It is an indicator that you would hope would show correlation.
Unfortunately, there are people who will just patent every single idea they have regardless of merit. Still, since some level of uniqueness and practicality is required for a patent, the people who can make the most patentable ideas can probably also generate useful additions to the model of rationality the fastest.
On the other hand, since Eliezer is trying to spread rationality, it makes sense to target the largest population that can benefit, using Japanese would drastically reduce the number who can read his works. Even if you could show that the Japanese were on average slightly more rational to begin with, you would end up with less people who learned all he could teach them.
Yes, agreed on both counts. Though, I think there are also other big issues with using patents as an indicator of innovation:
Whether or not you should get a patent for a given innovation is a cultural and situational issue. For example, an Open Source software developer is unlikely to seek a patent for their work, as is a university researcher who is hoping to publish their results. Using patents as an indicator might be messed up by the two cultures having different emphases on styles and methods of innovation.
A patent contains a variable amount of innovation: depending on how you want to package it, a given idea might be encapsulated in one big patent or in ten small ones. So, using patents as an indicator might be messed up by the two cultures tending to, for whatever practical or sociological or economic reason, group up or split up their patent applications.
Finally, this would be the easiest thing to research, but I’m not sure if the US patent office and the Japanese patent office use the same thresholds for minimum innovation, de facto.
Yeah, all three of those points make using patents to judge Innovation almost useless. Until you compare the cultural differences and patent requirements more, it still counts as weak evidence though.
It is an indicator that you would hope would show correlation. Unfortunately, there are people who will just patent every single idea they have regardless of merit. Still, since some level of uniqueness and practicality is required for a patent, the people who can make the most patentable ideas can probably also generate useful additions to the model of rationality the fastest.
On the other hand, since Eliezer is trying to spread rationality, it makes sense to target the largest population that can benefit, using Japanese would drastically reduce the number who can read his works. Even if you could show that the Japanese were on average slightly more rational to begin with, you would end up with less people who learned all he could teach them.
Yes, agreed on both counts. Though, I think there are also other big issues with using patents as an indicator of innovation:
Whether or not you should get a patent for a given innovation is a cultural and situational issue. For example, an Open Source software developer is unlikely to seek a patent for their work, as is a university researcher who is hoping to publish their results. Using patents as an indicator might be messed up by the two cultures having different emphases on styles and methods of innovation.
A patent contains a variable amount of innovation: depending on how you want to package it, a given idea might be encapsulated in one big patent or in ten small ones. So, using patents as an indicator might be messed up by the two cultures tending to, for whatever practical or sociological or economic reason, group up or split up their patent applications.
Finally, this would be the easiest thing to research, but I’m not sure if the US patent office and the Japanese patent office use the same thresholds for minimum innovation, de facto.
Yeah, all three of those points make using patents to judge Innovation almost useless. Until you compare the cultural differences and patent requirements more, it still counts as weak evidence though.