The data shows a downward trend in interstate conflicts, colonial conflicts, and per capita battle deaths since the 40's. Civil conflicts peaked in the early 90's, but nukes aren't a factor in those.
There is other research[0] (also recently posted to HN, though it got a lot less traction than Gates and Pinker), that claims this trend may very likely not be there. It's very hard to prove.
Source: https://files.prio.org/publication_files/prio/Gates,%20Nyg%C...