Live data from Hacker News

AI killed the tech interview. Now what?

kanenarraway.com

621–630 of 651 posts

Re: AI killed the tech interview. Now what?

#621
> Using apps like GitHub Co-pilot and Cursor to auto-complete code requires very little skill in hands-on coding.

this is a crazy take in the context of coding interviews. first, because it's quite obvious if someone is blindly copy and pasting from cursor, for example, and figuring out what to do is a significant portion of the battle, if you can get cursor to solve a complex problem, elegantly, and in one try, the likelihood that you're actually a good engineer is quite high.

if you're solving a tightly scoped and precise problem, like most coding interviews, the challenge largely lies in identifying the right solution and debugging when it's not right. if you're conducting an interview, you're also likely asking someone to walk through their solution, so it's obvious if they don't understand what they're doing.

cursor and copilot don't solve for that, they make it much easier to write code quickly, once you know what you're doing.

Re: AI killed the tech interview. Now what?

#622
post #136

Earlier quoted context omitted.

They generally do not know what they are looking for. They are generally untrained, and if they are trained, the training is probably all about using leetcode-type problems to give out interviews that are sufficiently similar that you can run stats on the results and call them "objective", which is exactly the thing we are all quite correctly complaining about. Which is perhaps anti-training. The problem is that the…

> The Western philosophical mindset doesn't handle this sort of thing very well. Mind elaborating on that?

It is a gross oversimplification but you can look at the Western mindset as being a reductionistic, "things are composed of their parts" sort of view, and the Eastern mindset as a holistic mindset where breaking things into their components also destroys the thing in the process.

The reality isn't so much "in between" as "both". There is a reason the West developed a lot of tech and the East, despite thousands of years of opportunity, didn't so much. But there is also a limit to the reductionistic viewpoint.

In this case, being told that the only way to hire a truly good developer is to make a holistic evaluation of a candidate, that you can not "reduce" it to a checklist because the very act of reducing it to a checklist invalidates the process, is something that a lot of Western sorts of people just can't process. How can something be effectively impossible to break into parts?

On the other hand, it is arguably a Western viewpoint that leads to the idea of Goodhart's law in the first place; the Eastern viewpoint tends to just say "things can't be reduced" and stop the investigation there.

This is highly stereotypical, of course, and should be considered as an extremely broad classification of types of philosophy, and not really associated directly with any individual humans who may happen to be physically located in the east or west. Further as I said I think the "correct" answer is neither one, nor the other, nor anything in between, but both, so I am not casting any shade on any country or culture per se. It is a useful, if broad, framework to understand things at a very, very high level.

Re: AI killed the tech interview. Now what?

#623

Earlier quoted context omitted.

I've never seen hiring completely in the domain of HR. HR filters incoming candidates and checks for culture fit etc, but technical competency is checked by engineers/ML folks. I can't imagine an HR person checking if someone understands neural networks.

HR involvement is unavoidable at big companies; and basics like "years of experience for payband" can cause issues. They fundamentally do not understand the job, but somehow have to ensure its not a biased hiring process.

Yes, and it is kind of necessary when hiring people from outside trusted networks. HR makes sure if people are who they say they are, background checks, and so on. Years of experience and so on are crude filters and should be bypassable by the team/hiring manager if the candidate meets the requirements. I know in large companies this can become political.

Re: AI killed the tech interview. Now what?

#624

Earlier quoted context omitted.

Yeah, that sounds hypercapitalist. At least I hope those engineers make enough money to retire when they burn out.

Unfortunately hedonic adaption often prevents that.

I could see that happening, especially given that they're young and feel powerful, nearly invincible. I experienced a similar attitude form younger folks outside of FANG, some kind of derision of experience and advice from people who have been around and choosing instead to obsolete it. After the first burnout may of them become a lot more thoughtful and humble. I may have been like that too when I was young, I don't remember being exactly like that but maybe I was just not aware. I still did respect experience and I still do to this day.

Re: AI killed the tech interview. Now what?

#625

The best interview process I've ever been a part of involved pair programming with the person for a couple hours, after doing the tech screening having a phone call with a member of the team. You never failed to know within a few minutes whether the person could do the job, and be a good coworker. This process worked so well, it created the best team, most productive team I've worked on in 20+ years in the industry,…

Some of us find the prospect of pairing with an unknown person in an unknown environment, and against the clock, to be very stressful.

I've had a great career and very positive feedback from every employer. I would do absolutely terrible in this kind of interview.

Re: AI killed the tech interview. Now what?

#626
post #248

The best interview process I've ever been a part of involved pair programming with the person for a couple hours, after doing the tech screening having a phone call with a member of the team. You never failed to know within a few minutes whether the person could do the job, and be a good coworker. This process worked so well, it created the best team, most productive team I've worked on in 20+ years in the industry,…

I dislike pair programming interviews - as they currently exist - because they usually feel like a time-crunched exam. You don't realistically have the freedom to actually think as you would in actual pair programming. i.e. if you wag your tail chasing a bad end for 15 mins, this is a fail in an interview, but it's pretty realistic of real life work and entirely a non-problem. It's probably even good to test for at i…

When I was doing interviews at my old job we did an hour of pair programming and we very intentionally designed the exercise such that

1) if you're qualified to do what we're hiring you for this stuff should be your bread and butter. We were doing restful web apps so the prompt might be "A restful API that takes in a string and returns that same string but backward." If you chase your tail on that for 15 minutes, you're not right for this job.

2) We weren't looking for a specific implementation. You want to make it a GET call with a query param that returns just the string? Neat. A POST with request and response DTOs? Also neat. We're not gonna ding you for doing it one way rather than the other but we're definitely gonna talk to you about why you made the choice that you did and maybe try to tease out another choice you could have made so that we can ask you to compare them. Again, the only wrong answer here is one you can't defend.

3) I'm not here to test your ability to memorize a particular language's syntax, your IDE's autocomplete or your ability to google. If you can articulate that you're trying to create a POST endpoint I'll give you that the correct annotation with Spring is @PostMapping or little things like that.

Re: AI killed the tech interview. Now what?

#627

Earlier quoted context omitted.

I'm curious, what are the recent cases that were brought up where the company lost?

Meta had one around 2020 with EEOC. The Harvard Case I mentioned above (supreme court) , Activision, Dell, a few others on the top of my mind.

I just looked them up, as I recalled those cases differently, and it doesn’t look like anything has to do with declining an applicant due to them not being the right fit for the team.

Re: AI killed the tech interview. Now what?

#628

Earlier quoted context omitted.

“High end” was meant as shorthand for FAANG …high comp, not necessarily high tech complexity. You are correct about the deficiencies of the whiteboard interview. It is not a sane way to hire an individual. It makes sense as a way to hire someone in the top 20% from a large unfiltered pool. So wrt high/low, that’s what FAANG companies have to do, and for many nontechnical companies they outsource this work or emulate…

Ah, makes sense. I just abandoned the code interview altogether and ask them questions about process. It's a very simple workaround, but very effective. I'll admit it helps that there are very few problems these days outside of specific problems to solve that require a high degree of technical competency to tackle.

For the more senior candidates that I interview, I've seen people who talk a great game but cannot deliver, and others who can. Having them write fizzbuzz-type code is a primitive check that has led to surprising results - 25 year industry veterans with all kinds of great projects delivered, and...they can't write a for loop? They can't declare a variable? They don't know what final/const really does? I hate asking them to do it but I can't really stop.

Re: AI killed the tech interview. Now what?

#629
post #256

Earlier quoted context omitted.

> Company A wants to hire an engineer, an AI could solve all their tech interview questions, so why not hire that AI instead? Interview coding questions aren't like the day-to-day job, because of the nature of an interview. In an hour-long interview, I have to be able to state the problem in a way the candidate can understand, within 10 minutes or so. We don't have time for a lecture on the intricacies of voucher cal…

>Interview coding questions aren't like the day-to-day job, because of the nature of an interview. You have missed his point. If the interview questions are such that AI can solve them, they are the wrong questions being asked, by definition. Unless that company is trying to hire a robot, of course.

That’s silly. It’s like saying that a car can do a 100m dash faster than a human so it’s not a useful test for selecting players for an NFL team.

Re: AI killed the tech interview. Now what?

#630
post #302

Earlier quoted context omitted.

Part of the problem is there just aren't a lot of people out there who can correctly judge that level of experience, and looking up the spectrum tends to simply look weird.

I agree. Do you have any thoughts on how to mitigate this? After all, its in your best advantage and the companies to hire someone with talents because of the value they can bring.

Not really, except having a lot of experience on board from the start.

Which will also naturally attract more of the same.

Post reply on HN