Ai is a hot button topic for a lot of people, especially artists of all kinds. There comes into play a question of ethics, sustainability,  intellectual property infringement, and a whole lot more. Not to mention the countless people that make it their active duty to boycott and call-out Ai users. Some of the more popular anti-Ai groups on Facebook alone have over 120k members. This isn’t a small wave of nay sayers, it’s a pretty big deal to a whole lot of people. Maybe you’re one of those people, maybe you’re not. 

Rest assured, this isn’t a class dedicated to bashing Ai, it’s not an anti-Ai statement of any sort, and I’m not here to change any minds or preach about the dangers of our robot overlords.
But since you are all here I assume you’re at least, in part, curious about how to create over-the-top images all on your own. I’ll go over a few tidbits regarding Ai use, but for the most part this class isn’t at all about Ai, but rather about putting in an almost obscene amount of work and care to making sure you create something you’re proud of. So with that said, we won’t linger on the topic of Ai for too long. 

One thing you’ll see in this class is the level of detail and work put into every look. This would probably be the last class to pull you away from the alluring easiness of Ai, because the truth of all this is that there are no shortcuts or easy ways out. 100% of the creativity, the planning, the followthrough? That’s ALL on you. There isn’t a program telling you what steps to take. There’s no map or general groundwork just handed to you. You have to dig deep, visualize what it is you want, and make those dreams a reality. Simply put, it’s a whole lot harder than typing in prompts or booting up a computer program. Not only will each session be a lot more work, but it also takes a level of skill and practice, so you might have to continuously experiment before you find your groove and start producing images that truly align with your visions. So not only does it take a lot more work, it takes a lot more patience.
So, why would anyone want to do all of this when it truly is as easy as typing in a prompt at a fraction of the cost and time of doing things traditionally? For some people it’s pride or integrity, for others it may be to advance skills and carve a niche. But for me and for those like me? It’s all about the love of creating. It’s not just the destination to a finished product, but a genuine love for the journey of creating it every step of the way. 

I know this is a business conference, and I know most of us are focused on a paycheck. I personally love eating every day, as I’m sure most of you do as well. I won’t pretend that it’s easy to dive right into this kind of work and make money, but I won’t pretend that it’s impossible either. The best thing we as artists can do is to always explore new options, experiment with new ideas, and never stop learning. Investing time into truly polishing your style and skills is invaluable to you and your business. Remember, you can charge 400% more than your competitors if you are 400% better. Practice at anything makes perfect, and if there’s anything you’re truly dedicated to learning, you CAN learn it.
So what does all this extra work help you accomplish? It helps you become more in-tune with your own unique creative style, it helps you learn to follow through on ideas from visualiztion to conception, and it helps you grow a vast array of skills and techniques along the way. You’re trading time and work for skills and knowledge, which is a wonderful trade in my book! And even if you don’t end up utlizing these skills in your business, I hope you at least feel inspired to try something new, give it your all, and make something that is uniquely “you.” Afterall, there’s nothing wrong with picking up a new thing or two, and learning along the way. 

So what is generative Ai specifically, and how is it different than other forms of Ai? Simply put, generative Ai refers to using artificial intelligence to create something from “nothing”, all on it’s own, with little to no input from the user. This can be as simple as describing a scene and picking a few colors, or as involved as typing whole paragraphs about the style and feel you want in the final image. The generative Ai programs then take those text prompts and they scour the internet for bits and pieces potentially related to those styles and choices, and it chops it all up, re-aranges, and rebuilds that information into something new, finishing with an image as close to what you described as possible. Most anti-Ai sentiments are made against generative Ai specifically, not other forms of Ai. As an example, a program might have an Ai upscaling or denoise feature. This involves a process of scanning an image, learning what is there, mapping out all the different parts, and then adding or removing information as necessary to make the desired corrections to an already existing image. This isn’t generative in any way, and it only uses information you’ve created and given it to function as a more advanced version of tools we’ve had for a long time.
Generative Ai involves inputting a small amount of data and receiving back a finished or somewhat finished product. You may make a few choices, but ultimately the program itself will be making most of the final calls on stylistic choices. I’m also aware that a lot of folks use Ai to generate small bits of their final image, and these people may utlize countless programs and dozens of generated bits to create their final images. I’m not saying there isn’t also work involved with creating Ai pieces that look genuine and polished and real, but utlizing Ai effectively is also a skill all on it’s own. The truth is, while you may generate something and see no problems with it, there are still a lot of areas in which Ai struggles, and there are a whole lot of people that can spot sloppy Ai from a mile away. As an artist I’ve always believed in putting my best foot forward, and I want what I put out into the world to be as good as I can create, to the best of my ability. Ai has some pretty glaring weaknesses in terms of quality, and they become too detrimental for me personally to ignore.
That takes me into my next point, the pros and cons of Ai. The pros are pretty obvious; Ai is faster, it’s easier, and it’s cheaper. All pretty great selling points, I’ll admit.
For the cons, I’ll be ignoring social, political, moral, or ecological points and only focusing on the quality aspect. The biggest weaknesses, in my opinion, are Ai’s inability to accurately reproduce textures and noise, the inaccuracy in matching or stretching patterns, and the consistent knack for Ai to go arry and input the wonkiest, most out-of-pocket nonsense into an image.
Let’s use generative extend as an example. Say you have an image that you need to extend a bit, so you use generative fill to add those extra details and extend your background. If it worked as it’s supposed to, you may have a relatively great looking final result when zoomed out. However, when you zoom in you can quickly see that the noise and texture over the generated parts do not match at all with the rest of the image. The generative areas add their own texture or noise overlay, and that texture is very different than what is shown in the rest of your image. This means if someone zooms in on your high res, or if you print the image out in a larger size, these stark lines of texture differences can become painfully obvious, and they end up looking really sloppy. You may or may not be aware of a term called “pixel peeping”. For those that aren’t, this refers to zooming in on an image to inspect the true quality of that image. Pixel peeping is often done in competitions by judges, by moderators of photography websites looking for new work to feature, and even by your clients and local competition. Will most people seek out your work with the sole purpose of looking that closely? No, surely not. However there WILL be people who do. If someone is going to inspect my work looking for flaws, I want to try my hardest to make sure there’s nothing for them to find. I operate under the guise that everyone who ever sees my work will be meticulously inspecting them for quality and flaws. Is this realistic? Not at all! But it helps to keep me on my toes, and helps me ensure that what I put out there can withstand the peepers. As for patterns, let’s say you’re trying to do something simple, like remove a tag underneath some lace. Smaller sections might not be as obvious, but larger bits may be. Despite pulling information from the image and having access to a large area of that lace, it cannot understand the intricate structure of said pattern, so it replaces the areas with a muddled, hazy version of what it thinks the pattern “might” be,  but not what it actually is. This means up close you’ll be able to see where the edges of the generated bits don’t line up with the original image, and inside that area won’t be matching to the rest of the pattern. In the end, it can look a bit sloppy. As for wonky nonsense, I’m sure we’ve all seen images where someone was trying to swap a sleeve or something and it’ll add an extra random face, or instead of producing what you’ve told it, it’ll add something so far out in left-feild that it’s laughable and outright silly. It can be frustrating if you’re just trying to add earrings to a model and it keeps adding tiny baby faces or extra ears. For technology that is supposed to be cutting edge and revolutionary, it sure has it’s share of weaknesses and downfalls, ensuring it’s usage and mastery will be a continued source of frustration and annoyance.

​

Additional tips:

- Ai has that "look" because it's simulating "perfection". Ai pieces don't have flyaway hairs, or smudged lipstick, or things like that. Paying close attention to all the little details is invaluable when trying to create a hyper-polished image on your own. 

​

- Draw your ideas. Sketching out things is often times my very first step when I’m trying to visualize and make adjustments to an idea. Having an overall plan when you embark on an idea can help make things easier to manage and make the process less overwhelming overall. It’s easier to move things around on a drawing than it is to do it in 3d, although on-the-fly changes and adjustments are usually part of the process as well. It doesn’t matter if you can’t draw particularly well, as long as you’re able to help yourself visualize your concept, sketching out your ideas becomes and invaluable technique. 

​

- Use cheap materials for prototyping. When mapping out your ideas or experimenting with different techniques, practice of lower quality materials first. Move up to higher quality materials as your skills increase, much like getting new gear as your photography skills increase. As an example, I have a couple pieces here that were made out of high quality, durable materials, such as this bit here (corset piece worn by Violet) that is solid plastic and made out of Worbla, which is a type of plastic that can be heat formed and cools back to a hardened state. Worbla can be somewhat expensive, although fantastic and simple to work with. It creates solid, sturdy pieces that not only look good, but are made the same way as some of the big-name designers. This bit here, (skirt worn by violet) however, was made almost entirely out of foam. This is a combination of foam sheets that I have heat formed, carved, sanded, glued, and painted, and foam clay, that is essentially exactly what it sounds like, and is a moldable and formable clay that hardens into a foam state, that can then be sanded and painted like the other foam. Foam isn’t as sturdy, so I reinforce the backs of these pieces by overlaying mod podge and tissue paper or cheap non-woven fabric, much in the same technique as paper mache. It takes a bit longer when you have to reinforce pieces like this, but the cost difference is quite significant. Aside from all the beadwork and fabrics and whatnot, the foam used in both of these (blue spiked corset & skirt) is roughly $10 worth of materials, whereas the small bit of worbla for this much smaller piece is about the same cost. A tip for foam, use contact cement, not hot glue. Hot glue will drive you nuts and never offer the sturdiness and sustained hold that you need on wearable pieces, whereas contact cement fuses two pieces of foam together permanently. It’s the best adhesive for foam, by far. 

​

- Utilize knowledge you may have from different mediums. For example, something I see some people struggle with in post is dodging and burning or adding shadows & highlights in a realistic way. If you know anything about makeup, you probably know about contouring. With makeup, darker shades are added to minimize areas and highlights are added to accentuate other areas. So if I’m doing my makeup, I’m adding a slightly darker foundation along my cheeks because it’ll slim those areas out. I’ll add a darker shade of eyeshadow along my crease line because it give the appearance of more dimension. I’ll add a highlight to the inner corners of my eyes to brighten them, and a touch on my cheekbones to further slim my cheeks, and maybe even a bit on the tip of my nose just because I think it’s cute, but it also slims the appearance of my nose just a bit. I learned a lot about where to place shadows and highlights from learning how to do makeup and knowing exactly where to place my layers and precisely how much to blend each to make it appear natural. These skills also carried over to my drawings and paintings. I know what colors to select and how to shade a subject because I already knew how to do makeup. And onto paintings, a dynamic painting usually has some sort of light source, and many of them. When you’re painting light refractions into an image, you want to be extra careful to only place them exactly where they would naturally be. Studying light and how it works is the same in painting as it is in photography, and you’re really gaining two skills at once when you pay attention to your light and use light intentionally.

​

​So in closing, why bake a pie when you can just buy one? Well, when someone compliments the pie I brought, I’ll feel good knowing I provided that dessert regardless, and that it made someone happy. But if I made the pie from scratch? Their compliments fill me with pride, because I know it was my handiwork and my recipe and my ingredients that they are enjoying. But if I also grew the fruit to make the filling myself? Oh buddy you had better believe I’m going to be honored and overjoyed that I killed it, and that people enjoy something I put so much care and love into. There’s nothing wrong with a store-bought pie and there’s nothing wrong with making one from scratch, but it truly is the difference between a flavor that anyone can run to walmart and grab and the unique flavor you only get to taste every Thanksgiving when your great aunt throws down in the kitchen. And if you find that unique flavor for yourself, one that everyone pines over? That only increases the value of each and every piece, making each one more precious, more prized, and more profitable.