You really do only dabble in it. It requires a LOT more than that. First it requires getting the correct model you want to use, then you have to fine tune it to whatever you want to use. Then you have to proceed to use of the correct LORA to get the style you are going for. And all of this is the easy part. Once you get the first image you like, that is when the REALLY hard work comes into. Getting that same image in many different situations. This involves a very complicated creation of a LORA and another round of fine tuning and quite a bit of Photoshop and Label.As someone whose dabbled in AI art, I can attest to this. Requires a whole of trial and error, along with finding the right wording for prompts (both positive and negative) to get something resembling what you're looking for. And I'm not at claiming to be a visual artist, as I mainly use it to get simple visuals to use as reference or to provide a visual that doesn't exactly already exist
After a lot of work, you then have a working prototype that may work in most cases, but it also may only work on very stringent circumstances.