• 0 Posts
  • 14 Comments
Joined 5 months ago
cake
Cake day: May 6th, 2024

help-circle






  • The problem comes when producing work. A generative model will only produce things that are essentially interpolations of artworks it has trained on. A human artist interpolates between artworks they have seen from other artists, as well as their own lived experiences, and extrapolate […].

    Yes, but how does that negate its usefulness as a tool or a foundation to start from? I never made any assertion that AI is able to make connections or possess any sort of creativity.

    Herein lies the argument that generative AI in its current state doesn’t produce anything novel and just regurgitates what it has seen.

    There’s a common saying that there is no such thing as an original story, because all fiction builds on other fiction. Can you see how that would apply here? Just because thing A and thing B exist doesn’t mean that thing C cannot possibly be interesting or substantially different. The brainstorming potential of an AI with a significant dataset seems functionally identical to an artist searching for references on Google (or Pixiv).

    Having someone copy your voice to make it say things you did not say is something many will be very uncomfortable with.

    So is this your main issue? I’m just not sure that that is really a valid reason, since many people are very uncomfortable with like, organ donation, pig heart valves, animal agriculture, ghostwriters, real person fanfiction, or data collection by Google. I’m sure there is something in the world that most people see as either positive or neutral that makes you very uncomfortable. For me, it’s policing.

    On the economic front, I agree - these companies should have been licensing these images from the start and we should be striving to create some sort of open database for artists so that they are compensated. It’s possible that awarding royalties, while flawed, may be a good framework since they could potentially be paid for all derivative works and not simply the image itself. But that may be prohibitively expensive due to the sheer number of iterations being performed, so it’s hard to say.


  • It’s not theft, the artist still has their work. If anything, it’s copyright infringement. When some 16-year-old aspiring artist uses another artists’ work as a reference or traces something, what’s that?

    I guess you could call it practice, but then doesn’t AI do the same thing by iterating based on its dataset? Some AI outputs look terrifying and janky - so did my art when I was younger.

    I dunno, like this issue isn’t as simple as I used to think it was. If we look outside of economics (because artists need money to survive, like all of us) is there actually a problem here?

    I’m still trying to figure out how I feel about all of this, but it’s pretty obvious AI isn’t just gonna go away like NFTs did. I really am interested in discussion, I’m not trolling.


  • So like, I guess I’m just wondering how that refutes his point that it’s a tool for artists then.

    I personally am aware of people who run local LLMs trained on their own art so they don’t have to spend as much time sketching or doing linework.

    Maybe you’re just not as open-minded about this as you could be? It’s being used in sketchy ways by a lot of people, but that doesn’t mean it doesn’t have a place, especially at the idea stage.


  • He argued that it’s just the same when an artist draws inspiration from other peoples’ art and creates their own - which is just plain false.

    Hey, can you articulate the difference though? Stating this as a plain fact seems kinda like you’re constructing reality to fit your opinion and maybe that’s what your friend is pushing back on.

    It’s true that AI is often trained on copyrighted images, but artists use copyrighted images as references all the time. I know AI can’t be literally “inspired,” or have artistic intentions, but like, what actually is the difference? Other than philosophical differences that involve like, the inability to emulate actual creativity.

    Seems like AI is just faster, because it’s a computer that can do tons of adjustments instantly instead of iterating over time like a human. Anyway, just food for thought. I don’t think AI is going to replace artists entirely but a lot of companies are definitely going to try to see how far they can take it.




  • Hmm, you know, I think I liked Workman the most. Colemak has been the most difficult to adapt to, but I’m not sure if it’s because I had been away from QWERTY for so long (Colemak has a lot of keys in common with QWERTY) or because the layout has keys in locations that I don’t find the most intuitive.

    Honestly, if I’m still hovering around 80 WPM for much longer it might actually be a good move to switch back to Workman, lol. Dvorak was probably the best IMO for reducing finger and wrist movement and strain if that’s a concern for you.


  • I’ve tried Dvorak and Workman but switched to Colemak-DH about a year ago just for fun and to try something new (I have a split ortho keyboard with XDA keycaps so the profile is uniform).

    Works great, but my typing speed dropped by about 40 WPM from the low 110s to the mid-70s and I haven’t had enough time to practice and get back up there. Previous layouts have required at least a few months of practice with Monkeytype or Keybr and I’ve been too busy.