Yeah, sounds like the same one. I just now tried the prompt that generated the image I posted in this thread at the “AI test kitchen”:
lego of a beautiful great horned owl with an x-shaped bandaid on its groin and wearing a vet cone.
And got this
I tried the prompt I used for the photorealistic image (photo of a beautiful great horned owl with an x-shaped bandaid on its groin and wearing a vet cone) and got very similar underwhelming images.
For now nobody can touch Bing/Dall E 3 for the level of comprehending a complex prompt. In this case it took me more than 20 tries there to get a couple of images I liked for the photo and lego owls, but that was because it wanted to wrap these style bandages around the owl’s legs with every prompt, and because it kept wanting to put the x of bandaids too high on the owl.