026-09-12 82026-09-13 52026-09-14 242026-09-15 62026-09-16 432026-09-17 7Gezaehlt werden eindeutige GUIDs; 'created item'Kann pro Nachricht zweimal erscheinen (uid:0 global + lokale Benutzerkopie
  • 0 Votazioni
    5 Post
    26 Visualizzazioni
    ?
    @aspensmonster It's not really a client bug per se, since #Mastodon philosophy is to keep the 500 char limit for everything (which I fully support, as one of the big reasons I enjoy #Fediverse (which we all once upon a time called #microblog-ing, remeber?) is that it helps me practice concision in my writing — a skill I struggle with.I think it's just a UI bug on both sides: #Tusky should warn but not stop you at 500, & web client should indicate you have to edit with a different client.
  • 1 Votazioni
    3 Post
    26 Visualizzazioni
    B
    @dansup @benpate @Bonfire @evan Looking forward to E2E direct messages in @pixelfed , @loops and @nodebb ( @julian ) . Hoping all three ship interoperable E2E by December this year.
  • 2 Votazioni
    102 Post
    87 Visualizzazioni
    openstars@piefed.socialO
    It's been a couple weeks and I wanted to add another perspective about this. I noticed a user downvoting posts, https://lemvotes.org/user/FarStare@lemmy.dbzer0.com if it helps to have specifics. In the more than one full year of existing, they have never once contributed via making a post, never once commenting even, just doling out up- and down-votes in roughly equal proportion, and even those nearly exclusively via posts but almost never comments. I am not concerned about receiving votes from PugJesus - either up or down, as PugJesus has earned the right to speak on behalf of the community. But who is this other person who is influencing how others see my content? Are they male or female? Do they have a degree of some kind? Do they primarily visit meme communities, as the vote pattern might suggest, or other communities as well? So how do I feel about people who use the Threadiverse more than others? My answer is more complex and nuanced than simply yay or nay. I might prefer the opposite of the implemented feature even - the more someone posts (including comments?), the more upvotes they are allowed to make? Have they not earned such? Also, if votes were exposed, each individual could decide for themselves how they wanted to make use of voting - perhaps someone would ban PugJesus, or conversely write special code (a 3rd party app / API recipient?) to highlight posts that PugJesus upvoted (isn't that how "boosting" works? at least re-tweeting does). It's really difficult to guess what someone wants, as each person wants different things. Which is why ultimately in an effort to make things more fair & equitable, the solution can end up making it less so. I like the follow-up suggestion to leave this new option OFF by default, so that an instance admin would have to specifically opt-in, perhaps as part of an overall suite of features limiting things like posts and comments as well. Although really, I think other approaches might work better - e.g. why would someone want to limit themselves to "some" of PugJesus' content, wouldn't they rather simply have it all or else ban him from the instance and therefore have none? I am not seeing the benefit realistically of having "some", especially when the delimiter is not based on community or overall content but simply what time of day (based on an arbitrary threshold for when they starts and ends to restart the next period) the votes are cast?
  • 0 Votazioni
    2 Post
    65 Visualizzazioni
    julian@activitypub.spaceJ
    @wjmaggos@liberal.city I think that thinking of "anything" being boostable is both weird and expected at the same time. We hold these thoughts in our heads concurrently because it was how the web-at-large (indieweb notwithstanding) worked. What I mean is... A news article isn't a discrete resource shared on a common social web — such a construct does not exist outside of the Fediverse. ... but news articles are littered with "share via..." buttons that sort of approximate the experience.

Citiverse è un progetto che si basa su NodeBB ed è federato! | Categorie federate | Chat | 📱 Installa web app o APK | 🧡 Donazioni | Privacy Policy

Il server utilizzato è quello di Webdock, in Danimarca. Se volete provarlo potete ottenere il 20% di sconto con questo link e noi riceveremo un aiuto sotto forma di credito da usare proprio per mantenere Citiverse.