Pic of the Day

Out of context: Reply #127257

  • Started
  • Last post
  • 127,325 Responses
  • sted3

    • Art by Georgy Kurasov https://www.artpeopl…grafician
    • would make an interesting AI prompt additionscarabin
    • I feel really pissed off with myself for instantly thinking ‘is that AI?’ When it’s kind of obvious it’s notIanbolton
    • AI diffusion models struggle with more than one people in a scene (distorted limbs, etc.)grafician
    • I was thinking of more of a style transference. Shitty that this is my first thought seeing the work. “This is mine now”scarabin
    • AI already has changed the way we look at and think about art apparentlyscarabin
    • At first, like Cessepe.Miesfan
    • the details on that dress. damnsarahfailin
    • Nah no model can reproduce anything this intricate lolgrafician
    • pretty sure human input + AI can do this jpeg too. but it can't reproduce the real painting, as object, or as experience, or the experience of creating it.uan
    • 2min prompts in dall-e (without naming the artist): https://i.imgur.com/…uan
    • LOL uan that's from stock photosgrafician
    • AI will 100% be able to do this as it grows. Machine learning only needs ideas, and the irony is talented people aren't giving away their ideas for free.garbage
    • Midjourney is making people pay $30 a month for their ideas.garbage
    • composition generators aren't some standalone entities they need to be proper fed and that is a pretty swampy area right now.sted
    • Yeah, right now. People don't learn / adapt/ input /produce on an exponential scale. The swamp monster will learn if it keeps getting fed.garbage
    • I've used all the new AI diffusion models, NONE can match the details in real illustrations, artworks, etc.

      1 in 100s of tries is good enough, if that
    • But nobody is admitting how many tries and time wasted to get a good result they post online after looking "amazing"grafician

View thread