Source: Bet On It
by Bryan Caplan
“Qualitative history lends itself to confirmation bias. Even when you have some quantitative measures of economic policy, it’s easy to put extra weight on the measures that deliver the answer you’re looking for. I’m not immune to motivated reasoning. I loathe and despise both socialism and Nazism, so it’s pleasant for me to equate them. I freely admit it. What to do? I could spend a year reading more about this topic. I could even spend a few years getting to the research frontier so I could credibly publish on the question, ‘How socialist was Germany under National Socialism?’ But while this is a fascinating issue, the opportunity cost of seriously deepening my understanding is just too high. Or to be more precise, the opportunity cost was too high. AI has drastically slashed the costs of quantification — and simultaneously drastically increased the credibility of quick quantification.” (06/29/26)