Response

Cole Krivak

I had conflicting feelings about AI coming into this studio, and I still do.  I think that the technology is incredible, and it's pretty wild that it works as well as it does this early in its "lifetime."   However, it does have some pretty significant problems and implications. For the most part, I think that the problems with how AI gets its source material are things that can be worked on, fixed, and ultimately resolved. Licensing AI art for usage is a slightly bigger problem but not something without a solution. And I think that if these issues are addressed and fixed, AI art will be in a significantly better position. I think the bigger issue is the people online basically telling artists to fuck off, and that this machine that can't even get the number of fingers on a hand right is going to replace their passion and often, their lively hood.  However, I don't think that artists are entirely in the right either, I think that this IS something that is going to catch on fast, but not in the ways that the prophets of doom on Twitter are saying. I think that the use of AI for stock photos and fast concept art is the most like path it's going to take, and I think that even if that industry becomes dominated by AI, there will still be room for human artists to work, where personality and nuance are needed.  I think that artists shouldn't push back against AI art in its entirety, but just against the things that are ACTUALLY wrong with it, such as questionable methods of obtaining source material and people who are using it to devalue(?) other artist work. (this is something that happens even without AI being involved and is a much larger problem.) But I do think that AI art and human art should stay in their own lanes, and have minimal to no overlap at all.  


During my creation process, I used Midjourney as a method of gathering reference images for scenes or details that I had envisioned in my head and didn't want to go scrolling through google image search to find. While it was certainly faster than manually looking for the images, it has it's own very significant (at least for me) downside.  That being if I wanted a reference for a scene, I needed to know what that scene was going to look like to get the image to reference to make the scene.  This isn't a massive problem if you're just looking for lighting reference or how a curtain folds and wrinkles, but if you want to use AI as a starting point for a larger piece, you either give it a very vague prompt and end up with a pretty generic image, or you give it something very detail, but that that point, just draw it.  Overall I think it's a very interesting tool to use, especially in the context of innovation, but the nuance and inaccuracy of it make it somewhat frustrating.


I could see myself using Midjourney to get very interesting visual precedents for projects or to help generate a mood board, but I wouldn't go much further than that.  I believe that if I can't sketch something or have some sort of vision in my OWN head then I shouldn't be able to build it.  There is just something about doing that groundwork yourself that makes the final product feel a) more satisfying to pull off, and b) like your own work and almost part of you.  I think that there is something very different about coming up with your own idea and solving those problems yourself, and building something that a machine spits out.


My thoughts are very much all over the place, and I don't think three paragraphs are enough to represent all of them, but I hate writing so I'm just going to leave this as it is.