4 ms·

how would you judge the choices you have made? what would be a better choice, and how would you know it was better?

TameAntelope 6y ago
Poorly, because hindsight is hard, I just would skip the steps related to hiding my biases behind overly precise numbers.
- chordalkeyboard 6y ago
  "poorly" and "bad" related to what? what measure are we using to evaluate these outcomes? Whats the magical thing that we use to tell whether "rationality" is bad at achieving "it"?
  - TameAntelope 6y ago
    Poorly and bad refer to the inability to accurately assess outcomes objectively and accurately in a way that identifies predictors for future outcomes. In other words, the techniques praised in lesswrong writings dress up what is effectively bias shuffling, which results in the same problem of being unable to assess the predictors of outcomes, just now better hidden behind falsely confident figures instead of admission of uncertainty.
    - chordalkeyboard 6y ago
      > Poorly and bad refer to the inability to accurately assess outcomes objectively and accurately in a way that identifies predictors for future outcomes. So at some point you were able to assess outcomes objectively and update your perceptions? What would you call that process? > In other words, the techniques praised in lesswrong writings dress up what is effectively bias shuffling, which results in the same problem of being unable to assess the predictors of outcomes, just now better hidden behind falsely confident figures instead of admission of uncertainty. I’m asking how you can identify a truth-finding process that is not bias shuffling and what you would call it (since “rational” is now taboo).
      - TameAntelope 6y ago
        You're barking up the wrong tree. I offer no solutions, I only spot plenty of problems with what lesswrong tries to sell as one.
        chordalkeyboard 6y ago
        Are those really problems with their advocated solutions or are they the same kind of stupid human tricks you see humans do everywhere humans do things?
        TameAntelope 6y ago
        That's exactly my point; lesswrong offers more stupid human tricks, nothing more. So if I'm going to pull stupid human tricks to make myself feel better, let me do it in a way that keeps me aware of how stupid I am, rather than try to mask it under the guise of a superior method for eliminating the stupidity.
        chordalkeyboard 6y ago
        I'd agree there are plenty of stupid humans there (and elsewhere), but I'm not sure that a catalogue of heuristics with the apparent intent of lessening cognitive bias is a worthy target for your ire. Maybe it would make more sense if you identified some actual problems with their pastiche philosophy.
        TameAntelope 6y ago
        I specifically don't like how the use of approximate estimations of certainty are used through Bayesian probability theory. My chief concern is that through the application of figures to beliefs, you hide the inputs to those beliefs behind something that appears more objective than it actually is, leading to a meaningfully less thorough examination of why you arrived at that figure. A secondary concern I have is that it falsely proposes that emotion isn't part of (or at least not playing a significant role in) a decision process, and that the more we eliminate emotion, the more successful our outcomes will be. I don't understand that to be the case at all, rather it seems clear that emotion is how we know when to stop collecting data and which data to select to include as relevant in our decision making process, and that is how bias creeps in. You end up in a position where you feel you've applied some kind of process for arriving at a decision, and your feeling is not correct, but because it feels correct, you don't examine it further.

Poorly, because hindsight is hard, I just would skip the steps related to hiding my biases behind overly precise numbers.

chordalkeyboard 6y ago
"poorly" and "bad" related to what? what measure are we using to evaluate these outcomes? Whats the magical thing that we use to tell whether "rationality" is bad at achieving "it"?
- TameAntelope 6y ago
  Poorly and bad refer to the inability to accurately assess outcomes objectively and accurately in a way that identifies predictors for future outcomes. In other words, the techniques praised in lesswrong writings dress up what is effectively bias shuffling, which results in the same problem of being unable to assess the predictors of outcomes, just now better hidden behind falsely confident figures instead of admission of uncertainty.
  - chordalkeyboard 6y ago
    > Poorly and bad refer to the inability to accurately assess outcomes objectively and accurately in a way that identifies predictors for future outcomes. So at some point you were able to assess outcomes objectively and update your perceptions? What would you call that process? > In other words, the techniques praised in lesswrong writings dress up what is effectively bias shuffling, which results in the same problem of being unable to assess the predictors of outcomes, just now better hidden behind falsely confident figures instead of admission of uncertainty. I’m asking how you can identify a truth-finding process that is not bias shuffling and what you would call it (since “rational” is now taboo).
    - TameAntelope 6y ago
      You're barking up the wrong tree. I offer no solutions, I only spot plenty of problems with what lesswrong tries to sell as one.
      - chordalkeyboard 6y ago
        Are those really problems with their advocated solutions or are they the same kind of stupid human tricks you see humans do everywhere humans do things?
        TameAntelope 6y ago
        That's exactly my point; lesswrong offers more stupid human tricks, nothing more. So if I'm going to pull stupid human tricks to make myself feel better, let me do it in a way that keeps me aware of how stupid I am, rather than try to mask it under the guise of a superior method for eliminating the stupidity.
        chordalkeyboard 6y ago
        I'd agree there are plenty of stupid humans there (and elsewhere), but I'm not sure that a catalogue of heuristics with the apparent intent of lessening cognitive bias is a worthy target for your ire. Maybe it would make more sense if you identified some actual problems with their pastiche philosophy.
        TameAntelope 6y ago
        I specifically don't like how the use of approximate estimations of certainty are used through Bayesian probability theory. My chief concern is that through the application of figures to beliefs, you hide the inputs to those beliefs behind something that appears more objective than it actually is, leading to a meaningfully less thorough examination of why you arrived at that figure. A secondary concern I have is that it falsely proposes that emotion isn't part of (or at least not playing a significant role in) a decision process, and that the more we eliminate emotion, the more successful our outcomes will be. I don't understand that to be the case at all, rather it seems clear that emotion is how we know when to stop collecting data and which data to select to include as relevant in our decision making process, and that is how bias creeps in. You end up in a position where you feel you've applied some kind of process for arriving at a decision, and your feeling is not correct, but because it feels correct, you don't examine it further.