For a really unusual interview process, see how Jane Street, the trading firm, does it. They expect people to do well at this.[2]
Their actual interview process starts each candidate with a stack of poker chips. In each interview, candidates get to make bets on various things that look like financial trades. If they run out of chips, they're out.
A software system differs from the chaos the trading world represents. You can try to evaluate traders, but the deep truth is that very good traders can become very bad traders seamlessly because nothing guarantees any success of any kind in that domain.
When interviewing software engineers you shouldn't care that much about the abilities that someone shows during the time of a technical interview on a specific day; that's not an appropriate way to measure the skills you need for your business. Leet code is an appropriate way to measure how much such engineer is able to perform as a competitive coder. You will never need a competitive coder in your team; because what to be done in SE is not about coding less than 10 lines of code in less than X minutes to transform data A into data B. But back to candidate selection, due to the short amount of time you have to meet applicants, you finally decide to hire whatever person that succeeded to find the fastest solution to the almost scholastic problem you proposed because in the end, you fail to understand what could be the appropriate selection method : aren't all candidates more or less equivalent if you stop proposing leet code problems ? What can prove that X person with 5 years of exp is less good than that one that worked at google for 6 years ? Does it really matters than candidate Y doesn't know Go when all the languages we use aren't that truly different ? Is someone from nowhere, without any education of any sort, but that seems to be an hobbyist programmer since his childhood really a bad candidate ?
You never know, but because whether you're HR or the person with technical authority in the team, you don't want to risk anything, and you believe that a good leet code solver can do the work.
And finally, to me, that's the truth : more than 50% of the persons you decided to offer an interview would have done the work with success, and at least 20% seemed very nice, and you couldn't choose so leet code was the thing.
For a really unusual interview process, see how Jane Street, the trading firm, does it. They expect people to do well at this.[2]
Their actual interview process starts each candidate with a stack of poker chips. In each interview, candidates get to make bets on various things that look like financial trades. If they run out of chips, they're out.
[1] https://en.wikipedia.org/wiki/In-basket_test
[2] https://figgie.com/how-to-play