I do not believe that we use the same definition of steelmanning.
Steelmanning is making the best case possible for the idea in question.
Bad steelmanning consists of bottom-lining the idea in question and amassing as many soldiers for it as possible. One nails idea X in place and asks one’s brain, what does that world look like? What might be true to imply that X is true? The virtual outcome pump in one’s head obligingly imagines something, which you write above X as an argument to prove X. Repeat the process on those premises until you have something that looks like a coherent argument for X, but resembles one only in the way that a painting of a bridge looks like a bridge. It does not stay up because of its sound construction as a bridge, but only because of its sound construction as a painting: the paint is stuck to the canvas.
Even good steelmanning has a hint of the bottom line about it, but that is because it is a technique of anti-irrationality, not directly of rationality. To the perfect reasoner, there is no such activity as making the best case possible for an idea, only the best argument possible relating to the idea, whichever way it turns out. The imperfect reasoner’s task is to force themselves to find actual good reasons for X even while flinching away from the task. It is futile to build a straw man and give it a coat of engine paint.
As for the original scenario, the everyday world provides far better examples where, by virtue of diminished responsibility, protection of some greater good, or various other reasons, one may be justified in forcibly or covertly thwarting someone else’s wishes. The scenario of driving under the influence of alcohol with a passenger who would refuse is a really bad one, and there is no point in putting a finger on the scales to make the decision come out in favour of driving.
Steelmanning is making the best case possible for the idea in question.
Bad steelmanning consists of bottom-lining the idea in question and amassing as many soldiers for it as possible. One nails idea X in place and asks one’s brain, what does that world look like? What might be true to imply that X is true? The virtual outcome pump in one’s head obligingly imagines something, which you write above X as an argument to prove X. Repeat the process on those premises until you have something that looks like a coherent argument for X, but resembles one only in the way that a painting of a bridge looks like a bridge. It does not stay up because of its sound construction as a bridge, but only because of its sound construction as a painting: the paint is stuck to the canvas.
Even good steelmanning has a hint of the bottom line about it, but that is because it is a technique of anti-irrationality, not directly of rationality. To the perfect reasoner, there is no such activity as making the best case possible for an idea, only the best argument possible relating to the idea, whichever way it turns out. The imperfect reasoner’s task is to force themselves to find actual good reasons for X even while flinching away from the task. It is futile to build a straw man and give it a coat of engine paint.
As for the original scenario, the everyday world provides far better examples where, by virtue of diminished responsibility, protection of some greater good, or various other reasons, one may be justified in forcibly or covertly thwarting someone else’s wishes. The scenario of driving under the influence of alcohol with a passenger who would refuse is a really bad one, and there is no point in putting a finger on the scales to make the decision come out in favour of driving.