Matt Ridley graciously replies to my critique of his Julian Simon Award Lecture in the comments.  Ridley’s in blockquotes, I’m not:

First, I know of a lot of people who are not conventionally clever but
who contribute to innovation by making tiny tweaks to machines or
routines. That’s very much my point, that innovation is not a series of
intuitive leaps but often unplanned micro-steps that come together.
(Kevin Kelly’s book What Technology Wants is a brilliant addition to
this literature.) I am thinking of one guy in particular who can adjust
or mend almost anything, but could never pass an exam or hold much of a
conversation. I just think there is a strong prejudice among academic
types (of which I am one) to miss the point that most innovation
consists of bottom-up infinitesimal cumulative improvements and to think
in terms of big ideas: the inspiration, rather than the perspiration.

“Tiny tweaks” are great, but the idea that a significant fraction of tweaks come from people who “could never pass an exam or hold much of a conversation” is quite implausible.  Even most smart people are stuck in a rut.  And the less smart you are, the more likely your “tweak” is to actually make things worse.

As for small countries leading the world, I would argue it is almost the
rule. To Athens, Genoa, Holland, let me add New York (for much of
America was superfluous to America’s economic explosion in the
nineteenth century), San Jose, CA (for there was a remarkable
concentration of innovation there in recent decades) Singapore,
Hongkong, or going back in time Tyre, Sybaris, Pataliputra.

So the United States – the third most populous nation on earth – doesn’t count against your view because the actual innovations tend to be concentrated in particular cities like New York and San Jose?  I say you have to step back and realize that the probability that an innovative city arises in a country depends on the country’s (and, with immigration, the world’s) total population.

when it embarked on the industrial revolution had less than half the
population of France: this was clearly no disadvantage. I think it is
rather striking how infrequently big countries like Russia, France, the
Roman empire, the Ming Empire do lead the world economically.

Britain wasn’t the most populous county in the world at the time, but it was one of them.  According to Angus Maddison’s data, the only countries with higher population than Great Britain in 1850 were China, India, Russia, France, and Japan (and the countries that later became Germany, if you want to conceptually consolidate them).  I’m not saying that the population:innovation link is perfect, but it’s fairly positive – as you can see when you check the correlation between population and almost any observable measure of innovation.  Challenge for Ridley: I know you don’t like Nobel prize or patent numbers; is there any measure of innovation that you do accept that doesn’t positively correlate with population?

Your counter-examples are definitely thought-provoking, but I still say you’re missing the big historical picture: Most innovation has come from regions of the world with large, connected populations (especially Eurasia), and higher-population periods of human history are much more innovative than lower-population periods.

As for the notion that both population and connectedness matter, sure up
to a point. To have 10,000 people is a lot better than to have 5,000.
But to have a billion instead of 500m? I’m not convinced.

Suppose we randomly deleted half the countries on earth.  Would you seriously not expect innovation to sharply decline?