I think you missed what I see as the main point in “What they might have considered, however, was that there was no valve in their pipe. Aiming to purify the American state, they succeeded only in corrupting the American mind.” Not surprising, because Moldbug (the guy quoted about the Mugwumps) is terribly long-winded and given to rhetorical flourishes. So let me try to rephrase what I see as the central objection in a format more amenable to LW:
The scientific community is not a massive repository of power, nor is it packed to the gills with masters of rhetoric. The political community consists of nothing but. If you try to run your new party by listening to the scientific community without first making the scientific community far more powerful and independent, what’s likely to happen is that the political community makes a puppet of the scientific community, and then you wind up running your politics by listening to a puppet of the political community.
The more you promote “Do what the NSF says”, the more Congress is going to be interested in using some of those billions of dollars to lean on the NSF and other similar organizations so that you will be promoting “Do what Congress says” at arm’s remove. No overt dishonesty needs be involved. Just little things like hiring sympathetic scientists, discouraging controversial research, asking for a survey of a specific metric, etc.
Suppose you make a prediction that a law will decrease the crime rate. You pass the law. You wait a while and see. Did the crime rate go down? Well, how are you measuring crime rate? Which crimes are you counting? To take an example discussed on Less Wrong a while ago, if you use the murder rate as proxy for crime rate over the past few decades, you are going to severely undercount crime because of improvements in medical technology that make worse wounds more survivable.
Obviously you can fix this particular metric now that I’ve pointed it out. But can you spot and fix such issues in advance faster and better than people throwing around 30 billion dollars and with a massive vested interest in retaining policy control?
When trying to solve something like whether P=NP, you can throw more and brighter scientists at the problem and trust that the problem will remain the same. But the problem of trying to establish science-based policy, particularly when “advocating loads of funding for science”, gets harder as it gets more important and you throw more people at it. This is a Red Queen’s Race where you have to keep running just to stay in place, because you’re not dealing with a mindless question that has an objective answer floating out there, you’re dealing with an opposed social force with lots of minds and money that learns from its own mistakes and figures out how to corrupt better, and with more plausible deniability.
Thankyou—this statement of the idea was much, much clearer to me. :)
It seems like the solution—well, a possible part of one possible solution—is to make the social science research institute that everyone listens to have some funding source which is completely independent from the political party in power. That would hopefully make the scientific community more independent. We now need to make it more powerful, which is… more difficult. I think a good starting point would be to try and raise the prestige associated with a social science career (and thus the prestige given to individual social scientists and the amount of social capital they feel they have to spend on being controversial) and possibly give some rhetoric classes to the social science research institute’s spokesperson. Assuming the scientists are rational scientists, this gives them politician-power with which to persuade people of their correct conclusions. (Of course, if they have incorrect conclusions influenced by their ideologies, this is… problematic. How do we fix this? I dunno yet. But this is the very beginning of a solution, but I’ve not been thinking about the problem very long and I am just one kid with a relatively high IQ. If multiple people work together on a solution, I’m sure much more and much better stuff will be come up with.)
--
I think you missed what I see as the main point in “What they might have considered, however, was that there was no valve in their pipe. Aiming to purify the American state, they succeeded only in corrupting the American mind.” Not surprising, because Moldbug (the guy quoted about the Mugwumps) is terribly long-winded and given to rhetorical flourishes. So let me try to rephrase what I see as the central objection in a format more amenable to LW:
The scientific community is not a massive repository of power, nor is it packed to the gills with masters of rhetoric. The political community consists of nothing but. If you try to run your new party by listening to the scientific community without first making the scientific community far more powerful and independent, what’s likely to happen is that the political community makes a puppet of the scientific community, and then you wind up running your politics by listening to a puppet of the political community.
To give a concrete relatable figure: The US National Science Foundation receives about 7.5 billion dollars a year from the US Congress. (According to the NSF, they are the funding source for approximately 24 percent of all federally supported basic research conducted by America’s colleges and universities, which suggests 30 billion federal dollars are out there just for basic research)
The more you promote “Do what the NSF says”, the more Congress is going to be interested in using some of those billions of dollars to lean on the NSF and other similar organizations so that you will be promoting “Do what Congress says” at arm’s remove. No overt dishonesty needs be involved. Just little things like hiring sympathetic scientists, discouraging controversial research, asking for a survey of a specific metric, etc.
Suppose you make a prediction that a law will decrease the crime rate. You pass the law. You wait a while and see. Did the crime rate go down? Well, how are you measuring crime rate? Which crimes are you counting? To take an example discussed on Less Wrong a while ago, if you use the murder rate as proxy for crime rate over the past few decades, you are going to severely undercount crime because of improvements in medical technology that make worse wounds more survivable.
Obviously you can fix this particular metric now that I’ve pointed it out. But can you spot and fix such issues in advance faster and better than people throwing around 30 billion dollars and with a massive vested interest in retaining policy control?
When trying to solve something like whether P=NP, you can throw more and brighter scientists at the problem and trust that the problem will remain the same. But the problem of trying to establish science-based policy, particularly when “advocating loads of funding for science”, gets harder as it gets more important and you throw more people at it. This is a Red Queen’s Race where you have to keep running just to stay in place, because you’re not dealing with a mindless question that has an objective answer floating out there, you’re dealing with an opposed social force with lots of minds and money that learns from its own mistakes and figures out how to corrupt better, and with more plausible deniability.
Thankyou—this statement of the idea was much, much clearer to me. :)
It seems like the solution—well, a possible part of one possible solution—is to make the social science research institute that everyone listens to have some funding source which is completely independent from the political party in power. That would hopefully make the scientific community more independent. We now need to make it more powerful, which is… more difficult. I think a good starting point would be to try and raise the prestige associated with a social science career (and thus the prestige given to individual social scientists and the amount of social capital they feel they have to spend on being controversial) and possibly give some rhetoric classes to the social science research institute’s spokesperson. Assuming the scientists are rational scientists, this gives them politician-power with which to persuade people of their correct conclusions. (Of course, if they have incorrect conclusions influenced by their ideologies, this is… problematic. How do we fix this? I dunno yet. But this is the very beginning of a solution, but I’ve not been thinking about the problem very long and I am just one kid with a relatively high IQ. If multiple people work together on a solution, I’m sure much more and much better stuff will be come up with.)