Humans? Over-recognising the human face? Now there's a shocker.
Speaking of which....
Often times skeptics talk about people who have paranormal beliefs as seeing patterns that aren't necessarily there. We even have a word for it, pareidolia (which I invariably have to look up the spelling for). This seems to make sense because our neocortex is an incredibly powerful pattern matching engine. So if some people see patterns more easily in vague or obscure stimuli then that implies there is something different going on in their brains.
This 2002 article from New Scientist magazine reports on research conducted by Dr. Peter Brugger at University Hospital in Zurich. Using a group of 20 self-described "believers" (I wonder exactly what they believed in) and 20 self-described skeptics he asked each group to distinguish between scrambled images of faces and images of actual faces flashed upon a screen. A second trial asked the two groups to tell real words from nonsense words. As you might expect, the believers were much more likely to see a face or a word where there wasn't one.
Both groups were then given L-DOPA to increase the levels of dopamine in their brains (L-DOPA is a synthetic precursor to dopamine and is able to cross the blood-brain barrier). This led to both groups making more errors, but it also led to the skeptics being more likely to see the scrambled faces and words as real. Interestingly, the increased levels of dopamine in the believer's brains did not seem to increase the likelihood of their interpreting the spurious data as real which might mean there is a threshold above which more dopamine does not make a difference.