I have seen multiple posts about this "remembering EVERYTHING", and I think they miss the point. Also they don't quote the context:

> So, coming back to the initial starting point that “you don’t have to remember anything”. The opposite is true. You have to remember EVERYTHING.

I see it like this: it is absolutely wrong to think that you don't have to remember anything. In fact, ideally you would remember everything. The more you remember, the better you can think. Now in practice, it's impossible to remember absolutely everything, so we should strive to remember as much as we can. And of course we need to be clever in how we select what we remember (but that seems obvious).

The point is really that it is common to say "it's useless to remember it because you can ask your calculator or an LLM", and the article strongly disagrees with that.

> More important than specific factual knowledge, you need two things: strong conceptual models for whatever you're doing and tacit knowledge

And the more experience with computers I get, the more I realize that there's actually not that many pure unique and mutually orthogonal _concepts_ in computer science and software engineering. Yes, a competent engineer must know, feel, live these concepts, and it takes a lot of work and exposure to crystallize them in the brain from all the libraries, books, programs, architectures one has seen. But there's not a lot of them! And once you are intimate with all of them, you can grok anything computer-related quickly and efficiently: because your brain will just wuickly find the "coordinates" of that thing in the concept space, ans that's all you'll have to understand and recall later.