So the system is solving a high-res inpainting puzzle that - if filtered - looks similar to a low-res input? The results are impressive because our brain can‘t

So the system is solving a high-res inpainting puzzle that - if filtered - looks similar to a low-res input? The results are impressive because our brain can‘t do this quickly. They contain absolutely no additional information but they _seem_ to do, so this may lead to much harm.