@davidboxenhorn@normonics@DavidDeutschOxf When we say "that doesn't sound right" we are really doing a Bayesian probability calculation of the right phrasing to use based on what we've heard before. That's what GPT does.
@pmontu33 Deep. In a sense, wisdom probably depends more on the interpretation engine than the input data. A wise engine can be fed any random thing - say, gardening tips -- and draw some deep insights from it. Even random noise fed into the engine would output deep wisdom.
@nntaleb Injecting more random noise reduces random noise, allowing the signal to emerge. Lindy emerges from b** in a similar fashion. (I think I learned this in FBR over a decade ago -- speaking of Lindy)
@nntaleb the BS debunking is just as helpful as the concepts you teach. Via negativa, rather have no map than the wrong map. Your best work is on the fat tails -- affirming on the right and BS debunking on the left.
@nntaleb A symmetry of this principle might be in how you can't judge a contemporary person as virtuous for adhering to the un-evolved standards of virtue from an older era.