moving away from consistency...

as current AI tools develop and more people use them. i have noticed subtle consistencies that pop up in stable diffusion. finding the consistencies and adapting them will be the difference from the average user to a unique user.

one of the more obvious consistency are poses for portrait images. there are only a few poses the model stable diffusion will produce; front-on, side cut-off from mid torso and 3/4 cut-off around stomach. these poses are very static with variation coming in scale and facial features interpreted by the prompt. the initial amazement of AI creations quickly dissipates after outputting thousands of images. so making dynamic variation with portrait images will be my goal for the near future.

dynamic, movement and character will create more powerful images that can be justified as art and help remove the current stigma with how 'easy' these tools make creating...