AI doesn’t “see” the way you do, and that could be a problem when it categorizes objects and scenes

Even without fur in the frame, you can easily see that the photo of a hairless Sphynx cat is a cat. You wouldn’t mistake it for an elephant.

But many artificial intelligence vision systems would do that. Why? Because when AI systems learn to categorize objects, they often rely on visual cues – like surface textures or simple patterns in pixels. This tendency makes them susceptible to confusion caused by small changes that have little impact on human perception.

A visual system more closely aligned with human perception – one that emphasizes shape, for example – could still confuse the cat with another similarly shaped mammal, such as a tiger; but it is unlikely that it is an elephant.

The types of mistakes an AI makes reveal how it organizes visual information, with potential limitations that can be concerning in more demanding environments.