As I understand it, the stated goal of the text was to build a concrete scenario despite this problem. Even though increased concreteness reduces the probability of a particular scenario from the cluster, it has other useful properties.
For example, it is probably easier and more interesting to analyze a concrete story. Building one and reconciling it with reality can be an exercise that improves one`s ability to make predictions.
and yet you do not identify any of these, supposed “other useful properties”.
How can you reconcile a prediction of algorithmic breakthroughs with reality? When would that reconciliation take place. Nobody is ever going to look back and say “I predicted algorithmic breakthroughs and there were none”. At best they’ll say that “the breakthroughs took longer than I expected but my predictions were good if you ignore that”.
and yet you do not identify any of these, supposed “other useful properties”.
I did it in the second paragraph.
How can you reconcile a prediction of algorithmic breakthroughs with reality? When would that reconciliation take place. Nobody is ever going to look back and say “I predicted algorithmic breakthroughs and there were none”
Maybe many people would do that, but I think at least some would be able to acnowledge the mistake and not rationalize away their prediction. To reconcile their prediction with reality, one, as an option, can make the prediction very concrete and narrow in the beginning. And that’s what people here generally try to do, as I see it.
Downvoted. See Burdensome Details. I particularly dislike predicting “Algorithmic Breakthroughs”
As I understand it, the stated goal of the text was to build a concrete scenario despite this problem. Even though increased concreteness reduces the probability of a particular scenario from the cluster, it has other useful properties.
For example, it is probably easier and more interesting to analyze a concrete story. Building one and reconciling it with reality can be an exercise that improves one`s ability to make predictions.
and yet you do not identify any of these, supposed “other useful properties”. How can you reconcile a prediction of algorithmic breakthroughs with reality? When would that reconciliation take place. Nobody is ever going to look back and say “I predicted algorithmic breakthroughs and there were none”. At best they’ll say that “the breakthroughs took longer than I expected but my predictions were good if you ignore that”.
I did it in the second paragraph.
Maybe many people would do that, but I think at least some would be able to acnowledge the mistake and not rationalize away their prediction. To reconcile their prediction with reality, one, as an option, can make the prediction very concrete and narrow in the beginning. And that’s what people here generally try to do, as I see it.