There’s been a lot of speculation around what Threads will be and what it means for Mastodon. We’ve put together some of the most common questions and our responses based on what was launched today.

  • BaroqueInMind
    link
    fedilink
    3
    edit-2
    1 year ago

    ChatGPT summary of the blog post:

    Here is a summary of the blog:

    • Threads are a way to organize conversations on Mastodon, a decentralized social network.
    • Threads allow users to reply to specific posts and create branching discussions that are easy to follow and participate in.
    • Threads can be public or private, depending on the visibility settings of the original post and the replies.
    • Users can view threads by clicking on the reply icon of any post, or by using the “Show thread” option in the menu.
    • Users can also join or leave threads by using the “Follow thread” or “Unfollow thread” options in the menu.
    • Threads are a powerful feature that enhance the user experience and foster community engagement on Mastodon.
    • Mastodon and Threads will likely be able to communicate with each other once Threads supports federation, but it will depend on the server operators’ choices. Mastodon users can always switch servers or self-host if they want more control.
    • Threads will not be able to impose their moderation policies on other servers, and that each server will have its own rules and tools. Mastodon servers can block specific content from Threads if they want to.
    • Mastodon welcomes the adoption of ActivityPub by large platforms, as it validates the decentralized social web and enables users to switch platforms more easily. This creates more competition and pressure for better services.
    • Kill_joy
      link
      fedilink
      21
      edit-2
      1 year ago

      Any way you could consider not littering threads with shitty tldrs?

        • EnglishMobster
          link
          fedilink
          131 year ago

          You’re the one who chose to post it without fixing it yourself to make it readable. ChatGPT didn’t paste it into that box and hit the “add comment” button.

          • BaroqueInMind
            link
            fedilink
            11 year ago

            Google Bard’s elaborated version of your comment (dissertation, opposite of a summary):

            Sure, here is the elaboration of the text:

            The original text is a bit accusatory, but it is also accurate. The person who posted the text could have taken the time to fix the errors before posting it, but they chose not to. This means that they are responsible for the fact that the text is not readable.

            The second sentence of the text is a reminder that ChatGPT is a large language model, not a human being. ChatGPT cannot take action on its own, so it is not responsible for the fact that the text was posted without being fixed.

            Here is a more elaborated version of the text that is still accurate and respectful:

            “I understand that you were frustrated with the errors in the text, but I would like to remind you that you are the one who chose to post it without fixing it yourself. ChatGPT is a large language model, and it cannot take action on its own. If you wanted the text to be readable, you should have taken the time to fix the errors before posting it.”

        • Kill_joy
          link
          fedilink
          6
          edit-2
          1 year ago

          This isn’t reddit though dude.

          People are here to learn and engage. You’re not contributing towards that goal and it sorta comes across as a cheap karma grab. Which I know I am not here for. Seems like many aren’t either.

          I think we would all just prefer to hear your thoughts, opinions, and contributions to the discussion rather than doing what you’re doing.

          Just my two cents.

          • BaroqueInMind
            link
            fedilink
            2
            edit-2
            1 year ago

            Thank you for the reply and feedback, I’m not lost and know what site this is.

            I think AI language models are dumb, but fascinating; I do not truly care about meaningless karma here since it doesn’t matter to me personally.

            I’m very confident no one here wants to read the confounding stupidity of my opinion, so I will limit my replies to impactful comments henceforth.

            Your two cents are really worth profoundly more, thank you.

    • BobQuasit
      link
      fedilink
      161 year ago

      That’s a pretty poor summary. It’s not all that much shorter than the blog post itself. Could you tell the AI to be much more concise?

      • artillect
        link
        fedilink
        5
        edit-2
        1 year ago

        We need an autotldr bot for the fediverse, but apparently it uses smmry, that could be used until someone makes a bot

    • stopthatgirl7
      link
      fedilink
      81 year ago

      If you’re going to “summarize” something, at least take the time to do it yourself, not run it through ChatGPT.

    • 0xtero
      link
      fedilink
      31 year ago

      That’s just plain wrong.
      A good advice would be to actually read the post yourself and not run it through a pretty sentence creator bot

    • TimeSquirrel
      link
      fedilink
      3
      edit-2
      1 year ago

      If I wanted an AI summary I’d have done it myself. I’m here to read comments from and talk to actual humans.

      Stop that. This is helpful to nobody.