Another thing you notice when you see animals in the wild is that each species thrives in groups of a certain size. A herd of impalas might have 100 adults; baboons maybe 20; lions rarely 10. Humans also seem designed to work in groups, and what I've read about hunter-gatherers accords with research on organizations and my own experience to suggest roughly what the ideal size is: groups of 8 work well; by 20 they're getting hard to manage; and a group of 50 is really unwieldy.There you have it, from one of Paul Graham's essays: the maximum size of a company should be 20; if you could keep it down to 8 to 10, that would probably be ideal.
Whatever the upper limit is, we are clearly not meant to work in groups of several hundred. And yet—for reasons having more to do with technology than human nature—a great many people work for companies with hundreds or thousands of employees.
What does this mean for big monoliths like Microsoft? I'm not sure. I, for one, would rather just ignore such organizations, and hope that such organizations would ignore me--and to the extent that such a company could harrass me, it will be because they have a certain amount of the force of Law on their side, whether wielded legally or not. Surely, though, if Microsoft is too big, then so is the government that would try to break it up.
But even in a big monolith like Microsoft, at one point, it seems that individuals were given great power to get things done.
Undoubtedly, certain projects are so large that large organizations--or, at least, organizations of organizations--are useful, if not needed, to create them. Perhaps operating systems fall into this category. Even something as small as a car almost does. When you consider the regulations that have been placed on the car industry, a large organization artificially becomes necessary. And even if those organizations aren't strictly necessary, I would add that they are at least very useful for achieving certain goals.
Even so, I have often wondered: would it have been possible to create something as big as Microsoft Windows with Office, or something like Apple OS, with loosely-coupled groups, instead of one big company? The growth of Linux would seem to indicate that yes, this is possible--but is it possible to do so, and be profitable? This line of thinking definitely deserves exploration and experimentation.
Ultimately, though, I'm not afraid of large non-government organizations. Their powers are limited to buying and selling products and services, to making contracts, to hiring and firing people...oh, and to suing people. When they interfere with my life, it's usually made possible only because it's backed up by government force--which, in turn, is often made possible by copyright and patent law, mostly. (And, at this moment, I won't even touch on why these things are evil!)
Governments, however, even the smallest ones, have powers that the biggest Megacorporations can only dream of: to pass laws to make what I'm doing illegal; to drag me to court at gunpoint; to fine me, imprison me, even kill me; to tax what I'm doing--often in an effort to get me to stop doing it.
And this is why I am far more concerned about the size of Government, than I am the size of corporations. Ironically, many of those people who think large corporations are evil, are also those who champion Government as it grows exponentially.
Of course, large corporations are evil, if only for one thing: the leaders of these organizations are often in league with government, in attempts to secure their market-dominated positions and thwart competition--and even in atempts to "improve" society itself, by forcing other people to do what they think is best, independent of the mission of the corporation itself. (Tell me again, just why is Bill Gates in a position to lecture us about Estate--ah hem, Death--Taxes? Taxes he so diligently makes sure he won't have to pay?) But then, this is far more an issue about power, than it is about organization. That is, corporations become evil by blurring the line between corporation and government.