In the last paragraph of a Scientific American article of July 1990 that can be found here
http://www.math.ucsd.edu/~ronspubs/90_06_ramsey_theory.pdf
Graham and Spencer write
"Today we can easily recognize the constellations in the night sky as a consequence of Ramsey theory."
Really? What exactly is done and how? Please could someone provide some details and/or references.
I wouldn't take the comment too seriously, but I think it's just saying that for a large and quasi-random set of 3D points (stars), there is a 2D projection (our perspective of them), which contains subsets of stars that form different patterns (constellations). It's a bit of a stretch to say that this is Ramsey theory (because constellations aren't really straight lines or any consistent geometric shape, and it's not clear what is being proved here (eg is it any 2D projections that would do this)) - I would say this is more of an indication of evolution to create over-active pattern recognition in humans!