Do diseconomies of scale actually exist in practice?

I remember hearing Steve Keen say economies of scale (at least can) continue forever. He cites companies like Amazon which, instead of flattening out and seeing a diseconomy of scale as predicted, they just keep scaling and getting more efficient as they expand to become enormously big. Keen says diseconomies of scale do not hold up to empirical scrutiny. Is this true? What are some example of diseconomies of scale in practice?