in the famous pre-LLM-bullshit video of Miyazaki where some misguided joker tries to pitch procedural animation as a labour-saving device to, predictably, one of the people on the planet least receptive to it, he is met with (to me) ultra-cathartic disdain. in this post i relate that to the fake art generated by what we now laughingly/resentfully call “ai”
the junk that was being pitched in the video wasn’t on the level of what’s happening now, and was not at all the same thing technically, but it was the same thing creatively: cranking out some gross-looking shit that can never have meaning because it was created in an inherently thoughtless way, pitching that as a replacement for (or augmentation to – just as unwelcome) emotionally-involved human work, and in this case pretty tactlessly doing it in a way that feels mocking to something important to him (a person who has lost control of their body), which i imagine he wouldn’t have an issue with a tasteful/thoughtful/human depiction of.
anyone thoughtful can see immediately that there’s no application of that technology that makes things better instead of worse, and i don’t think you need technical expertise to correctly deduce from the animation that there is nothing to it. miyazaki’s is the reaction i would expect and hope for from someone whose thing is huge time investment in thoughtful art.
i don’t hate procgen, i do procgen, but there’s definitely a trap you can fall into of trying to generate something, not in order to better serve the needs of some creative project, but in the hope of generating something, anything that you can then, working backwards, invent a project whose needs are served by it; and then hopefully automate every other process involved in that project (which you haven’t the moxie to do, or you wouldn’t be here), with the end goal of reducing the effort involved in producing a work to zero, which is an arsehole goal – although you don’t have to be an arsehole to have it for a while.
because procgen work requires actual thought and process and technical ability, people can come out the other side of that trap with thoughts about why it doesn’t work, or at least aware that it doesn’t – if you’re lucky, even understanding that taking all the work out of a process also takes all the art out of it. but the AI/ML people with the same arsehole goal have no ground beneath them to get traction on, no process of iteration; they are stuck smashing their heads against the same wall forever. if you’re that invested in investing nothing, you might never figure out you’re getting ripped off
Leave a Reply