As ThePrimeagen said, if you just read the A.I.‘s explanation rather than deriving it yourself by scrutinizing the code, it won’t stick, because you haven’t exercised that mental muscle.
As I see it, A.I. for programmers as a learning aid is likely to shape up to be like those belly-jiggling machines from the 1950s and all the other things we humans have desperately made in an effort to try to get the effects of exercise without actually exercising.
As for the non-learning aspects… the whole point of the post is about how much the author felt their abilities had atrophied when the A.I. was down.
A.I. for programmers as a learning aid is likely to shape up to be like those belly-jiggling machines from the 1950s and all the other things we humans have desperately made in an effort to try to get the effects of exercise without actually exercising.
I don’t know if “robust” is the right word for it. Fierce maybe?
It contains quite big accusations without backing them up (“works for Amazon” is not sufficient).
It’s not precise with facts nor terminology. LLMs are overhyped, but framing them as just a dumb autocomplete is an exaggeration in the other direction.
I see your post and raise you ThePrimeagen’s coverage of AI is Creating a Generation of Illiterate Programmers.
As ThePrimeagen said, if you just read the A.I.‘s explanation rather than deriving it yourself by scrutinizing the code, it won’t stick, because you haven’t exercised that mental muscle.
As I see it, A.I. for programmers as a learning aid is likely to shape up to be like those belly-jiggling machines from the 1950s and all the other things we humans have desperately made in an effort to try to get the effects of exercise without actually exercising.
As for the non-learning aspects… the whole point of the post is about how much the author felt their abilities had atrophied when the A.I. was down.
omg this is hilarious :D can’t unsee now.
Robust rebuttal: https://anatol.versteht.es/blog/llmnt/
I don’t know if “robust” is the right word for it. Fierce maybe?
It contains quite big accusations without backing them up (“works for Amazon” is not sufficient).
It’s not precise with facts nor terminology. LLMs are overhyped, but framing them as just a dumb autocomplete is an exaggeration in the other direction.
nah, it really isn’t. the rest is PEBKAC.