• 38 Posts
  • 19 Comments
Joined 1 year ago
cake
Cake day: June 24th, 2023

help-circle


















  • Compared to some other Splotters, FCM was downright professional and nice looking haha.

    ikr!! FCM was one of the earlier Splotter games i encountered, so The Great Zimbabwe was quite a shock aesthetics wise for me!

    im afraid i dont have any links on hand, but i feel the complaints (at least i heard on reddit back then) were because the design for FCM looks more like a prototype than that of a finished game, especially the map tiles and “clip-art” graphics combined with the high price point. this might also have been people’s first introduction to Splotter games in general (it was for me) so that probably didnt help too

    i definitely loved the art style though, so at least I dont have to worry about the one time run of Lucky Duck’s deluxe. Horseless Carriage on the other hand… its no Ian O’Toole’s Kanban EV id give it that








  • the thing is this indie group, have been creating boardgames since before genAI models for artwork were popular. their first game in 2016 (top 10 since its release as rated by hobbyists among over a thousand other games) and subsequent expansions on kickstarter did really well even with public domain artwork that dont even look like they fit into a cohesive set. the expansion fetching usually close to a million dollars on kickstarter each time even before retail release

    what makes the game appealing in-spite of the public domain artwork have long been discussed. so to me and possibly the journalist it seems like a question why they felt the need to use genAI art now with so many successful releases without it in the past seems to come off like not wanting to pay for better than public domain artwork



  • the vast majority of employed artists aren’t making anything as creative as cover art for a hobbyist board game.

    its not just the cover art for a hobbyist board game, it is art for every card in the game. for hobbyist card games, it can go to several hundred to thousand artworks each from an artist. for a game like Android Netrunner the art of each card works with the theme and mechanics of the game acting like a brief window into this futuristic society world you compete in. (also blatant shilling, this is a great game if anyone is into cyberpunk and card games, unlike anything Magic the Gathering can ever hope to achieve), there is also graphic design for games like Kanban EV (by Ian O’Toole) which is unlike anything ive seen. boardgame hobbyists can and do regularly buy these things with quality visuals

    maybe im too emotionally invested into games but i think these art, and the art for things like beloved character design for computer games, decorative tarot cards, novel artwork which take you to another world even if just for a brief moment, is worth encouraging, putting up with Barbie Monopoly and paying for

    the alternative i fear would be these people’s time being spent instead on working soulless jobs like labelling training data for genAI models, manual work which so far only humans are cheap enough for and figuring out how to squeeze more money out of consumers


  • do let me know if im coming off as combative and this isnt the place for it, i do admit i definitely am a pessimist

    Is something that only the rich have access to right now, enable creative expression beyond our wildest imagination for all of the people who don’t have 5 to 10 years of their life to dedicate to learning art.

    isnt this possible just by commissioning an artist from fiverr or deviantart with your own prompt of an image you want. for the amount of times a person wished they had spent time learning how to draw, we would let many more companies get away with not paying artists for every piece of art available in a board/card game so they could make more money

    Sure, but we quite enjoy having prerecorded music nowadays and we would never give that up in exchange for live artists.

    would we give that up instead for genAI created music? no one has the time for 5 to 10 years of vocal training too

    Because humans like to express themselves and share that expression is widely as they can for no other reason than the active sharing and having their works seen by many.

    when genAI models can learn from art faster than a human can, art becomes a working professional artist’s only competitive advantage if they wish to live off of their work. while it may be shared, but possibly only behind a glass screen in a private gallery with metal detectors prohibiting cameras at the front, considering how futile anti-AI art filters may end up

    Why do you doubt the most pure form of art? Art as a hobby. Art as a form of self-expression?

    because people are unwilling to spend 5 to 10 years learning art as a hobby to express themselves when they can still earn some money from it as their passion now


  • just like vinyl and other vintage works, i do think it will be a shame that human produced art will become scarce and likely only for the rich to enjoy. i dont see why they would share it freely anymore

    And even if they somehow totally disappear, people will find plenty of new and exciting ways to continue to push the boundaries of what AI can do

    this assumes that genAI models can improve without any new input. but to be honest, it feels more like a, once they wipe out a generation of artist, they are free to increase the price of their “Skill as a Service” out of the reach of an average person for more profit. the GPU and water the genAI models run on arent getting any cheaper so no risk of anyone spinning up their own cluster



  • People who use AI will create a better cheaper product

    i feel like this assumes that there will still be human produced art to train on to improve the genAI model when there isnt any incentive for humans to spend so much time to learn to make art when it can be used for training and when machines can churn out pieces at a faster cheaper rate

    (c) Restrictions. You may not … (iii) use output from the Services to develop models that compete with OpenAI;

    from section 2ciii of OpenAI’s Terms of Use somehow while its justifiable for corporations to use human produced work to train a machine that competes with humans, using corporate machine produced work to train a competing machine is not