The flip side is that a DS&A toy problem grind is an actionable item that approximately anyone who can already bear the cost of their own training can work on, with legible indicators of their progress.
I know it's popular to throw shade on interviewing for "wasting" candidates' time on this instead of "learning to build actual things", but even if DS&A toy problems aren't the best-correlated with actual work performance, the fact is that (a) they're well-enough correlated, by the standards of scalable interview assessments (b) contrary to popular belief around these parts, presenting "actual things" is much easier to bullshit your way through than a few hours of talking through toy problems (c) every job opening gets a deluge of bullshitters.
You do realise that these leetcode rounds are virtual and there are “agencies” on the internet that would take this test up for you?
When I interview people, I have to make sure that the person actually wrote the code by asking some arbitrary questions about it in a subsequent rounds
We’ve had cases where people committing this fraud did manage to slip through the cracks but thankfully here it’s very easy to fire people once we found out.
I know it's popular to throw shade on interviewing for "wasting" candidates' time on this instead of "learning to build actual things", but even if DS&A toy problems aren't the best-correlated with actual work performance, the fact is that (a) they're well-enough correlated, by the standards of scalable interview assessments (b) contrary to popular belief around these parts, presenting "actual things" is much easier to bullshit your way through than a few hours of talking through toy problems (c) every job opening gets a deluge of bullshitters.