Some things are moving fast, though; solar comes to mind as an excellent example of an area with exponential price drops. At any given time, some things are improving fast, and most things are improving slowly or not at all.
My feeling is that there are a lot of basic tech improvements just on the horizon, we start figuring out more ways to bring machine learning into hard fundamental problems of industrial design, medical diagnosis, and so on.
Which of course leads to inevitable discussions of displaced highly skilled labor; it's not clear that the areas of fast improvement will lead to a good distribution of positive outcomes, as we've discussed here so many times before...
On the whole, though, I agree with the other three points.
Funny how 'subsidies' is a bad word and 'VC funding' isn't.
The idea, in this case, is exactly the same: Invest enough to achieve the economy of scale, and get the actually-better solution to the point where it's cheaper than the happens-to-be-currently-dominant solution.
(obvs, subsidies can also be used for other purposes, like protectionism or ensuring that a vital-for-national-security industry isn't outsourced. But I believe that for solar it's about investment, more than these other uses.)
I like this comparison, actually. But realize that it is expected that most VC fundings fail. Subsidies is a bad word because it is often an attempt at rigging the game. With the assumption that we fully know what the answer should be.
So, if we can move subsidies to be more like experiments where it is expected that a significant percentage fails, I'm all for it.
My feeling is that there are a lot of basic tech improvements just on the horizon, we start figuring out more ways to bring machine learning into hard fundamental problems of industrial design, medical diagnosis, and so on.
Which of course leads to inevitable discussions of displaced highly skilled labor; it's not clear that the areas of fast improvement will lead to a good distribution of positive outcomes, as we've discussed here so many times before...
On the whole, though, I agree with the other three points.