Post

Green missed the mark with regard to Brandolini's law and AI. First, AI makes it easier to generate a higher volume of credible sounding pseudo science faster. And repetition is key to people believing the b.s. So AI does make the problem worse. (1/x)

(2/x) you can't count on AI to summarize the paper for you to do the debunking. You actually have to read it and understand it yourself. In addition, if new research is required to settle some aggressive claim that is in a scientific grey area, that research takes time.

(3/x) AI tends towards lowest common denominator analysis. It doesn't understand what constitutes quality research and mistakes volume for accuracy. It is pretty stupid in this regard and there is no reason to think it will get better because no one seems to think it needs to improve... so it won't