AI Art and Mediums

rewrite this in my voice much shorter: Yes yes yes. AI art is easier. But those 4-5 years teaching yourself how to draw also teaches your eye. Most of art is teaching yourself how to see so you can create or recreate the areas of tone and value and such. Mediums compound. What you learn from one medium translates to others. While it only took you a couple of months to get AI art down, I still haven't yet, I guess I am just slow, but just because you have mastered a medium does not mean you have learned all the ways to artfully apply it. That is what I find so exciting about AI art. I work mostly with text now rather than images for AI generated art. I prefer oil painting for images and photography, like the pictures of my cat and art that I have like a thousand pictures I made each day every day when I was teaching myself programming all those years. But what I mean to say is that your work in one medium compounds the skills, that is, training the eye, in other mediums. To me though, it makes more sense to stick to text for AI art. It just seems like my preferred method. But that is just what I think right now. I plan on getting back into ComfyUI and maybe even do video as well, it is just that I am engrossed with the text medium at the moment. I waiver between writing and painting. I still think that diving into one medium makes me better at other mediums even though I am not working on them at the moment because of how I continue to grow as a person.

AI art is easier, sure. But those years spent learning to draw train your eye—art is mostly about learning to see. Skills from one medium carry over to others. You picked up AI art in months; I still haven’t, guess I’m slow. But mastering a tool isn’t the same as knowing all the ways to use it artfully. That’s what excites me about AI art. I mostly work with text now, while I still prefer oil painting and photography for images—like the thousands of cat photos I took while teaching myself programming. For me, text makes the most sense for AI art, but that could change. I plan to get back into ComfyUI and maybe even video. I bounce between writing and painting, but I know that diving deep into one medium makes me better at all of them—because I keep growing.

do the same with this: I got the same pushback from the art world when I started putting pictures of my cat in front of animations of close ups like this one of my oil paintings. I got the same pushback when I started drawing on a Wacom tablet and using Camtasia to screen capture and then AfterEffects to then edit and time lapse the videos into a 2 hour film composed of hundreds of hours of hand drawings on a Wacom. I think the same people who gatekeep AI/Art as not being art are the same people who did so when I started using digital mediums rather than physical mediums. People in the art world are really out of touch with reality half the time as well. The "successful" ones are often born so rich that they are really outside of the common experience of humanity and thus they oppress others with their "education" from books rather than the streets. I lived on the streets and started my life over with a set of colored pencils and DXM. I drew what I saw. Which did not make sense to people but that is why they liked it. I was born rich though, so I am no better than the other "professional" artists in the art world whose lives are parasiticly wealthy and disconnected from reality. The only difference between me and them though is that I did not stay rich. I had to rebuild everything from nothing but a set of colored pencils Chris gave me. I went and saw Santa, aka Chris, at the old bus stop the other day. Not Chris that gave me the colored pencils, he is dead, but Santa Chris, he really is Santa but he looks like a homeless guy at the bus stop, which he was for years until he moved into the same transitional housing where I lived for 8 years.

I got the same pushback when I started putting cat photos in front of my animations. Same when I switched to digital—drawing on a Wacom, screen-capturing with Camtasia, and editing time-lapses into a two-hour film. The same people who gatekeep AI art were against digital art too. A lot of the art world is out of touch. The "successful" ones are often so rich they have no real grasp of common life, enforcing knowledge from books instead of experience. I started over on the streets with nothing but colored pencils and DXM, drawing what I saw—people didn’t get it, but they liked it. I was born rich, so I’m no better than the art elites, except I didn’t stay rich. I had to rebuild from nothing but those pencils Chris gave me. Saw Santa Chris at the bus stop the other day—not the Chris who gave me the pencils, he’s gone. But Santa Chris, the real deal, used to be homeless before he moved into the same transitional housing I lived in for eight years.

now with this: So? It actually is not. But even if it were. Why do you feel the need to point it out? Probably because of some inadequacy. Probably from a fear of using AI to help yourself. Yes I use AI to make comments. I do it all the time. But it is not what you would think. Rather than simply give it what I want to create and have it created, I write long comments myself and then use the AI to summarize. I created PersonaGen which allows you to maintain a similar style in AI generated content. I use it all the time to edit what I generate to smaller more concise messages. Why? User experience. No one likes to read my long sprawling messages. They would rather read a shorter summary that still gets my points across.

So what? It’s not, but even if it were, why point it out? Probably some insecurity—maybe even fear of using AI to help yourself. Yeah, I use AI for comments, but not how you think. I write long ones myself, then have AI summarize them. I built PersonaGen to keep a consistent style in AI-generated content, and I use it to refine my writing into something more concise. Why? User experience. No one wants to read my long, sprawling messages—they prefer a sharp, to-the-point version that still gets my ideas across.

now with this: Yes. I did so. I created PersonaGen. It allows you to extract the style or Persona from a writing sample and then input any new prompt to use the same style. https://github.com/kliewerdaniel/ngpg07 Except I did not train a model. I simply used prompt engineering. The entire program can be distilled into two prompts, an encoding prompt and a decoding prompt. The encoding prompt generates a JSON response which is then parsed to values in a database. Then I used a Vite frontend for a UI which allows you to tweak those values so you can adjust the output generated. Now I have found even better ways, such as using RAG. https://github.com/kliewerdaniel/KnowledgeGraph01 This is what I am going to implement for the next generation. So you can upload your content and use the knowledge graph instead of the basic sqlite I used in ngpg07 version of PersonaGen.

Yeah, I built PersonaGen. It extracts a writing style from a sample and applies it to any new prompt—no model training, just prompt engineering. The whole system runs on two prompts: an encoding prompt that outputs JSON, which gets stored in a database, and a decoding prompt to apply the style. I built a Vite frontend to tweak values and fine-tune the output. Now, I’m moving to something better—using RAG. My next version will replace SQLite with a knowledge graph for storing and applying writing styles. You can check it out here: 🔗 [PersonaGen (ngpg07)](https://github.com/kliewerdaniel/ngpg07) 🔗 [KnowledgeGraph01](https://github.com/kliewerdaniel/KnowledgeGraph01)

Yes, I did. I built PersonaGen—a tool that extracts a writing style from any sample and applies it to new prompts. I didn’t train a model; I just used prompt engineering. The whole thing boils down to two prompts: one encodes your text into JSON for a database, and a Vite UI lets you tweak it. Now I'm exploring even better methods with RAG (https://github.com/kliewerdaniel/KnowledgeGraph01) for the next-gen version, so you can upload your content and use a knowledge graph instead of the basic SQLite from PersonaGen (https://github.com/kliewerdaniel/ngpg07).

now with this: False. I already learned how to communicate. I just use AI to improve my abilities. You should try it sometime. It is thought provoking rather than ablating. You learn so much. Mostly I use AI to help me code. I have taught myself so much about AI from software programming. I am hardly lazy. Why do you think I am lazy? Is that a reflection of how you would use AI? Do you only use AI to be lazy so that is why you think everyone else that does so is also lazy? Tsk Tsk. I assume because you can only conceptualize people using AI to be lazy that you must indeed be insanely stupid.

False. I already know how to communicate—I just use AI to enhance my skills. You should try it sometime; it’s thought-provoking, not mind-numbing. You’d learn a lot. I mostly use AI for coding and have taught myself so much about AI through programming. Hardly lazy. So why assume I am? Is that just how *you* would use AI? Do you only see it as a shortcut, so you think everyone else does too? Tsk tsk. If you can only imagine AI being used for laziness, that says more about you than me.