Interestingly, this seems like an argument for the inadequecy of academia. They are clearly inadequate at incorporating the ideas of people who don’t sufficiently respect academia!
If my decision on whether to update on new information depends heavily on how polite the person making the suggestions is being, that seems like a serious mistake if I’m optimizing for an accurate model of the world or funding the best research, and not a mistake at all if I’m optimizing for institutional and/or personal prestige and success.
Certainly you could reply that such rudeness constitutes some Bayesian evidence, but by your own observation they’re updating far too much from that standpoint.
At a minimum, it suggests a heuristic that you should update the scientific consensus in favor of anyone who have new and/or useful ideas and/or information, and is being rude to academia.
Interestingly, this seems like an argument for the inadequecy of academia. They are clearly inadequate at incorporating the ideas of people who don’t sufficiently respect academia!
They are probably right much more often than they are wrong. Before you go calling them inadequate, you need to show they could do better with their resources. How much physics would get done if every physics crank got a careful assessment.
(Scott Aaronson writes:
“Yesterday several people asked my opinion of a preprint claiming to solve the Graph Isomorphism problem in deterministic polynomial time. I responded:
If I read all such papers, then I wouldn’t have time for anything else.” )
If my decision on whether to update on new information depends heavily on how polite the person making the suggestions is being,
This is about things much more specific than oridnary rudeness. Consider the rationality of giving high quality feedback to someone who has already signalled that they dont respect you and don’t listen to feedback.
At a minimum, it suggests a heuristic that you should update the scientific consensus in favor of anyone who have new and/or useful ideas and/or information, and is being rude to academia.
Not in any way. You can’t update the consensus using Bayes anyway, because y ou can’t just average out fundamentally different ideas.
I notice I am confused. Why would whether or not something is adequate to a task depend on whether they could have done better with the resources they had available? Academia is inadequate to teaching my three year old to play the violin. That’s not because they’re bad at it, or doing anything wrong, or that someone should be adequate to that task. It’s just not something they can do.
You’re arguing, it seems to me, that academia is acting correctly when it refuses to be adequate at examining such claims. That it would not be a good or rational use of its resources to be adequate. I’m not even going to disagree with that. If I decide to devote more resources to an area, presumably I should be able to do better than those who think the problem isn’t worth their time. Even if it isn’t worth their time!
That last sentence I don’t understand at all. The consensus is a probability distribution, or else it’s deeply terrible. If I get new information I should update on that information using Bayes. If the consensus says it’s (probably) not raining and I look outside and see rain, I update or I’m being insane. People can and do incorporate fundamentally different ideas into their world models every day. To suggest it can’t be done confuses me.
There is no consensus on what the consensus is, or on how to do epistemology generally. There is no separate standpoint on the sum total of topics that academics discuss from which it can be evaluated. This meta-convoluted recursion is why you can’t apply one-size-fits-all mechnaical solutions like Bayes. See Chapman on Bayes.
If I get new information I should update on that information using Bayes. If the consensus says it’s (probably) not raining and I look outside and see rain, I update or I’m being insane
You can’t infer from simple cases to complex ones.
People can and do incorporate fundamentally different ideas into their world models every day.
That’s not evidence that it’s being done with Bayes.
Why would whether or not something is adequate to a task depend on whether they could have done better with the resources they had availab
Inadequacy of a kind worth complaining about is inadequacy is inadeqaucy that can be reasonaby be fixed. That leads to the possibility of a kind of fallacy where someone is condemned for inadeqacy that isnt actually culpable. “The government are wicked becasue they haven’t given eveyone gold palaces to live in”.
Inadequate equilibria is about the other kind. The point about an inadeauate equilbrium is that there is some possibility of getting into a better one.
This started with a complaint that academia wasn’t taking enough notice of Bostrom’s work. Well, if they are failing to notice it becasue of some fixable problem like bad incentives, then that is one kind of inadeuacy. If they are just too damn busy, like Aaronson, that is the other kind.
Interestingly, this seems like an argument for the inadequecy of academia. They are clearly inadequate at incorporating the ideas of people who don’t sufficiently respect academia!
If my decision on whether to update on new information depends heavily on how polite the person making the suggestions is being, that seems like a serious mistake if I’m optimizing for an accurate model of the world or funding the best research, and not a mistake at all if I’m optimizing for institutional and/or personal prestige and success.
Certainly you could reply that such rudeness constitutes some Bayesian evidence, but by your own observation they’re updating far too much from that standpoint.
At a minimum, it suggests a heuristic that you should update the scientific consensus in favor of anyone who have new and/or useful ideas and/or information, and is being rude to academia.
They are probably right much more often than they are wrong. Before you go calling them inadequate, you need to show they could do better with their resources. How much physics would get done if every physics crank got a careful assessment.
(Scott Aaronson writes:
“Yesterday several people asked my opinion of a preprint claiming to solve the Graph Isomorphism problem in deterministic polynomial time. I responded:
This is about things much more specific than oridnary rudeness. Consider the rationality of giving high quality feedback to someone who has already signalled that they dont respect you and don’t listen to feedback.
Not in any way. You can’t update the consensus using Bayes anyway, because y ou can’t just average out fundamentally different ideas.
I notice I am confused. Why would whether or not something is adequate to a task depend on whether they could have done better with the resources they had available? Academia is inadequate to teaching my three year old to play the violin. That’s not because they’re bad at it, or doing anything wrong, or that someone should be adequate to that task. It’s just not something they can do.
You’re arguing, it seems to me, that academia is acting correctly when it refuses to be adequate at examining such claims. That it would not be a good or rational use of its resources to be adequate. I’m not even going to disagree with that. If I decide to devote more resources to an area, presumably I should be able to do better than those who think the problem isn’t worth their time. Even if it isn’t worth their time!
That last sentence I don’t understand at all. The consensus is a probability distribution, or else it’s deeply terrible. If I get new information I should update on that information using Bayes. If the consensus says it’s (probably) not raining and I look outside and see rain, I update or I’m being insane. People can and do incorporate fundamentally different ideas into their world models every day. To suggest it can’t be done confuses me.
There is no consensus on what the consensus is, or on how to do epistemology generally. There is no separate standpoint on the sum total of topics that academics discuss from which it can be evaluated. This meta-convoluted recursion is why you can’t apply one-size-fits-all mechnaical solutions like Bayes. See Chapman on Bayes.
You can’t infer from simple cases to complex ones.
That’s not evidence that it’s being done with Bayes.
Inadequacy of a kind worth complaining about is inadequacy is inadeqaucy that can be reasonaby be fixed. That leads to the possibility of a kind of fallacy where someone is condemned for inadeqacy that isnt actually culpable. “The government are wicked becasue they haven’t given eveyone gold palaces to live in”.
Inadequate equilibria is about the other kind. The point about an inadeauate equilbrium is that there is some possibility of getting into a better one.
This started with a complaint that academia wasn’t taking enough notice of Bostrom’s work. Well, if they are failing to notice it becasue of some fixable problem like bad incentives, then that is one kind of inadeuacy. If they are just too damn busy, like Aaronson, that is the other kind.