A Rule of Thumb for (Engineering) Decisions
In engineering, there is this rule of thumb about “one-way” versus “two-way” decisions. In a sentence: there are situations in which we can afford to experiment, fail, and roll back without much consequence beyond the time invested itself. These are two-way decisions. Other situations require much more scrutiny, because going back is not really feasible in context. These are one-way decisions. This paradigm for decision-making was famously championed by Jeff Bezos in Amazon’s 2016 Letter to Shareholders. Many decisions are, in fact, of the two-way kind, he argues, and their (ir)reversible character affects who gets to decide, how fast an organization moves, and therefore how quickly it can innovate and grow. I recently saw it very well explained in this interview with Matteo Ferrando, a fal.ai engineer. The general wisdom says that we should try to identify two-way doors as much as we can. ...