• 2 Posts
  • 2.53K Comments
Joined 11 个月前
cake
Cake day: 2024年8月21日

help-circle


  • it still is, which is why shit like this is so crazy because it happens a lot.

    the model keeps basically nothing of the original image, less than a single bit per work it ingests on (the LAION-B dataset is almost 6 billion images, most models have more than double the input data, the models are 5-6 GB, and each image is 1024x1024 pixels), so the image can’t be in there. and yet most of these models can manage to stitch together their input data almost perfectly. it’s like the model splits images into their constituent parts and builds them back the same.

    from a technical standpoint it’s amazingly unlikely. from a human perspective it’s scummy. from a legal perspective it’s 100% plagiarism.













  • why would you take the least charitable interpretation? there is no need to be hostile.

    and the answer, of course, is that it can be, as long as the information copied is meaningful for displaying to the user.
    you’re basically asking the equivalent of whether putting things into an array is an algorithm, which of course has the answer “it can be, depending on how you put it in”. so basically, the operation you’re highlighting is not the point.




  • the reason stated in the report was that base load failed to correct. it was a novel failure case, but very much related to a grid with intermittent power tacked on.

    of course there is no technical reason it wouldn’t work, but we are in the real world, where technical solutions are not implemented in a vacuum. the grid is yet to be completely retooled for intermittent wave-following sources, and accidents like this are how we figure out what not to do.

    also “us disinformation pigfuckers”? really? the us has a vendetta against spain now?