Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Stop Interviewing with Leet Code

fev.al

Post reply on HN
Print view View original

581–590 of 675 posts

Re: Stop Interviewing with Leet Code

#581

Post by adam_arthur »

overrun11 wrote: post #309
kanonieer wrote: post #152

Considering the normalization of spending hundreds of hours grinding LC questions and the industry built around whiteboard interview preparation, my (n=1) conclusion is that LC interviews is not about technical assessment at all. It's an assessment that's designed to find people who are ready to submit to an endless grind with little to no skepticism. Developers who question the technical usefulness of LC interviews…

The average college graduate makes ~60k out of school. The median law school new grad makes ~80k. Yet a large group on Hacker News acts like anyone who studies enough to get into a FAANG company fell for some horrible scam. I've studied less than 200 hours in my life and now make ~320k and expect to get ~450k when I switch jobs later this year (both remote)– I have only a high school diploma. Chalk me up as another v…

Imagine misattributing FAANG comp to any kind of personal competency.

My FB interview panel was the biggest clowncar of unwarrantedly self important people.

Anyway, pretty orthogonal to your comment, but kind of hilarious seeing what comp can do to people's ego, and perception of self

I made it rich through hustle and major contributions to a startup from the early days. Most big tech employees are a cog in the machine, along for the ride

Re: Stop Interviewing with Leet Code

#582

Post by ericbarrett »

kccqzy wrote: post #514
ericbarrett wrote: post #298

Earlier quoted context omitted.

In the FAANG interviews I've done you're never allowed to ask questions...? Maybe for clarification of the problem space, but not about the algorithm or the promise of a particular solution.

If I'm the interviewer and you don't ask questions I'm going to rate you very low on your communications skills. You may have the best algorithm in your head and write the most elegant code, but working at a company requires you to communicate your ideas and plans and code and everything else to your team. And no, communicating only at the end (code review time) is not enough. This is not a school assignment that you…

I mean, that's great, and I feel the same way. Yet every time—since the introduction of leetcode questions, anyway—as the interviewee I've been asked not to ask questions about the algorithm or the solution, just clarification of the problem space. FWIW I have been employed at several of the FAANGs or whatever they're called now and have also been on the hiring side, where I certainly did not discourage asking questions of any sort.

Re: Stop Interviewing with Leet Code

#583

Post by ystad »

Nice article - Leet code style interviews give minimal signal. The fact that we use the word Leetcode makes me think the company is hiring for the masses and it's going to be a boring job. I call such a company a Dinosaur.

I think companies should offer a choice to interviewers if they prefer to give code samples , an at home problem solving or an in-person exercise. This addresses careful thinkers, adapts for anxiety during an interview.

I do appreciate when companies ask relevant questions that they have come across rather than mundane Sudoku questions.

I have interviewed with a few companies, and Stripe's interview style stands out. Coding questions are relevant day to day style questions.

I would say Google, Amazon and Facebook set this trend and have spoiled it for all.

Unfortunately, some companies cannot think on their feet to set a different approach. Maybe it's in your best interest to avoid these places.

Re: Stop Interviewing with Leet Code

#584

Post by dekhn »

UncleMeat wrote: post #413
dekhn wrote: post #366

Earlier quoted context omitted.

You are factually wrong. Not only that, you're conflating several different processes and team rules. Many googlers use brew to install applications on their laptops. This not against policy. Other googlers work with code stored directly on their laptop. There may even be developers who are obtaining deps (for their own builds) from brew. The problem with the brew author is that he had every opportunity to make himse…

It is true that this was the problem with the brew author. Homebrew being part of the typical Google workflow is entirely independent of the situation. But, as usual for internet discussions, it is fun to rathole on side conversations.

The best part is, I was defending the use of homebrew (which I absolutely hate) and local development (which is far inferior, IMHO, to blaze/forge/citc/piper). I had really hoped releasing abseil/bazel would help but sadly, it was done too little, too late.

Re: Stop Interviewing with Leet Code

#585

Post by dekhn »

visarga wrote: post #564
dekhn wrote: post #481

Earlier quoted context omitted.

My question is fairly open and allows me to continue asking significantly more complicated questions. So far, nobody has (for example) asked enough clarifying questions to determine that a bloom counting filter could potentially solve the problem- most people just store ASCII string keys in a hash table, which wastes tons of memory and requires a lookup for every string. So usually I end up after 45 minutes finding t…

Bloom filters have collisions. How would that work with counting sequences?

Because (and this is absolutely a clarifying question I would love to get), approximate counts are acceptable.

https://bmcbioinformatics.biomedcentral.com/articles/10.1186...

Re: Stop Interviewing with Leet Code

#586

Post by jjav »

brhsagain wrote: post #148

All of these supposed "flaws" of leetcode are by design. Big companies want people who are smart enough to do the work, but obedient enough to put up with all the bullshit that comes with working at a big company. What person better matches that than someone who's able and willing to study for and pass a tech version of the SAT? Every anti-leetcode article I read is some version of "leetcode is bad because it measure…

> No, we all know it measures those things, and those are exactly the things the measurers want to measure.

I wouldn't be so sure everyone realizes it doesn't measure anything relevant to actual job performance although it should be obvious (since nobody whiteboards algorithm puzzles all day in an actual job).

In all of these threads there will be many people saying they only want to hire the top% of developers (as if it was a strict linear scale, as opposed to hundreds of intertwined skills) and they justify leetcoding as the way to do that in a belief that top leetcoder is somehow a top software engineer, instead of realizing these are unrelated skillsets.

Re: Stop Interviewing with Leet Code

#587

Post by bko »

brhsagain wrote: post #148

All of these supposed "flaws" of leetcode are by design. Big companies want people who are smart enough to do the work, but obedient enough to put up with all the bullshit that comes with working at a big company. What person better matches that than someone who's able and willing to study for and pass a tech version of the SAT? Every anti-leetcode article I read is some version of "leetcode is bad because it measure…

I think you're overthinking it. I think very simple questions are appropriate just because you want to see if the person knows how to code. Not knows how to balance a binary tree, but literally knows how to code. You'd be surprised how many programmers I've interviewed that literally couldn't get anything to compile. To be fair, you see the same in the finance world where someone lists Excel and data analysis as skill sets and then doesn't know how to do a vlookup.

People write all sorts of stuff on their resume. They took a C course in college 10 years ago and write C as a skill. They could be programming managers and not know what a pull request is.

I think its definitely gone too far, but basic coding proficiency is essential IMO.

Re: Stop Interviewing with Leet Code

#588

Post by dekhn »

lazyant wrote: post #573
dekhn wrote: post #481

Earlier quoted context omitted.

My question is fairly open and allows me to continue asking significantly more complicated questions. So far, nobody has (for example) asked enough clarifying questions to determine that a bloom counting filter could potentially solve the problem- most people just store ASCII string keys in a hash table, which wastes tons of memory and requires a lookup for every string. So usually I end up after 45 minutes finding t…

The problem with this question is that outside the junior people who won't understand the hints, many good programmers would pass or fail just based on whether they have seen a similar problem or know what a Bloom filter is. CS is so broad that you may have been working for many years and never come across one topic that the interviewer deems standard.

I don't fail people who don't suggest a bloom filter. What I meant was: knowledge of probabilistic data structures and knowing this is a situation where they could be used is a sign of a more experienced programmer (likely one who's worked in clickstream processing).

If all you do in my interview is know that a hash table or other associative data structure is good for maintaining counts, that using a full ASCII byte to store symbols from { A G T C } is wasteful and compress the letters to 2 bits each, you can pack multiple 2 bits into bytes, and can write a function that codes and decodes such data, I'm happy and you get a pass. I'm even happy to sit there and help people through nearly all the steps of the codec. Not trying to trick anybody or select for obscure CS knowledxge.

To me the real question is, how much hinting is reasonable for the coding and decoding function? Many programmers (including senior ones) struggle to implement:

x Since I never use bit munging in my day job (it's 90% python data science) should I really ding somebody for not knowing about left shift or or-assignment?

What I really don't get is why people immediately jump to trying to implement huffman coding, RLE, or lookbacks as a solution to reduce the size of the DNA string. I still wonder if how I present the question is giving everybody a fair chance to shine.

Re: Stop Interviewing with Leet Code

#589

Post by adregan »

foobiekr wrote: post #436
adregan wrote: post #302

Earlier quoted context omitted.

A simple problem with constraints can be really fun. For example, implement FizzBuzz without using any looping constructs.

The variant of fizzbuzz that I use is "read and categorize a line at a time from stdin." You'd be amazed how this one small change makes the discussion quite different.

In your variant do the integers increase in order?

Also is the input always integers or is there some “junk” mixed in?

Re: Stop Interviewing with Leet Code

#590

Post by TulliusCicero »

cmrdporcupine wrote: post #467
TulliusCicero wrote: post #426

Studies show that one of the most effective interviewing tools is a test for general mental ability: https://www.semanticscholar.org/paper/The-validity-and-utili... Now, IQ tests are of dubious legality, at least in the US, but algorithmic coding questions basically get you an IQ test crossed with a programming skill check: win-win. All the ire about how you don't actually invert binary trees or whatnot during your r…

Sorry, I've had both my kids go through the WISC-V "IQ" test in the last few years and there's no single "intelligence" number that comes out of that that would be meaningful for employment. It measures all sorts of things and you can be off the charts high in one kind of mental / reasoning skill while being average or below average in others and yet that still might tell you nothing about how well someone writes cod…

Studies have general mental ability tests as comparable to work sample tests for predicting job performance.

That people are guessing that it wouldn't work is irrelevant, when we have evidence that it does.

Post reply on HN