For a long time, a lot of the mechanisms of this were opaque to me -- how does a corporation form this collective thinking organism? Doing that seems quite difficult.
From my current internal view of a small company (between 200 and 300 people) and my previous extensive experience doing the same thing in a number of similar-sized companies, I now get it. I have been one of the corporate mini-minds, of several different types, in several corporations, which gives me a better idea of what they do.
A corporation normally takes its "blueprint" from one or more parent companies, usually those the founders have previously worked for. That blueprint is specifically a set of roles for people, and what kind of person such a role should be filled with.
This blueprint may change over time, and normally does, moving in the direction of one previous company or another of various high-ranking people in the company. In a tech company, a prominent Amazon executive hire will tend to move a number of things in the direction of how Amazon does it. Over time, if it was a good hire, it shifts the blueprint.
"If it was a good hire?" Yes. An executive hire is, primarily, a shift in how the corporation sees itself, how it looks for prominent people, and what kinds of roles it thinks prominent people should occupy. These changes, in hiring and structure, are some of the highest-leverage changes a corporation ever makes. They cause a "pull" toward one future or another, which is much more effective in creating long-term change than any "big bang" sudden change could be.
The blueprint is made of mini-minds -- not a full corporate overmind. If there's one of those, the CEO does most of the dictating of it. But under that overmind (or similar high-level controlling structure) are a number of much more special-case minds, which constantly answer questions like "how should we achieve this specific engineering goal?" and "how should we allocate the high-level advertising budget?"
A corporation is alien to the human idea of mind in several ways.
First: the mini-minds are chosen consciously. You don't have to ape an existing structure -- though you usually do, because an untried arrangement of mini-minds usually fails, killing your company. You normally pick something that looks a *lot* like a setup you've seen succeed. But in theory, any arrangement of mini-minds in any arrangement of purposes can be chosen and paid for. The limit occurs in time, when the corporation either makes enough money to sustain itself or does not.
Second: the mini-minds can be re-chosen at any time. A corporation chooses the kinds of questions and actions a mini-mind should incorporate. A given person serving in that role may choose to do something (slightly) different, but their compensation and authority in the company will be based on the chosen actions and questions. The corporation will reward them for doing what it says, not what it needs. In the extreme case, a corporation can veto a given mini-mind entirely, usually by firing the person performing that function (or changing what they do.) A given set of actions and questions can be completely removed from the corporate capabilities or consciousness, just by having nobody doing them.
Third: the mini-minds are chosen by the voluntary, inspected action of a human being, in a way that is frequently public within the company. The CEO or one of his chosen managers creates every job role, and thus every function of mind within a corporation.
How is that alien? How is that effectively unlike a human, whose internal structure is mostly unknown, and who may also be composed of mini-minds for all we know?
Difference the first: a human has a powerful pre-built structure as a result of heredity and training. While a corporation usually mostly draws its structure from an existing successful corporation, a human is *always* trained and draws most of its assumptions from a group of existing successful humans.
Difference the second: a human has powerful inertia built in. While a corporation can fire a "conscience" or an "awareness of certain consequences" at any time, a human is stuck trying to use denial -- which is such a bad idea that we have constant consequences of guilt and cognitive dissonance that attempt to prevent us. Trying to remove a part of a human mind entirely is remarkably difficult. The equivalent for a corporation is easy, even when it's a terrible idea.
Difference the third: since the creation of each mini-mind, and each change to the blueprint, is known, there is a sense of collective shame in certain actions. A given thing being unpopular (or otherwise subject to veto) can be enough to prevent it, even if it essentially required.