The moderator rebellion is crushed, there are no longer any third party apps competing with the official Reddit app and Reddit seems to be as popular as ever.“It’s a nice time right now,” he says. “I think we’re executing really well.” In 2024, the company plans to focus on three pillars. 1. Maximizing ad revenue 2. Charging AI companies for training on its data3 . Enabling users to make money from Reddit

  • athos77@kbin.social
    link
    fedilink
    arrow-up
    0
    ·
    11 months ago

    Thing is, he complained from the start of the protest that all these AI companies had trained their products on reddit’s data, and that not only had those companies not paid reddit for it, but that reddit had actively ended up paying for all the bandwidth and everything to transfer that data multiple times to multiple companies. He knew from the start who was really sucking up all the API bandwidth. He could’ve gone after the corporations and worked with the app developers, given them time to adjust.

    But he didn’t. He didn’t because he panicked when he realized how he had let reddit foot the bill for massive corporations, because he’s always been butthurt that no one likes reddit’s shitty app and he wanted to force them to move over, because he’s too fragile in his masculinity to admit he’s made a mistake and change course, yet he’s too much of a pussy* to actually stand up to the corporations that screwed reddit over.

    He’s a shitty person and a worse CEO.

    * term used deliberately, because he thinks of himself as this big alpha male “in charge” type but he’s really really really not.

    • FaceDeer@kbin.social
      link
      fedilink
      arrow-up
      0
      ·
      11 months ago

      I seriously doubt that each company that wanted to train an AI on Reddit’s data went and downloaded a fresh copy of the data from Reddit anyway. There were already pre-packaged dumps of Reddit’s data floating around on archive sites. Still are, so making Reddit’s API harder to use doesn’t really change anything going forward either. It may make those archives slowly get more out of date, but there are better ways to make an AI knowledgeable about current events anyway. We’ve learned a lot about AI training in the past year and I think the value that was imagined around data sources like Reddit may have actually just been a brief blip.