Having used LLMs for development extensively, I’ve come to realize just how important it is to have deep knowledge of the problem domain to use them effectively. If you don’t understand which algorithms are correct, what data structures make sense to model the problem, etc., then you have no way to evaluate what the LLM spits out. And it seems like universities are doing a really poor job adjusting the curriculum to ensure students actually learn fundamentals, and are able to apply them. I recall even back when I was in the university there was already a trend where the focus was primarily on coding and getting people ready for the industry rather than the theory part. And theory is really the part you need to be learning.
Having used LLMs for development extensively, I’ve come to realize just how important it is to have deep knowledge of the problem domain to use them effectively. If you don’t understand which algorithms are correct, what data structures make sense to model the problem, etc., then you have no way to evaluate what the LLM spits out. And it seems like universities are doing a really poor job adjusting the curriculum to ensure students actually learn fundamentals, and are able to apply them. I recall even back when I was in the university there was already a trend where the focus was primarily on coding and getting people ready for the industry rather than the theory part. And theory is really the part you need to be learning.