Love me a long dog but this is either AI or some fucked filter
AI.
The first image is near perfectly centered and level which isn’t a dealbreaker but is still something AI does that real images can’t do very well.
Second is the depth of field. I’ve noticed in Sora and other AI video generators, this same depth of field blur is always used. I suspect it’s because AI don’t do well with fine detail so blurring it away is a good way to avoid those errors when one wants to fake images of the same environment.
Now you can have depth of field from even a phone camera but it’s not common and it’s uncommon that it looks this good.
The next issue is how clean the textures are. The hat is perfectly smooth. Is that a real posssibility? Sure, but shine almost always shows tiny scratches, also why is the lighting in the wrong place? Why is it reflective/shiny on the sides but mat on the front?
Another interesting thing is the material of the strap. Hell, even the placement of the strap is odd. Is the strap on the left in front of the ear or not? Are there two or just one?
Next, Who has a head that size? Even a child’s helmet would likely be larger. You’d need much more depth if this was supposed to cover part of someone’s forehead. I’d imagine a real helmet on a dog wouldn’t sit very well and would tilt to one side due to gravity unless it was strapped on tight which this one isn’t.
Then again I’m not an expert on helmets… Oh wait there’s a brand name maybe I can look it up… yeah Bell helmets does have that logo, but not that helmet or anything very close to it. Mostly just BMX stuff.
Anyway, definitely AI, but isn’t it kind of cool/scary we live in a world where it is getting so difficult to tell what is real and what isn’t? How long before it becomes impossible to tell?
The background is fairly consistent between the two images though, even in the blurred out of focus area. Maybe the photos were edited using an AI model to add the helmet (rather than buying a helmet fit for a dog)?
Have you seen the videos made with Sora? The background can stay consistent even when in motion so long as it’s blurry. It’s very impressive
I did not think about taking frames from a video. Yeah, that can be the case too.
Cute! But “Toto Airways” involves a lot of collisions with random objects… am afraid this is under-armored.
Aye eye
Aww Bell looks cute.
AI images really are getting quite good nowadays aren’t they?
I’m sure this is an old meme.
Idk about old but it seems possible, looks like the original wasn’t AI:
Edit: you are right about old, found one that’s 8yrs old
