IBM, Microsoft, Google--they have one or a few core selling products, the rest of their activities could be explained as self-promotion to those that would otherwise invest their time and skills elsewhere. In short, survival.

Stretching it, such large companies may become countries in the sense that they cover the needs of their citizens. Like a VM on an OS. The lucky ones will have wise benevolent dictators (the best political option in Plato's government evolution stages--not that I am endorsing it.)

* A theme running through many of pg's essay is that startups are a far more productive use of resources. Pg sometimes uses acquisitions as an example taking the same resources & applying them to a large organisation. The result is reduced productivity. Doesn't that imply that acquisitions shouldn't take place in the first place? A transaction is supposed to take place when it produces some sort of a net surplus.

* Can a startup complex survive without acquisitions?

* If size is a disadvantage, isn't this a disincentive to grow? Wouldn't that put a company in danger of being out competed by smaller players?

* If growth is capped by anti - economy of scale effects, is there still enough incentive for startups?

* Today's startups often float around winner take most areas. Isn't this at odds with small organisations being dominant? The solution to this one seems alarming: Small organisations with huge revenues. If this scales we get a large number of Googles being run entirely by a few hundred employees & responsible for a level of wealth creation grossly disproportionate to their number of employees.

My ideas are pretty uncooked. But what I'm saying is that this prediction might be more robust if it explored a separation startup & big business. IE what does a startup look like in an economy dominated by startups. Foxes can never outnumber rabbits.