Ask HN: Statistical learning and non-Statistical learning for humans
I have been pondering a lot about this question, at least since the rise of LLM models that may genuinely get pass Turing's test. I'd like to put down conclusion here: I believe humans learn both from a Statistical method and non-Statistical method, and we are not very different from LLM in the first method.
Statistical method: I'm an introvert. I do not like interacting socially f2f (online is a completely different matter as I can bubble for days non-stop). I also happened to migrate to a different country with a different culture after I matured. I believe I learned how to social -- or, to be more precisely, how to keep others think that I'm too much of a weirdo to be fit in a team -- with a Statistical model. Here is what I meant: social interaction between humans face to face starts with greeting, and then small talks to warm up, and then maybe more formal discussions, and then small talks to chill down, and finally say good-bye. I learned to do all these things by observing my friends and colleagues and just mimic them. I mimic them by taking note "which small talks are statistically more probable to occur, and which responses are considered to be normal, or even funny, cunning", "statistically how do people transition between the different stages of social interaction", etc., and then apply them, adjust my method according in practice. I do feel that I'm acting like an LLM, and so far has been very successful, considering I never failed a team interview, and I never got fired by being awkward.
Non-Statistical method: On the other hand, I feel that learning Math/Programming/Physics is a COMPLETELY different thing. I can read a ton of books, but I can't "Statistically" create a proof for a new problem that I never encountered in those textbooks. For sure the problem is similar to some of the examples or past exercises, but there is no Statistics playing here. I feel that I "understand" (for whatever it means) the proof, the logic behind it. I do copy and reuse the method of proving (e.g. by offering counter examples), but I feel that I agree with the logic behind those methods "innately".
What do you think? Do think this makes sense? Do you think that the Non-Statistical method, the innate understanding of logic, is NOT as "innate" as I thought, but still something that I learned with a "Statistical" method, back when I was younger?
Either way, I do believe that PART of me is not too different from a LLM model.