That's empirically false. Breakthrough technologies are novel by definition, their discovery is very improbable conditional on prior thinking.