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

AI killed the tech interview. Now what?

kanenarraway.com

Post reply on HN
Print view View original

191–200 of 651 posts

Re: AI killed the tech interview. Now what?

#191

Post by Clubber »

dennis_jeeves2 wrote: post #114
Clubber wrote: post #101

Earlier quoted context omitted.

>You can often tell an inexperienced interviewer from the extremely domain specific question they ask which _they_ are familiar with. Lol a bit touchy aren't we? Like I said, it's not really relevant in today's languages. It was for a Delphi/Pascal position. If you do any type of database code (like T-SQL), you would also know it. If your experience is mainly in C type languages, everything is a function so it doesn'…

>Lol a bit touchy aren't we? No. I just looked up other responses to your post. It's obvious you got exposed as being inexperienced (or an idiot), while posing to know the definitive with your "gotcha". Being inexperienced (or ignorant) is not a problem, but being cocky is.

I think you're projecting. You aren't posting this using your real name are you?!

Here's something to consider, Dennis. Instead of using any type of reasoning that maybe I'm interviewing for a language you aren't familiar with where functions and procedure differences matter, you decided to just go off the handle and call me inexperienced and/or an idiot. This is what we call in the hiring business a "huge red flag." I recommend maybe use some of that big brain you have and apply some deductive reasoning instead of just calling people names.

Re: AI killed the tech interview. Now what?

#192

Post by conjectures »

jghn wrote: post #139
CharlieDigital wrote: post #73

Code reviews. Teams are really sleeping on code reviews as an assessment tool. As in having the candidate review code. A junior, mid, senior, staff are going to see very different things in the same codebase. Not only that, as AI generated code becomes more common, teams might want to actively select for devs that can efficiently review code for quality and correctness. I went through one interview with a YC company…

This is harder than it sounds, although I agree in a vacuum the idea is a good one. So much value of the code review comes from having actual knowledge of the larger context. Mundane stuff like formatting quirks and obvious bad practices should be getting hoovered up by the linters anyways. But what someone new may *not* know is that this cruft is actually important for some arcane reason. Or that it's important that…

It's not so hard. One of the interview stages I did somewhere well known used this.

Here's the neural net model your colleague sent you. They say it's meant to do ABC, but they found limitation XYZ. What is going on? What changes would you suggest and why?

Was actually a decent combined knowledge + code question.

Re: AI killed the tech interview. Now what?

#194

Post by dingnuts »

aleph_minus_one wrote: post #137
ToucanLoucan wrote: post #131

Earlier quoted context omitted.

What is the functional difference between copying an AI answer and copying a StackOverflow answer, in terms of it being "cheating" during an interview? I think the entire question is missing the forest for the trees. I have never asked a candidate to write code in any fashion during an interview. I talk to them. I ask them how they would solve problems, chase down bugs, or implement new features. I ask about concepts…

> Languages are largely teachable, it's just syntax and keywords. This is like "learning a natural language is just 'cramming vocabulary and grammar' - voila, you've become a fluent C1 speaker". :-) Seriously: if you argue this way, you have only seen a very biased set of programming languages, and additionally, your knowledge of these programming languages is very superficial (i.e. you have never gotten to the "inte…

No, the comparison to natural languages is what is whack. If you understand the underlying concepts that programming languages pick and choose from as features, all you have to learn is what keywords map to those concepts and the language's syntax.

The comparison to natural languages would be if you could learn one language and then quickly pick up every other language of that "class" after learning how that single language works. That's not really how natural language works at all, but it does work with programming languages.

Re: AI killed the tech interview. Now what?

#195

Post by 0_____0 »

placardloop wrote: post #178
nottorp wrote: post #143

Earlier quoted context omitted.

Problem is, company A doesn't need an engineer to solve those interview questions but real problems.

“Real problems” aren’t something that can be effectively discussed in the time span of an interview, so companies concoct unreal problems that are meant to be good indicators.

Really? How short are your interviews, and how big are these Real Problems such that you can't get a sense of how your candidate would start to tackle them?

Re: AI killed the tech interview. Now what?

#196

Post by datadrivenangel »

jakubmazanec wrote: post #67

The problem isn't AI, the problem is companies don't know how to properly select between candidates, and they don't apply even the basics of Psychometrics. Do they do item analysis of their custom coding tests? Do they analyse the new hires' performances and relate them to their interview scores? I seriously doubt it. Also, the best (albeit the most expensive) selection process is simply letting the new person to do…

How do you control for confounders and small data?

For data size, if you're a medium-ish company, you may only hire a few engineers a year (1000 person company, 5% SWE staff, 20% turnover annually = 10 new engineers hired per year), so the numbers will be small and a correlation will be potentially weak/noisy.

For confounders, a bad manager or atypical context may cause a great engineer to 'perform' poorly and leave early. Human factors are big.

Re: AI killed the tech interview. Now what?

#197

Post by shermantanktop »

aleph_minus_one wrote: post #137
ToucanLoucan wrote: post #131

Earlier quoted context omitted.

What is the functional difference between copying an AI answer and copying a StackOverflow answer, in terms of it being "cheating" during an interview? I think the entire question is missing the forest for the trees. I have never asked a candidate to write code in any fashion during an interview. I talk to them. I ask them how they would solve problems, chase down bugs, or implement new features. I ask about concepts…

> Languages are largely teachable, it's just syntax and keywords. This is like "learning a natural language is just 'cramming vocabulary and grammar' - voila, you've become a fluent C1 speaker". :-) Seriously: if you argue this way, you have only seen a very biased set of programming languages, and additionally, your knowledge of these programming languages is very superficial (i.e. you have never gotten to the "inte…

In C, implicit type narrowing/widening behavior stumped me as a noob working on noob problems. “Deep C Secrets” was a revelation when I finally found it.

Then again, that’s C.

Re: AI killed the tech interview. Now what?

#198

Post by Rhapso »

nottorp wrote: post #145
Rhapso wrote: post #96

Earlier quoted context omitted.

Yeah, very bad fit. Surprised they made an offer. Folks getting mad about whiteboard interviews is a meme at this point. It misses the point. We CANT test you effectively on your programming skillbase. So we test on a more relevant job skill, like can you have a real conversation (with a whiteboard to help) about how to solve the problem. It isn't that your interviewer knew all the languages, but that the language di…

> can you have a real conversation (with a whiteboard to help) about how to solve the problem And do you frame the problem like that when giving interviews? Or the candidates are led to believe working code is expected?

Do I? yes. I also teach my students that the goal of an interview is to convince the interviewer you are a good candidate, not to answer the questions correctly. Sometimes they correlate. Give the customer what they need not what they asked for.

Do I see others doing so? sadly no.

I feel like a lot of the replies to my comment didn't read to the end, I agree the implementation is bad. The whiteboard just isn't actually the problem. The interviewers are.

Unless they change mentality to "did this candidate show me the skills i am looking for" instead of "did they solve puzzle" the method doesn't matter.

Re: AI killed the tech interview. Now what?

#199

Post by alabastervlog »

FirmwareBurner wrote: post #76
realharo wrote: post #68

Earlier quoted context omitted.

The amount of time people spend grinding leetcode instead (if you believe what they say about it online) is just as much or more.

Those are bubbles. I am aware such people exist but nobody I know grinds leetcode or does hobby programing.

It's that trimodal-comp thing. The 6-hour leetcode gauntlet is only the overwhelming norm in the top hump of that graph. It's not normal in the other two (not that it's never seen, but the companies dumb enough to do it without offering FAANG-tier comp are shooting themselves in the foot). I've had, IDK, ten plus tech jobs and have been solidly in the middle comp tier for most of those, and have exactly once encountered anything resembling a leetcode question in the wild.

I've also done hiring, and have no idea what good leetcode would do me. I'm convinced these people thinking they "dodged a bullet" on "fakers" either spotted a real faker who they would have caught with a normal conversation anyway, or else are assuming someone with a good resume who failed their test did so because they were lying and not because they choked under the unique sort of pressure interviews present when you turn them into a dancing-monkey routine (it's approximately the same kind of stress as doing an open mic night or karaoke in front of a crowd of strangers—most people have trouble with that and lots of them fall apart at least some of the times they try it). Meanwhile, anyone who can talk about the job and their career in any depth, convincingly, without giving away that they're actually either very-green or have no idea what they're doing, while in fact not knowing how to do the job, possesses a skill at least as valuable as programming, and I find it hard to believe most such folks haven't figured out they can apply that skill directly in exchange for money instead of trying to fake their way through conversational tech interviews.

I do see how leetcode is valuable if you want to ensure that most of the candidates in your pipeline would do fine before you even evaluate them, because you offer high comp and need a way to discourage candidates who definitely can't make it before they even apply, and/or if you want to make job-hopping painful as a wink-and-nudge collusion way to keep comp suppressed. It makes sense for FAANG, in a certain way, but not because it's a good way to evaluate candidates per se.

Re: AI killed the tech interview. Now what?

#200

Post by shermantanktop »

TransAtlToonz wrote: post #172

I had no idea people took hackerrank as a serious signal rather than as a tool for recent graduates to track interview prep progress. Surely it has all the same issues AI does: you have no way of verifying that the person who takes the interview actually is responsible for that signal. I don't see AI as a serious threat to the interview process unless your interview process looks a lot like hackerrank.

Your “unless” covers a huge swath of this industry, at the low end and at the high end. Excluding places that do that leaves you with what exactly? Boutique shops filled with 20 year veterans?
Post reply on HN