People are still missing the point

I am seeing a lot of people in my last post getting defensive because FAANG and this and that still asks leetcode. Obviously a 10 year hiring standard does not vanish overnight, but let's not be so narrowsighted. Mind you I went through the SAME grind to get myself to 2400 on lc. And that is exactly my point. Even at that rating, I know lc does not prepare you for real engineering work. So I really doubt people spending this much time on lc are as strong in those areas as they think they are. If every person in the pool has 500+ problems solved, then having 500 solved is worth basically zero as a differentiator. It is the new high school diploma. It might help you get through an ATS or OA, but it does not get you the offer. The truth is people are overinvesting in the easiest and most straightforward part. You solve a problem, get a green checkmark, and feel productive. Meanwhile the stuff that actually takes longer to build intuition for gets pushed aside because it is slower, messier, and way harder to measure. You do not build engineering intuition from a few videos or from memorizing patterns. That comes from building, debugging, breaking things, fixing them, and slowly understanding why systems behave the way they do. I have people reaching out to my company with 1000+ lc problems but they still cannot explain how to handle backpressure in a distributed pipeline or why their async logic causes race conditions. A lot of people are grinding lc like crazy while being completely underdeveloped in the areas that actually matter once the OA is over. So stop arguing about whether lc still exists and start worrying about whether you have any real engineering skill to show for after the screen. If you do, good, this post was not meant for you.