This is really funny actually, considering what basic Photoshop tools are capable of out of the box :)
This is really funny actually, considering what basic Photoshop tools are capable of out of the box :)
1. Generate me an eye
2. now place 3 of these on a head where I choose
3. merge it in for me, in some way that fits the overall image.
The 4th step (missing due to time it takes) iterate over a few iterations till I see one I like.
3 years from now a novice will be able to do more advanced photoediting than a pro inside photoshop right now.
It's already the case that photoediting for novices are a reality, but the level of control is lacking.
https://lh3.googleusercontent.com/OFIpc-iFtO1AQty-IJ6beLEGe6...
He already had a good enough lantern, all he had to do is to set a tasteful setting for this.
https://imgur.com/a/SgQ38bw https://imgur.com/a/Gi0IZtB
Can you apply feather edges for me and compare the results (and let's use the gentlemen's agreement that you will not spend more than 20 seconds)?
I can't do it in the time constraint, if that's the goal of course AI wins.
Model output being defined by the training data is still a valuable observation for many unaware of the limitations, regardless of how obvious it might be for others.
But half the fun is trying to do it entirely inside the generative art tools available. If I were really trying to make something good I'd use both together.