Why is AI slop so easy to spot but hard to detect?

When easy to spot we spot it easily when hard to spot we just never spot it. So we conclude it's always easy. I can't remember the name of this cognitive bias.