Raw LLM Responses

As an artist, who relies on traditional media, digital and AI. I can tell you that it is not that easy. Most recently, for an ad campaign, I was tasked to have a little girl, taming a lion in a circus, while in the background, dozens of acrobats wearing specific logos on jumpsuits, were performing. I am not given a budget to hire acrobat performers to be photographed, nor a budget to hire a CGI firm for animation. While you may think that it is as easy as just typing in a google search. I on the other hand, need repeatable results, with specific details. Not just random garbage from a generic model for public use online. To do this, I actually went to four local circuses, and photographed acrobats performing, then took those images, and compiled them a model data set. I then created a model for the jumpsuits, by creating a 3d rendered jumpsuit and placed it through an animation suite to capture basic movements and poses, and created a model set of the companies logo. I downloaded model sets for lighting and enviromental effects, created by other people. and created a database of wire frame poses i specifically desired the performers to display. Through the magic of coding in Python Script, (Which i also had to learn, in order to do this kind of work) was able to use all these meticulously crafted data sets, to generate an image. Actually 312 of them. Each generation being redefined, and tweaked through prompt and negatives, till it produces the desired results. And now, i have a reliable model set, which will generate the kind of repeatable imagery needed for the clients series of ads. As far as the little girl and lion, I sketched those by hand, then scanned them into adobe illustrator to refine and finish.