@mfortunato This is indeed a learning forum. Part of the problem I have with gen AI in this context is that it robs the human behind an art piece of the experience of creating that work. That means you don’t learn from it, instead atrophying your preexisting skill a little.
I personally avoid using gen AI these days, in all circumstances. I’d prefer if no one here used it as part of the creative process, but for reasons other than I’m talking about here. If you’ve found it to be a useful and fulfilling part of the creative process that allows *your* skills and *your* ideas to shine better, then I won’t argue about it now.
Perhaps this is what @gallmerci has done - found a balance between outsourcing the experience and doing it themself that they are satisfied with and has allowed them to focus on what they want to: the texturing. Without going into different concerns about gen AI, this might be ok in a strictly practice sense, where the output is not valued, only the experience texturing.
I’ll happily admit that gallmerci’s texturing is excellent and has improved on the plain mesh immensely, but it still is painted on the mesh. A thin layer on top of what was generated by the AI. Much of the character of the pirate can already be seen in the viewport shading. The texturing certainly allows the model to shine better, but the texturing wouldn’t look very impressive without the model. In this sense, it doesn’t matter that the spotlight is directed at the texturing, it can’t avoid shining on the model as well.
@FedPete
It is the era we live in now. That doesn’t mean we should use it. If we embrace it because it is emblematic of our era, we treat it as if it is an entirely neutral technology. There are very real concerns about it, whether or not you agree with them, you ought to at least consider them.
Indeed I do go to IKEA when I need some furniture. Division of labour is an important part of the world, and has been for as long as we’ve organised into communities. Especially so in our modern cosmopolitan world, where even the most basic products around us, such as a stool, have roots from all around the world. There’s simply no way for any one person to do everything.
However, that’s not what we’re trying to do here. We’re trying to learn and improve our blender skills. Perhaps that also strays to some other programs as well, and I don’t see why it shouldn’t, as long as it’s still us creating it. It may say blender in the title, but I see it as more of a collab about art, mainly digital, exploring a theme. Using gen AI seems the antithesis of that goal, as it prevents us from learning what it gives us a shortcut for. It also can’t be original, so it prevents us from truely exploring it and finding something new in the theme. I don’t think using gen AI should be accepted in this context, just as I doubt it would be accepted in a carpentry competition, if someone submitted an IKEA stool, whether or not they admitted where they got it, or even if they painted it very well.
FedPete, you also talk of it as a tool.
Using this ‘tool’ is not like using a more traditional digital tool such as Blender, where you can manually adjust the model thousands of times until it fits your artistic vision perfectly. At best, using gen AI is like being a manager of a derivative artist, where you tell them what you want, and get something back similar but not quite identical to what you wanted. At worst, it’s going to IKEA and selecting a stool. It’s plain, it’s boring, it’s mass produced. The only art the stool can truely fit into is a piece commenting on the unoriginality of it all.
Using polyhaven or Textures .com to get materials is a shortcut. Something that outsources the experience of finding and photographing textures, meaning I won’t learn how to do that. I’m ok with that, because that’s not what I want to learn. If my goal was to have a extensive library of textures perfect for my specific projects, I then wouldn’t take that shortcut, because it then would cut out the very experience I want. Additionally, there is still some challenge in finding a texture good enough for my purposes on these sites, and then adapting it to my needs. There is a skill there, of being able to take something mundane in the real world, with it’s imperfections for my useage and all, and adapt it to a new purpose in Blender.
A note here: This is the same for procedural textures in Blender. I don’t want to sit around creating a noise texture manually, because it’s random, and therefor by definition, boring. That’s not a skill I feel the need to devote time to get better at.
When someone uses gen AI for something so central to a piece as gallmerci has done, I can’t see how it is the same as using smaller shortcuts to allow you to focus on the expression of your ideas and the artistic skills you wish to develop, but a shortcut so large you all but cut yourself from the art.
@gallmerci I appreciate your honesty about the usage of AI, and truely your texturing is excellent. If the problem is time, then perhaps we should try to allow more time rather than rush our entries into AI. What is the point of a deadline if it pushes us out of the art, rather than motivate us to create something?
I found it quite small amount of verts compared to the detail.
