If you want to be pedantic, building the uranium bomb was easy - it was just a gun that fired one subcritical mass into another; the technical challenge was ref

If you want to be pedantic, building the uranium bomb was easy - it was just a gun that fired one subcritical mass into another; the technical challenge was refining enough of the required isotope, and that really boiled down to organisation and resourcing. The plutonium bomb required creating a sphere that would implode uniformly enough to trigger the reaction; this was significantly more complicated, but again a fairly bounded problem, probably the most significant challenge being the lack of cheap and available computing power to run the numbers. The point is that the science was already almost completely known, the Manhattan project was primarily an engineering challenge. We are nowhere near that stage of maturity in software development yet.