Or we are going to be training Junior engineers with outdated skill sets and build outdated infrastructure.

Sometimes top down planning works, but I think I would personally need to trust the leadership and the incentives.

Unfortunately incentives on these are typically to do bare-minimum so you can meet the criteria to get subsidies. Sure the chips will be usable, but you spent a multiplier of the cost it takes to get it, and the chips aren't cutting edge.

I'd a bit more interested in some moonshot idealism than just some military buildup.