About induction, deduction, and transduction

I have noticed a new posting by Ya’acov Ritov on arXiv that discusses what the limits of the scope of Statistics should be:

“The paper argues that a part of the current statistical discussion is not based on the standard firm foundations of the field. Among the examples we consider are prediction into the future, semi-supervised classification, and causality inference based on observational data.”

I do not have currently enough free time to read it at a detailed enough level to make a sensible comment, but this sounds like an interesting discussion! At this stage, I cannot decide whether this is yet again a point about model shifts or if there is a more fundamental issue at stake. (Thankfully, Popper is not mentioned! But Taleb is…) It seems however that the paper claims that prediction about a single object is not statistically valid:

“We believe that predicting the future, that is, predicting one most important future event, is not a statistical task.“

and thus that statistics requires a long sequence of experiments to achieve validation, hence falling upon a frequentist justification…