Why should I relearn algorithms?

I read Dan Luu's programming books page recently, on which it says about algorithms: "Why should you care? Well, there's the pragmatic argument: even if you never use this stuff in your job, most of the best paying companies will quiz you on this stuff in interviews. On the non-bullshit side of things, I find algorithms to be useful in the same way I find math to be useful. The probability of any particular algorithm being useful for any particular problem is low, but having a general picture of what kinds of problems are solved problems, what kinds of problems are intractable, and when approximations will be effective, is often useful."

I learned algorithms in college, many years ago. I still have a good idea of when to use data structures, generally. I probably couldn't pass a FAANG-style interview without a couple months prep now, despite giving them every other week. I couldn't tell you when to use a red-black tree over a b-tree anymore. Frankly, as a dime-a-dozen megacorp SWE, that's not really something that would come up in daily work. IME, management isn't going to give you time to care about performance specifics - maybe, if there's a library that offers a convenient API and some deceptive benchmark graphs, you use that. I think I've used a prefix trie, once, in my entire professional career. The rest is just standardlib abstractions where I don't really care about what's happening below. Someone much smarter than me is making optimizations based on the latest papers for any given sorted() implementation.

What other motivation is there to relearn algorithms except for clearing the interview hurdle? Especially when free time is limited, and you could instead _finally_ get around to reading Designing Data-Intensive Application or Crafting Interpreters, learning Rust/Go/CultLang, or dipping your toes into something "fun" like reverse engineering?