After all, image processing is only "image...

After looking around at several machine vision related sites, I wonder if it is really safe to write that machine vision can "do anything".
We have the strongest level of automated visual inspection technology called FlexInspector.
#Why do we call it "the most powerful"?
However, we still can't detect things that don't show up in the image. That is the limit.
After all, image processing can only make judgments based on images.
It is an obvious story, but I feel very uncomfortable to write something like "anything is possible" without explaining such risks.