By contrast, the transaction cost economics models make-or-buy decisions as rational choice between real, available alternatives, imposing the reality principle

By contrast, the transaction cost economics models make-or-buy decisions as rational choice between real, available alternatives, imposing the reality principle at choice time. In my experience of collective decision-making, it's often the case that more aggressive, less-proven technologies are rejected as unproven or unrealistic, largely because no one wants the reputation in the group of having championed a mistake. By contrast, people deciding alone often will take the more optimistic choice. In technology, that can mean that person/engineer who's now on the hook finds ways to make the new technology work (and avoid its flaws). That translates to high-achieving organizations giving individuals the power to decide, but also holding them responsible for the consequences. Whether the "move fast, break things" permission to fail in service of learning new technologies and the problem domain actually works depends on some real capture of knowledge. Probably the job cuts in tech now (particularly at Twitter) are driven by realizing this "real capture" ain't happening. So it's not enough to avoid the Nirvana fallacy. You also have to get past decision paralysis to learn, but show the lessons you learned are worth something to the company.