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

Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

chrlschn.dev

Post reply on HN
Print view View original

61–70 of 286 posts

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#61

Post by postalrat »

hcks wrote: post #30
bluehorizon2 wrote: post #26

People are forgetting why companies these days do Leetcode in the first place. The initial reason why we have Leetcode today is that Google originally determined through their interview research that the smartest candidates were the ones who were best at algorithms. Google wanted to hire the smartest people, not necessarily the best coders, so that’s why their interviews were mostly algorithmic. Of course everyone st…

It’s just a proxy measure for IQ really. They could have handed out Ravens matrices and got better outcomes prediction.

Is memorization how we should measure IQ? If so then LLMs have already far surpassed humans.

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#62

Post by indeed30 »

danwee wrote: post #41
swader999 wrote: post #6

I like this article a lot and have had great success with this tactic (as a hiring manager) over the years. I often give a page or two of code as a take home assignment to discuss at the interview. They get a chance to look at it and research it on their own. And they also get a bit of the crucible of face to face thinking and reacting on their feet while you discuss it with them in the interview.

In the year 2023 tech companies are not "buying" (supply > demand) but "selling" (supply < demand); that means the hiring process should at least be equal in terms of time spent on it. Candidates spend way more time in home assignments (hours to days), while companies spend in the order of minutes to review the assigments. We think this is "normal" because "the company is OFFERING you a job"... that couldn't be less…

That's not my perception of the industry right now. I think what you said was true two years ago, but not any more.

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#63

Post by brigadier132 »

codeTired wrote: post #51
nfriedly wrote: post #46

Earlier quoted context omitted.

Honestly, being "willing to grind" might be a good thing to target if it's a good proxy for being willing to do the work . However, I agree with the general sentiment that almost nobody is doing the research to ensure their interview process selects for the things they actually want.

I’m willing to grind if you present me with a real life problem you need solving. I’m not willing to run on a hamster wheel, it seems illogical to me.

1. Learning algorithms and practicing leetcode is not actually useless. A deep knowledge of algorithms falls under the category of high value knowledge that is not frequently needed. Its still high value when its needed and you cant predict when it is needed.

2. Its not a hamster wheel because you get a high paying job for your efforts

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#64

Post by moregrist »

codeTired wrote: post #36
bluehorizon2 wrote: post #26

People are forgetting why companies these days do Leetcode in the first place. The initial reason why we have Leetcode today is that Google originally determined through their interview research that the smartest candidates were the ones who were best at algorithms. Google wanted to hire the smartest people, not necessarily the best coders, so that’s why their interviews were mostly algorithmic. Of course everyone st…

The problem is that it might have been true at some point but now there are resources to “learn” for interviews, so currently you will get people willing to grind. Nobody happy with their pay and job is going to grind leetcode. They could just ignore leetcode and ask puzzle problems instead to test ability.

> They could just ignore leetcode and ask puzzle problems instead to test ability.

One of the hiring fads before leetcode was to ask puzzle questions, with a heavy emphasis on Fermi questions and trick questions. These had the same failure modes.

It turns out that hiring is hard, involves some risk, and always tends to illustrate Goodhart's Law: whatever process you put in place to avoid that risk will ultimately become worthless when people start optimizing for it.

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#65

Post by brigadier132 »

codeTired wrote: post #49
hcks wrote: post #43

I don’t get why people always hate on leetcodes. Anyone can train and get good at it, it makes interviews short and not time consuming, it’s a standard accros industry so you can focus on studying it and not have to read one book for each company you apply to. What would be a better alternative? Hiring based on connections? Take home assignments that takes 15 hours to complete? Domain specific knowledge grilling?

Because adults with hobbies and families have better things to do. How is this so hard to grasp? Bonus is that 90% of devs literally will not use most of that knowledge. My biggest problems are getting requirements, understanding the problem, working around a legacy system or deciphering old code, not actually writing code when requirements and documentation is on point.

> Because adults with hobbies and families have better things to do

If you really valued your time you would spend a week studying for the interview that gives you a 50pct or more increase in salary so you can work that many fewer years

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#66

Post by mjohn »

I've never tried it, but in the age of AI why not ask candidates to interact with ChatGPT to solve a problem?

In my (limited) interviewing experience the good candidates were the ones who had a good mental model for solving novel problems, were inquisitive and asked good questions. I think these attributes are also a reasonable indicator of general intelligence.

The people who I hired that were less successful were the ones who solved the coding exercises but had a poor mental model of how to solve novel problems and did not ask good questions.

Trying to use ChatGPT could take the pressure off from regurgiating coding minutae. The interviewer would be able to observe how the candidate approaches the problem and whether they can spot where ChatGPT goes wrong. It might also help introverted candidates since it feels less confrontational and there is likely to be less anxiety about unimportant details (after all any syntax mistakes would be ChatGPT's mistakes).

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#67

Post by mochomocha »

Be the change you want to see in the world.

I refuse to interview at places that are Leetcode-heavy. Never prevented me from getting 7 figures offers at FAANG-type companies.

Similarly when I interview candidates, I don't ask them Leetcode questions either. I like to get a sense of whether or not a candidate can reason in semi-unknown territories and has good intuition, not if they can parrot a textbook.

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#68

Post by kweingar »

mjohn wrote: post #66

I've never tried it, but in the age of AI why not ask candidates to interact with ChatGPT to solve a problem? In my (limited) interviewing experience the good candidates were the ones who had a good mental model for solving novel problems, were inquisitive and asked good questions. I think these attributes are also a reasonable indicator of general intelligence. The people who I hired that were less successful were t…

Could candidates opt out of using ChatGPT? I don’t think I’d be comfortable using it in an interview.

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#69

Post by indymike »

We use this process, and it has helped us hire some great developers without having to use third party staffing and headhunters:

1. Post well written job post on major job boards.

2. Immediately reply via text to every applicant (yes, we have opt in on the application)

3. Initial screening including several knock-out questions via text. Immediately let candidate know if they are not selected and why.

4. 30 minute can you write a working function from requirements code test. Allow one re-try. Immediately let candidate know if they are not selected and why.

5. Interview and code review session. Bring a side project, show us how it works.

6. Offers out.

7. Background checks start in parallel with on-boarding. First week is spent with business unit teams to learn what the company actually, really does.

8. End of first week, we occasionally have to let someone go on background issues. Otherwise, week 2 is first week coding, expect a working pull request by end of week. All candidates informed position is filled, ask if they would like to be alerted if similar job opens.

Works like a champ. Helps we write software that does this.

Average developer tenure is two years, company is three years old. We write telecom software that does texting, voice and video interviews.

Re: Interviews in the Age of AI: Ditch Leetcode – Try Code Reviews Instead

#70

Post by varispeed »

bluehorizon2 wrote: post #26

People are forgetting why companies these days do Leetcode in the first place. The initial reason why we have Leetcode today is that Google originally determined through their interview research that the smartest candidates were the ones who were best at algorithms. Google wanted to hire the smartest people, not necessarily the best coders, so that’s why their interviews were mostly algorithmic. Of course everyone st…

> The initial reason why we have Leetcode today is that Google originally determined through their interview research that the smartest candidates were the ones who were best at algorithms.

How did they measure "smartness"? The Leetcode algorithms are mostly solved problems and I know people who can "solve" any leetcode you throw at them, but they are very poor at actual real world problem solving.

It's quite a bit like saying people who memorised Wikipedia are smart, because they know things.

To me being smart is to know where to look.

Post reply on HN