5 votes

The paradox of control

5 comments

  1. [5]
    skybrian
    Link
    I suppose it’s true that when you fully understand a subject it becomes less interesting. But it seems like this argument assumes some sort of end of history, that there is little remaining to...

    I suppose it’s true that when you fully understand a subject it becomes less interesting. But it seems like this argument assumes some sort of end of history, that there is little remaining to explore.

    But that seems like a subjective impression that comes from not looking. There are plenty of scientific questions on which little progress has been made and little danger that scientists will be out of work.

    1. [4]
      entangledamplitude
      Link Parent
      Quite the contrary, the way I understand the thesis is that the more we try to control a phenomenon, the harder it becomes to actually get (more) useful information/understanding from it. In the...

      I suppose it’s true that when you fully understand a subject it becomes less interesting. But it seems like this argument assumes some sort of end of history, that there is little remaining to explore.

      Quite the contrary, the way I understand the thesis is that the more we try to control a phenomenon, the harder it becomes to actually get (more) useful information/understanding from it. In the same spirit as: any metric that is managed ceases to be a good metric.

      1. [3]
        skybrian
        Link Parent
        Rereading the article, it doesn’t seem very clear what the author means by “control.” Sometimes they are talking about measuring things, which is closer to observation, although it’s one part of a...

        Rereading the article, it doesn’t seem very clear what the author means by “control.” Sometimes they are talking about measuring things, which is closer to observation, although it’s one part of a control system.

        1. [2]
          entangledamplitude
          Link Parent
          Part of the essential idea is that we don’t have the sagacity to just measure but not optimize (Goodhart’s law). All of those attempts at control are only weakly correlated with well-being. The...

          Part of the essential idea is that we don’t have the sagacity to just measure but not optimize (Goodhart’s law).

          “We climb onto the scale: we should lose weight. We look into the mirror: we have to get rid of that pimple, those wrinkles. We take our blood pressure: it should be lower. We track our steps: we should walk more.” “We invariably encounter such things,” Rosa notes, “as a challenge to do better.”

          All of those attempts at control are only weakly correlated with well-being. The only thing they have going for them is that they provide concrete directives for action, which are seductive when one operates with a “bias for action”.

          1. skybrian
            Link Parent
            A scale is for measuring your weight, not your well-being. This doesn't seem like a paradox? Also, Goodhart's law is more about fooling other people, not yourself.

            A scale is for measuring your weight, not your well-being. This doesn't seem like a paradox? Also, Goodhart's law is more about fooling other people, not yourself.