I think most people define "Intellisense" as "IDE suggestions based on static anaysis results". Sometimes it blends a bit of heuristics/usage statistics as adde

I think most people define "Intellisense" as "IDE suggestions based on static anaysis results". Sometimes it blends a bit of heuristics/usage statistics as added feature depending on the tool. They are mostly deterministic, based on actual AST of your code, and never hallucinates. They may not be helpful but can never be wrong. On the other hand, LLMs are completely different -- based on machine learning and everything is random and about statistics. It depends on training data and context. It is more useful but make a ton of mistakes.