[–] 4 points 17 hours ago

Online translators seem to have gotten a lot better! Your comment, translated to Mandarin and back:

是的,我也会这样做。每次我尝试分段落写东西,收件人最后都会有一半内容没有回应。感觉他们读到第一个问题就停止阅读了。

这种 AI 扩写和总结让我想起了早期网络翻译刚出现的时候:你把一段话翻译成另一种语言,再翻译回原来的语言,结果得到的内容却完全变了。

Yes, I do this too. Whenever I try to write something in paragraphs, the recipient inevitably ends up not responding to half of what I wrote. It feels like they get to the first question and just stop reading.

This AI expansion and summarization reminds me of the early days of online translation, when you could translate something into another language and then translate it back, only to end up with something completely different.

  • source
  • parent
  • context
  •  

    cross-posted from: https://lemmy.world/post/40825996

    Digital artists and illustrators are abandoning Elon Musk’s blogging platform X (formerly Twitter) over the introduction of a new controversial image-editing feature powered by artificial intelligence (AI). According to reports, the creators claimed that the new tool could be used to modify others’ works without their consent.

    One such popular creator championing the boycott is Mu-jik Park, the renowned South Korean artist known by the pen name Boichi. Boichi is the creator of the hit manga series Dr Stone and Sun-Ken Rock.

     

    My understanding of federation is that it's like email. If one server is misbehaving, then they get defederated.

    So how come email spam still exists - why don't spammer domains get defederated? It seems like we've got the worst of both worlds, where it's hard to get your emails relayed when you run a small email server, and easy to get them relayed if you're a spammer.

    Is there anything about Lemmy's architecture that will prevent this problem?

    view more: next ›