> what the lamp from the second image would look like on the desk from the first image
The lamp is put on a different desk in a totally different room, with AI mush in the foreground. Props for not cherry-picking a first example, I guess. The sofa colour one is somehow much better, with a less specific instruction.
It looks like the same table, but some of the legs are missing. Parts of the lamp are also missing. And the scaling of it looks really wrong.
What a great, awful world we will have if people really start making decisions based on these results. I'm curious if in some years we will have people who fancy AI-trash chic seriously..
That one is an odd example.. especially since image #3 does a similar task with excellent accuracy in keeping the old image intact. I've had the same issues when trying to make it visualize adding decor, it ends up changing the whole room or furniture materials.
The lamp is put on a different desk in a totally different room, with AI mush in the foreground. Props for not cherry-picking a first example, I guess. The sofa colour one is somehow much better, with a less specific instruction.