Odd that Freeman Dyson thinks politicians and administrators are particularly difficult to persuade here. This is the whole point of why capitalism works better than having clever people run a command economy.

How many American politicians and administrators do you think were actually 'persuaded' into believing that capitalism works, in the sense that you mean to use the word? It's probably more like they were born into it.

You can be clever enough to notice you need roads and infrastructure, but no one is clever enough to predict what technologies will run the future (truly, this principle applies to almost every reasonably complex thing, not just technology -- the finance angle in particular is the standard phrasing, hence me bringing up capitalism).

What the Shuttle and public infrastructure have in common is that they're projects suited to the public sector, (the Shuttle was, at least, in the 1970s-80s), as opposed to the private sector. The Shuttle was so risky and costly in the late 70s that only a national government could consider trying it. This is one niche for the public sector in a mixed economy that is largely capitalist: projects that require lots of capital and incur lots of risk. But even if it's ultimately a public project and not a product or service in a market, we can incorporate some of the benefits of market competition with the suggestions that Dyson offers. The point is that if we allow competition, then we don't need to be as clever to predict what technologies will run the future, and it seems silly that the politicians and administrators would praise capitalism and limit their projects in this way.

But it shouldn't seem that silly to us, because we know that the capitalist dogma never told them that they should apply competitive principles in public projects, and because they're tasked with doing what other people want even though they don't in their heart of hearts want to do what other people want.