I like this idea, but measuring the magnitude of negative effects is hard. You can’t trust companies to report how many hours of sweatshop labor they use per unit, and in many cases that information is scattered between the manufacturers of many different parts and no one really knows. Similarly for energy used, types and amounts of toxins released, politicians bribed, small guys sued, and so on. It might also be a lawsuit magnet.
Yeah, cases where there’s strongly motivated deception are a real test case for techniques for reasoning under noisy data.
I’d sort of like to see an effort like this that embraced both viewpoint pluralism and nymity/reputation.
That is, let anyone who wants to write an article making out a case for the net impact of a given practice, but have a social norm requiring that they actually register under their real names, and provide a mechanism whereby users of the system can easily get a summary of what sorts of claims a given user makes.
Maybe even introduce a layer of site-endorsed evaluations on grounds like does this case depend on unsubstantiated empirical claims, does it take into account what is known/generally believed about the topic, etc., in cases where that kind of objective evaluation is possible. Not so much “agree/disagree” but “observed facts are consistent with/inconsistent with/irrelevant to”.
That could obviate the lawsuit risk, perhaps? I mean, if the site doesn’t endorse any particular position, but just functions as a clearinghouse for individual cases.
I like this idea, but measuring the magnitude of negative effects is hard. You can’t trust companies to report how many hours of sweatshop labor they use per unit, and in many cases that information is scattered between the manufacturers of many different parts and no one really knows. Similarly for energy used, types and amounts of toxins released, politicians bribed, small guys sued, and so on. It might also be a lawsuit magnet.
Yeah, cases where there’s strongly motivated deception are a real test case for techniques for reasoning under noisy data.
I’d sort of like to see an effort like this that embraced both viewpoint pluralism and nymity/reputation.
That is, let anyone who wants to write an article making out a case for the net impact of a given practice, but have a social norm requiring that they actually register under their real names, and provide a mechanism whereby users of the system can easily get a summary of what sorts of claims a given user makes.
Maybe even introduce a layer of site-endorsed evaluations on grounds like does this case depend on unsubstantiated empirical claims, does it take into account what is known/generally believed about the topic, etc., in cases where that kind of objective evaluation is possible. Not so much “agree/disagree” but “observed facts are consistent with/inconsistent with/irrelevant to”.
That could obviate the lawsuit risk, perhaps? I mean, if the site doesn’t endorse any particular position, but just functions as a clearinghouse for individual cases.