l will discuss the somewhat magical role that K3 surfaces and their moduli spaces play in Gromov-Witten and Donaldson-Thomas theory (starting from the Yau-Zaslow formula in the 1990's and continuing to the present (including recent results with G. Oberdieck)).
