[bayesianhorse]

As far as I understood it, Bill Gates was arguing that the worst of poverty is vanishing surprisingly well.

Kay says this is wrong, because the countries with intermediate economies aren't on the top yet.

I think the basic misunderstanding is that economic development - if it works - is a stochastic exponential process. Humans tend to underestimate this function badly. On the one hand it is very fast - after the US went from where Turkey is today to its current economy in "only" 50 years, but in every particular year, looking into the future, progress looks like a painstakingly slow crawl.