So, recently some fediverse admins (mostly Mastodon) and the founder of Mastodon, Eugen Rochko (Gargron), where contacted by Meta/Facebook for an NDA meeting. We know nothing about it, but we’re pretty sure that it was about this project92 thing that Meta/Facebook is creating to “compete” with Twitter.

So a lot of Mastodon admins already singed a pact to immediately block any Meta/Facebook activity in the fediverse as soon as it comes up. My Mastodon instance, fosstodon.org hasn’t singed that pact and I’m pretty worried.

The following image is an screenshot of Gargron and dansup (creator of Pixelfed) talking about this. These posts were deleted, even from the wayback machine.

  • Kernel@beehaw.org
    link
    fedilink
    English
    arrow-up
    16
    ·
    2 years ago

    Sounds troubling. But not all of their actions turn out to be completely nefarious, and this could represent a genuine effort to contribute to the next generation of infrastructure. The emerging network protocol appears to offer an opportunity for both non-profit and commercial ventures.

    • darius_drake@beehaw.orgOP
      link
      fedilink
      English
      arrow-up
      23
      ·
      2 years ago

      I’m not particularly interested in anything that Meta/Facebook has ever created. Everything that this company has made or touched has been horrible/dangerous/terrible for everyone and I don’t think that this isn’t the case. Also that article that you linked is behind a paywall ._.

      • Kernel@beehaw.org
        link
        fedilink
        English
        arrow-up
        8
        ·
        2 years ago

        And that’s probably the wiser assessment, given their other activities. Here’s the relevant part of that NYT article.

        But Meta said it saw no reason to keep its code to itself. The growing secrecy at Google and OpenAI is a “huge mistake,” Dr. LeCun said, and a “really bad take on what is happening.” […]

        “Do you want every A.I. system to be under the control of a couple of powerful American companies?” he asked. […] Meta’s open-source approach to A.I. is not novel. The history of technology is littered with battles between open source and proprietary, or closed, systems.

        • darius_drake@beehaw.orgOP
          link
          fedilink
          English
          arrow-up
          3
          ·
          2 years ago

          Meta open sourcing their AI is great and hypocritical at the same time. Like yes we agree that AI shouldn’t be controlled by a couple of powerful American companies, but we still need to have full control over the algorithm that literally manipulates users internationally on a daily basis.

      • Steeve@lemmy.ca
        link
        fedilink
        English
        arrow-up
        2
        ·
        edit-2
        2 years ago

        I get this is the fediverse and we’re supposed to hate big tech and all, but this “everything Facebook has done is evil” is objectively wrong and such a Reddit take. Facebook’s biggest problem has always been the people that use their products spreading hate and misinformation and their lack of moderation at their current scale, a problem that every single social media has hit, including the fediverse (i.e. Beehaw defederating with lemmy.world). And honestly, with governments refusing to take a heavy stance on regulating misinformation should it really be up to tech companies on what can and can’t be shared on their platforms? Then there’s data leaks like Cambridge Analytica, but that was a ticking time bomb, because data privacy back then wasn’t a concern anywhere. I’ve worked in the data industry for a decade now and it was the fuckin wild west back then lol.

        And then there’s the good shit they’ve actually done. The article you replied to (not behind a paywall for me for some reason?) talks about how they open sourced their LLM AI for research purposes. Their data center designed were open sourced to help other data centers hit net zero carbon emissions (they’re a huge contributor to the Open Compute Project). They’ve open sourced a ton of tools/languages as well.

        WhatsApp is still fully end to end encrypted messaging and they’re pushing the same on messenger now (their WhatsApp site even has a whole section for local law enforcement telling them they can’t provide message data).

        Their targeted advertising has rightfully gotten a lot of scrutiny, but there’s a lot of misinformation behind it, like “Facebook is listening to my calls” and “Facebook is reading my message data”, which they’ve denied and there’s no actual evidence of. I have family with small businesses that wouldn’t have made it through the pandemic without their advertising platform.

        I don’t think that they have any place in the fediverse, honestly I’d be surprised if they wanted in on it anyways, but in my mind they aren’t any more evil than any other corporation and the “Facebook is straight evil” attitude is just an attempt for Redditors to feel superior about their social media corporation choice.

        • whofearsthenight@beehaw.org
          link
          fedilink
          English
          arrow-up
          2
          ·
          2 years ago

          Facebook’s biggest problem has always been the people that use their products spreading hate and misinformation

          Facebook’s algo drives this. It’s a choice that they’ve amplified this content.

          The article you replied to (not behind a paywall for me for some reason?) talks about how they open sourced their LLM AI for research purposes.

          Can’t comment much on this one, so I won’t.

          WhatsApp is still fully end to end encrypted

          still because wasn’t it like that when they bought it?

          they’re pushing the same on messenger now

          Now, as in, they didn’t design it that way to begin with because it wasn’t the profitable thing to do. They have to compete with iMessage, and further, they gain just by being able to tell every cop shop “sorry can’t do it bro.”

          Their targeted advertising has rightfully gotten a lot of scrutiny, but there’s a lot of misinformation behind it, like “Facebook is listening to my calls” and “Facebook is reading my message data”, which they’ve denied and there’s no actual evidence of. I have family with small businesses that wouldn’t have made it through the pandemic without their advertising platform.

          Glad your family made it. Unfortunately, though, this is the masses not understanding how technology these days really works. They don’t have to read your messages or listen to your calls because they’re doing that all over the web and through their own users. The truth is more nefarious because for most people “listening to my calls” is scrutable, while adding tracking cookies across the web or computing social graphs based on your contact info being shared without your consent by a few of your friends, or doing some ML on every photo shared is not.

          I don’t think that they have any place in the fediverse, honestly I’d be surprised if they wanted in on it anyways

          Agreed.

          If I were to guess this meeting was probably a job offer if anything lol.

          Disagree. Fediverse and it’s growth as it stands now is not good for Facebook, so they’re trying to head it off at the pass. I’d be willing to bet this meeting was a feeler for them and I hope Eugen and others are smart enough of to say basically nothing, and they’re continuing the grand tradition of embrace, extend, extinguish.

          Is Meta evil? No. They’re probably not a standard deviation away from any other org in terms of how many are “evil” but their incentives today all align to a worse outcome for humanity. It’s kinda worse - it’s a collection of incredibly smart people compartmentalized enough from the “evil” the org does. Actually, don’t know that I would say “evil” so much as “sociopathic.”

          Meta should get no passes, and be met with absolute scrutiny related to the fediverse.

          • Steeve@lemmy.ca
            link
            fedilink
            English
            arrow-up
            1
            ·
            2 years ago

            Facebook’s algo drives this. It’s a choice that they’ve amplified this content.

            Facebook’s algo drives interaction, human interaction with this content amplifies this content, just like pretty much every other algorithm. Ragebait sells unfortunately, but this isn’t anything new. Seen any news channel since the 24 hour news cycle was introduced?

            still because wasn’t it like that when they bought it?

            This is a response to “everything Meta touches turns to shit”. WhatsApp is a fine example.

            Now, as in, they didn’t design it that way to begin with because it wasn’t the profitable thing to do. They have to compete with iMessage, and further, they gain just by being able to tell every cop shop “sorry can’t do it bro.”

            No, because it wasn’t a priority when Messenger was built in 2011 lol. Don’t rewrite history with what you know today, I don’t know if you remember the internet back then, but it was the wild west. Data privacy was not a concern to many.

            The truth is more nefarious because for most people “listening to my calls” is scrutable, while adding tracking cookies across the web

            This one is more subjective, because to avoid the tracking you can just not use Facebook. If they don’t have your personal information they aren’t tracking you, they only know some person somewhere interactioned on a site that has Facebook Pixel. They don’t even collect your IP address, hence why they use cookies for tracking here.

            Not to mention this is all to sell targeted ads, so “evil” would be dramatic imo. Just a company doing what a company does.

            or computing social graphs based on your contact info being shared without your consent by a few of your friends,

            This isn’t true until actual proof is provided.

            or doing some ML on every photo shared is not.

            Also subjective, just don’t use Facebook if you believe this to be the case. Their ML models don’t care about you if it can’t target you with ads.

            Is Meta evil? No. […] Actually, don’t know that I would say “evil” so much as “sociopathic.”

            “Meta is evil” is what I was discussing. Sociopathic sure, just like every profit driven corporation, but to change this we’d need to unwind our society from profit driven entirely, which I’m all for, but that’s a whole other conversation.

            I’m just really hoping that this place doesn’t get obsessed with Facebook ragebait like Reddit did.

        • PopOfAfrica@lemmy.one
          link
          fedilink
          English
          arrow-up
          1
          ·
          2 years ago

          Correct, they are no more evil than any other corporation…

          Which still makes them incredibly evil.

          • Steeve@lemmy.ca
            link
            fedilink
            English
            arrow-up
            1
            ·
            2 years ago

            They aren’t Nestle levels of evil imo lol, but yeah I think the problem with most corporations in general is the total lack of regulation and oversight. Ultimately social media is a tool, and allowing anyone to use it for any purpose is going to have shitty unintended consequences. But how do you reign that in at this point without the masses revolting at their loss of free speech? We’re in a mess I don’t know if we can unwind.

    • soundasleep@kbin.social
      link
      fedilink
      arrow-up
      11
      ·
      2 years ago

      Keep in mind XMPP had similar sorts of activity back when chat apps were the rage, and in the end the protocol was added to Google Talk (now dead), AIM (now removed), Facebook (now removed), and Skype (now removed). https://en.wikipedia.org/wiki/XMPP#Non-native_deployments

      I suspect existing orgs will want to contribute just as long as it takes to steal users and build a garden, that they can then wall off.

      • polygon@kbin.social
        link
        fedilink
        arrow-up
        6
        ·
        edit-2
        2 years ago

        The Silicon Valley way of doing things is “growth at any cost”. Of course Meta wants in on what might turn out to be the next big thing. Of course they want to use money and power to dominate the protocol, insert all sorts of monetization, and ruin the whole thing. And when it doesn’t work out because they’ve done the same dumb shit that already ruined Facebook and Reddit the protocol will have been destroyed and rendered useless. Meta goes back to Facebook and Instagram while the entire Fediverse project becomes defunct.

        This is the history of these companies. Thankfully “fediverse” is not something Meta can just outright buy and then destroy, but they can still throw their weight around with cash and the enshittification will quickly ensue. The Fediverse needs to resist. It’s hard to say no to money, but VC capital is what is destroying the internet. We need to do this differently if we want it to succeed.

      • flatbield@beehaw.org
        link
        fedilink
        arrow-up
        4
        ·
        2 years ago

        Yes. I was thinking the same thing. XMPP is a great example. They could have all federated, but then they chose not to. This will be the exact plan presumably.

      • killick@kbin.social
        link
        fedilink
        arrow-up
        2
        ·
        2 years ago

        I see you have been paying attention. Past behavior is the best predictor of future behavior, so I would say your prediction is right.