• 1 Post
  • 15 Comments
Joined 1 year ago
cake
Cake day: June 5th, 2025

help-circle

  • I also thought the same, but the more I watch it, the less it feels like AI. The dangling thing in the handle doesn’t change speed or disappear after she gets in front of it, and the speed changes accordingly when the treadmill lurches a little forward. I thought that her legs clipped through the thing dangling from the roof, but she actually just kind of hugs around it. The physics of her jump feel slightly weird but that could absolutely be just me overthinking.

    The video is ridiculously grainy so it’s hard to tell for sure whether it’s AI or not, but that getup and the fact that she’s doing that this casually towards fucking stairs makes me think that it’s nevertheless an edit of sorts.





  • As for why it looks like crap, tbh it’s likely just because our letters are very thin. Vertically, an l would take the same space in the line as an ー which makes absolutely no sense and makes the lines look like enshittified junk food. On the other hand, the characters in the languages you showed are squarish, which means that they look good in both orientations.

    As for why it’s hard to read, as someone who’s learning japanese, it’s likely because people are just not used to it. While I started reading R-L fairly early in my learning journey and had no problem getting used to it, I had a friend who only started reading R-L after he was already pretty much fluent in the language, and he told me that he found reading it a little tricky despite reading L-R no problem. I assume that he read words, not letters, and reading in an orientation he was not used to threw him off. On the other hand, I was still learning so I read each character individually, which was why I didn’t feel much of a difference.







  • This article gives me vibes that someone wrote a few lines outlining the situation and asked the AI to write the article itself. Interestingly though, I think most people would just rather read the outline, less time wasted and less llm.

    A part that screams AI would be:

    This wasn’t subtle venue security—your biometric data became part of the artistic statement, whether you consented or not.

    “This isn’t this–it’s that” is an extremely common AI sentence structure, further exposed by the fact that the part before the em-dash doesn’t even make sense to begin with. No one was asking themselves whether it was part of subtle venue security.

    As a sidenote, sometimes I read sentences like this and I wonder “could this ever even have been written by a human?” I think that there’s a very low chance that this article didn’t have at least some amount of AI involved, but I know that somewhere out there there must be some people who actually write like this. And that’s kind of sad.

    tbh I don’t even know why I even wrote this, the entire article appears to be one big example of generic AI writing