Oh, yeah--definitely! It's a game, and a tool for visualization.
I got a lot of blank backgrounds on the images I generated, almost as a default. Maybe if you specify "on a white background" you'd get a workable image. A guy I know did that so he could overlay the character image on a background for consistency in making a multi-panel strip. The first time he tried generating the full scene, the AI consistently misspelled words in the banner he has on the podium in the foreground.
I'm happy you're experimenting with this, 2cross. You might get bored with it, like I did, because getting something "exactly what you want" is so very difficult, as you observed about having to fix mistakes in Photoshop.
I think if people actually worked with this and saw what it can and can't do, they wouldn't have such a harsh reaction to it. "Ugh! Fire--bad!!!"
In my opinion, there will always be a market for good art created by real, live artists. No computer program can replicate the magic of people like Rukis and Dark Natasha, Zummeng and Margaret Carspecken, Phil Foglio and Tracy Butler. Original art is difficult to replicate, and the way an artist goes about designing and creating a composition is all about genius and soul.
Of course, there's also loads of generic and soulless art out there, made by live people--think of the covers of romance novels--but my point is, that I really don't think the artists who have year-long waiting lists need to be the slightest bit worried about our new robot overlords stealing their business.
I even tried plugging in "in the style of Looney Tunes" and "in the style of Walt Disney" into some prompts, and got nothing that even resembled that. So there's not much to worry about there.
Oh, yeah--definitely! It's a game, and a tool for visualization.
I got a lot of blank backgrounds on the images I generated, almost as a default. Maybe if you specify "on a white background" you'd get a workable image. A guy I know did that so he could overlay the character image on a background for consistency in making a multi-panel strip. The first time he tried generating the full scene, the AI consistently misspelled words in the banner he has on the podium in the foreground.
I'm happy you're experimenting with this, 2cross. You might get bored with it, like I did, because getting something "exactly what you want" is so very difficult, as you observed about having to fix mistakes in Photoshop.
I think if people actually worked with this and saw what it can and can't do, they wouldn't have such a harsh reaction to it. "Ugh! Fire--bad!!!"
In my opinion, there will always be a market for good art created by real, live artists. No computer program can replicate the magic of people like Rukis and Dark Natasha, Zummeng and Margaret Carspecken, Phil Foglio and Tracy Butler. Original art is difficult to replicate, and the way an artist goes about designing and creating a composition is all about genius and soul.
Of course, there's also loads of generic and soulless art out there, made by live people--think of the covers of romance novels--but my point is, that I really don't think the artists who have year-long waiting lists need to be the slightest bit worried about our new robot overlords stealing their business.
I even tried plugging in "in the style of Looney Tunes" and "in the style of Walt Disney" into some prompts, and got nothing that even resembled that. So there's not much to worry about there.