Stories
Slash Boxes
Comments

SoylentNews is people

posted by hubie on Monday April 03, @05:01AM   Printer-friendly

Twitter has released its content recommendation algorithm, known as "Responsible ML," to the public under an open-source license. The company says it hopes that the code will be used by other organizations to promote responsible development of artificial intelligence (AI) and machine learning (ML) systems. Twitter's Responsible ML system is used to make content recommendations to users, such as suggesting accounts to follow or tweets to engage with. The company says that the system is designed to promote healthy conversations on the platform and reduce the spread of misinformation.

Twitter's move to open-source the algorithm comes after criticism from some quarters about the company's use of AI and ML. In particular, there have been concerns about the way that algorithms can amplify harmful content, such as hate speech or conspiracy theories. Twitter says that it has taken steps to address these concerns, such as using human reviewers to evaluate content recommendations and implementing transparency measures to give users more control over their feeds.

The release of Responsible ML also comes amid wider debates about the regulation of AI and ML systems. Some experts have called for greater transparency and accountability in the development of these technologies, while others have warned about the potential risks of using algorithms to make decisions that affect people's lives.

The decision to open-source Responsible ML is part of Twitter's broader efforts to promote ethical and responsible development of AI and ML systems. The company says that it will continue to work with the academic community and other organizations to advance the field of responsible ML.

Submission created with ChatGPT


Original Submission

This discussion was created by hubie (1068) for logged-in users only, but now has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 5, Informative) by The Vocal Minority on Monday April 03, @06:14AM (2 children)

    by The Vocal Minority (2765) on Monday April 03, @06:14AM (#1299505) Journal

    Only the first one took me to a page that talked about the Twitter recommendation algorithm being outsourced.

    • (Score: 3, Touché) by Fnord666 on Monday April 03, @01:53PM

      by Fnord666 (652) on Monday April 03, @01:53PM (#1299531) Homepage
      Apparently ChatGPT just made up a bunch of links. I guess it’s not ready to take over the world yet.
    • (Score: 3, Funny) by captain normal on Monday April 03, @03:21PM

      by captain normal (2205) on Monday April 03, @03:21PM (#1299537)

      Maybe this is just another April Fools Joke, of course 3 days late as our usual timeline on stories.

      --
      "It is easier to fool someone than it is to convince them that they have been fooled" Mark Twain
  • (Score: 1, Offtopic) by Mojibake Tengu on Monday April 03, @07:07AM (2 children)

    by Mojibake Tengu (8598) on Monday April 03, @07:07AM (#1299508) Journal

    Advice of the day:

    Never trust a machine who just wants to make you happy. Even if it is a collective.

    --
    The edge of 太玄 cannot be defined, for it is beyond every aspect of design
    • (Score: 2, Insightful) by Freeman on Monday April 03, @01:53PM (1 child)

      by Freeman (732) Subscriber Badge on Monday April 03, @01:53PM (#1299530) Journal

      You are anthropomorphizing your machine. New advice: Don't attribute feelings/human characteristics to inanimate objects.

      --
      Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"
      • (Score: 5, Touché) by Zinho on Monday April 03, @03:23PM

        by Zinho (759) on Monday April 03, @03:23PM (#1299538)

        You are anthropomorphizing your machine. New advice: Don't attribute feelings/human characteristics to inanimate objects.

        Yeah, they hate that.

        --
        "Space Exploration is not endless circles in low earth orbit." -Buzz Aldrin
  • (Score: 2) by ElizabethGreene on Monday April 03, @03:48PM (5 children)

    by ElizabethGreene (6748) on Monday April 03, @03:48PM (#1299543)

    Per Mr. Musk's tweet [twitter.com] the recommendation engine source is here https://t.co/9ozsyZANwa [t.co] which unfurls to https://github.com/twitter/the-algorithm [github.com].

    As of this moment there are 200 pull requests outstanding. From a quick glance at the comments, it looks like you get a gold star or somesuch if they commit one of your pulls.

    • (Score: 0) by Anonymous Coward on Monday April 03, @07:04PM (4 children)

      by Anonymous Coward on Monday April 03, @07:04PM (#1299581)

      A gold star?! For learning the code base for a future employer??? Call me hardcore, I'm going in. Choose me Elon!

      • (Score: 0) by Anonymous Coward on Monday April 03, @08:28PM (3 children)

        by Anonymous Coward on Monday April 03, @08:28PM (#1299590)

        Yeah, but the joke is on you. The only people on the other side of the pulls are the green card workers who are stuck working as indentured servants, and they are too busy modifying the code to make sure Musk's tweets get more interaction and views than anything else. So if you want a gold star, make a contribution in that area.

        • (Score: 2) by Ox0000 on Monday April 03, @10:40PM (2 children)

          by Ox0000 (5111) on Monday April 03, @10:40PM (#1299603)

          You're just randomly stringing characters together, aren't you?
          Green Card holders are permanent residents (aka Resident Aliens), they can, and do, move from employer to employer and they can even be unemployed without repercussions for their status in the United States. Think of GC holders as US citizens without the right to vote, it's not quite like that but it's conceptually a not-unreasonable model.
          Visa holders on the other hands, and H1B's would be a prime example of this, are typically the equivalent of indentured servants. Their ability to be on US soil depends on stupid conditions, such as (for an H-1B) whether or not they are employed by someone who is sponsoring their visa. In the case of an H-1B, them losing their employment /could/ result in a "leave US soil, you have 7 days" kinda deal.

          If you're going to xenophobe, at least do it properly - but then I'd rather you didn't at all because I thought you were better than that.

          So, unless you're willing to go pick strawberries, and work at car-washes, and McDonalds, and all sort of other jobs you think you're too good to do for the money that is offered for those jobs, and which is currently done by immigrants that are willing to do those jobs, I'd strongly suggest you stop piling on or complaining about immigrants; because even though you do not see the myriad jobs they do at minimum wage, they make life as comfortable as it is for you.

          • (Score: 1, Informative) by Anonymous Coward on Monday April 03, @11:09PM (1 child)

            by Anonymous Coward on Monday April 03, @11:09PM (#1299606)

            It wasn't intended to be a knock on the visa holders (yes, you are correct, that is what I meant), it was actually intended to point out (indirectly) the fact that they are indentured servants to Twitter. The second half of it was to remind people what a shitty place it is to constantly have to cater to Musk's every whim [theguardian.com] or be fired [arstechnica.com], but their only real choice is to quit a shitty situation and go back to their country and start the whole byzantine visa process all over, or to hold their noses and stick it out. So I tried to pack a lot into those brief words in a rather smart assed way, none of which were intended to demean the people with the visas.

            • (Score: 2) by Ox0000 on Monday April 03, @11:20PM

              by Ox0000 (5111) on Monday April 03, @11:20PM (#1299607)

              Right-o, continue onward then, chap! Thank you for the explanation, this adds (to me) meaningful context to your posting and I appreciate that you took the time to clarify!

(1)