I'm hoping, as the "duck and cover" generation dies out, we might see a renaissance in nuclear engineering. Some revival is going on at the moment, but civilization only worked on nuclear ideas for about 30 years (1940-1970). We barely touched the surface. I imagine Musk would have a nuclear rocket division, if that was politically acceptable. Maybe a nuclear technology ecosystem is only possible off Earth.

The other problem is that groundbreaking ideas take increasingly more genius and luck.

This has always been true.

Therefore any kind of work aimed at making peak mental performance better would be the most important, to start cascade of improvement and fuel it.

It seems that the commercialization impact of R&D spending is more complicated than I had thought (https://www.natureindex.com/news-blog/measuring-the-impact-o...), but I would still encourage you to look at how much various places spend on R&D, and how many people they manage to move through STEM university programs (in terms of percentage of their own native-born population).

I think that making people constantly scramble to meet the next deadline puts a harmful damper on innovation and Big New Ideas. https://www.theguardian.com/science/2013/dec/06/peter-higgs-...

We also may need to spend more effort explicitly targeting our R&D industry towards finding mysteries and solving them. Basic scientific advancement, after all, comes precisely from exploring the unknown rather than from merely innovating within the bounds of the already-known.

And then there's patent reform and such to think about. Blah blah.

But you should take everything I say with a grain of salt, since I work for an R&D company.