[kevmo314]

I'm sure the post's author doesn't need interview advice anymore but in case there are any prospective interview candidates out there, completely freezing during an interview is a super negative signal. Even if you need to manually multiply out 2's on a whiteboard it would be more productive than saying "I don't know".

In my experience the only reason you should say "I don't know" is if you're going to follow it with "but if I had to guess" or similar. Sounds like the interviewer definitely came on strong but being able to ace the psychological part of an interview is often as important or more important than the actual solution.

[withinboredom]

Asking questions irrelevant to the position is also a pretty good red flag that you don't want to work there.

[cloverich]

It's a conformity test, though it's not often viewed that way. If you interview enough you get really good at passing them, surprisingly quickly. Ultimately they want to hire people that fit a certain mold, because they understand how to utilize them. It is only a red flag if you aren't that type of person, or don't enjoy pretending to be one.

IMHO it is also the reason they can't generally make their own (good) products any more, without acquiring / copying -- that is not the type of people they hire (more artistic / risk taker types).

[arandomusername]

The point of the question is to see your problem solving skills, not if you have your 2s memorized. Anyone with basic math skills should be able to arrive at the answer.

[withinboredom]

My answer would have been the same. When asked to work it out, I would have pointed out that I simply wouldn't; not because I don't know how, but because if I need to know the number of bits it takes to store such a specific number, there is probably a reason. That reason could be storage of some form, such as wire protocols or files, and I don't want to fuck it up because I miss a step or misremember something. So, no, I won't even bother memorizing it, let alone, calculate it by hand.

[arandomusername]

It seems like you aren't capable of differientating between an interview and real work. This isn't an exercise to see if you can figure out how many bits it takes to store a number. It is a test to see how you approach a problem, if you can break it down (2^10 = 1024, 2^10*2^10 = 2^20 = 1m), and arrive at a solution. It's a simple problem that doesn't require you to be a math wizard.

[notarget137]

I come to the interview to get a real job not to please the smart "gotcha!" recruiter.

[arandomusername]

Those aren't gotcha questions.

And part of a real job involves solving real problems. In an interview, you have a very limited time, so you have to come up with questions that test the candidate's ability to solve problems within a certain timeframe.

[subarctic]

It's still good to do well at interviewing even for jobs that aren't the best fit for you, because whether you want to take those jobs or not, it will help you when negotiating for the job you really want to have those other offers in hand.

[shadowgovt]

Yeah. As a former interviewer: this question was stupid. Don't ask a candidate to solve a problem the machine can solve trivially via brute force.

[nitwit005]

Very few of us have had to do difficult math in a high pressure social situation. Having someone sitting there and pushing for instant answers is just going to sink most candidates for no real reason.

I've found you often need to directly ask the interviewer not to do what they're doing when they try this sort of thing. Sometimes they're just bored and want to talk because there has been a moment of silence, or because they're enamored with their interview question.

[arandomusername]

Difficult math? You just have to multiply 2s until you reach >56mil, and you can just round up (i.e 16k -> 32k -> 64k).

[nitwit005]

Difficult enough to need some time to think, yes.

If it was trivial, there'd be no point in asking.

[arandomusername]

Few minutes maybe.

It's a beginning question to get you warmed up.

[toastercat]

I always remember this old blog post: https://aneccodeal.blogspot.com/2014/02/interviewing-for-anx...

I haven't done a whiteboarding interview for a while, but I remember them vividly. Hot flashes, sweating, stomach churning, anytime I'd be asked a question I am definitely capable of answering, my brain would shut down and refuse to start back up again. The most apt thing I could compare it to is stage fright. Even something like simply multiplying 2's would seem impossible to me in that state of mind.

Aside from seeking professional help for dealing with anxiety, I'd recommend programmers with anxiety to avoid whiteboarding interviews [1] or at the very least let it be known ahead of time that you get stage fright.

[1] https://github.com/poteto/hiring-without-whiteboards

[ryandrake]

Yea, the problem with the high-stakes, high-pressure whiteboard-hazing interview is that it is not testing a candidate's problem solving ability. It's testing their comfort level and ability to navigate high-stakes presentation/communication pressure cookers. Which may be great if you're looking for an unflappable smooth talking PM who will be doing presentations to VPs, but probably not great if you are looking for an engineer to solve your software problems while chilling out in their Aeron chair.

That's my big problem (both as an interviewer and as an interviewee) with the current "best practices" in tech interviews. We're evaluating the wrong thing. This is how smooth talking, charismatic phonies breeze through and we find out in 6 months they can't code. Unfortunately by then, they've often gotten themselves promoted to Director.

[fragmede]

thus, take-home interview questions, but unfortunately those can be gamed too. The real one is referrals - someone you know's actually worked with them before and has good things to say, but the idea of a programmer's guild does not jive for some.

[cloverich]

One of my favorite bombs is I was asked a "real" coding question, basically a function that returns some config or another function. Anyways they somewhat lead you into a simple if statement, or a switch in my case, and then dump more examples until you come up with using a map instead. Which is something I didn't do in the interview, and most definitely (implicitly) failed. It wasn't until I was back on my side project, happily solving the same problem but using a map of course (i dont generally use switch statements in my actual code). Oh, duh lol, that's what they were asking me. They probably think I don't know how to program.

That was a turning point for me, because as far as questions go it was very fair. And yet, it had the same effect. 20 interviews later I was acing far more difficult questions, being rather polished (and ahem, interview questions aren't generally very unique). So I guess I interviews at that place too early. Maybe its useful? IDK. For me personally, it seems like really good people have a track record of shipping good stuff (or doing other good things in life), and can generally explain what it is and how it works. They usually have former colleagues you can talk to as well. Some of them even have side projects you can look at. But ah when the interview mill is in full steam, its hard to individualize to the candidate. We have a feature factory sir, and what we need are some good cogs. That's how you maximize the value you achieve from a developer after all!

[racional]

I'd recommend programmers with anxiety to avoid whiteboarding interviews [1] or at the very least let it be known ahead of time that you get stage fright.

Unfortunately acknowledging that you are in any substantive way human is a huge turnoff to these companies.

And approaching them with the suggestion that the interview process should have at least some resemblance to a 2-way street will be seen as a huge red flag, and a sure giveaway as to what an incorrigibly self-entitled primadonna you must be for even brokering that suggestion, as well.

[Borg3]

If you do NOT know, then answer that you do NOT know, unless you can speculate. Noone is omniscent. I really dont like when people with no knowledge trying guess the answer or speculate without any background. Im sure I would NOT want to work with such person.

[goostavos]

I love these threads. They always show what an absolute dice roll the interview process is. OP rates people taking a best guess at a problem they don't know as a positive character trait, this one rates it as a character flaw.

Google didn't reject the candidate, the interview loop he got rejected him for any of the myriad of knee-jerk split-second decisions made on the candidate's character in an unnatural and stressful environment.

I've been in hundreds of interview loops at $Megacorp. You can stack the deck with preparation, but sometimes you just hit a loop that doesn't click with you.

[bowsamic]

Well the goal is to create companies or teams with a consistent sense of what is positive vs a flaw

[goostavos]

At least in my neck of the woods, the "team" is generally only represented by the hiring manager. The rest of the loop is a mix of people from different teams, orgs, and roles, all pulling in their own direction.

[zzyzxd]

It actually depends on what the interviewer wants to see in the candidate. Memorizing a specific implementation? Test their experience and skill for a new design? Or just some sort of personality test and see if they are honest?

As a candidate, if I do not know, I say I do not know. I can offer my thoughts on "but if I were to design this from scratch" based on my experience, but only if the interviewer is interested in that.

As an interviewer, I hate when candidates pretend (or believe) that they know something when they actually don't, and just confidently make stuff up like ChatGPT.

[arandomusername]

You should always try to arrive at an answer before giving up. You won't get anywhere if you won't even try.

[jonas21]

You're treating this as a binary thing -- either you know something or you don't -- like it's rote memorization of facts.

But that's not how problem solving works. You need work your way from what you do know to get to an answer for what you don't. It's not guessing -- it's taking the background knowledge you do have and applying it to the problem at hand (and every engineer should have some background knowledge, even if it's just basic arithmetic).

That's what these kinds of questions are testing. Someone who immediately gives up is probably not going to be resilient in the face of new challenges.

[notarget137]

Problem solving isn't a binary thing either. You are not just capable of solving problems on the spot or not. Different people have different approaches on the same problem under different conditions. Interview is not one of the things that really shows your problem solving skills rather your skills to solve arbitrary dumb problems right there on the spot.

I've quit several interviews because of these particular questions since they are simply wasting my time and not testing my "skills".

[bowsamic]

> Im sure I would NOT want to work with such person.

Well that's the opposite of what the interviewers are looking for. They DO want to work with such people.

[pelorat]

I don't know is the correct answer. I would have brought out "get me your boss" them followed by "your recruiter is an (r-word)"

[ninininino]

I think it's a negative signal to not have the imagination to realize that anyone who has frozen during an interview due to performance anxiety is well-aware that their freezing was likely a job-candidacy-ending mistake. This is not particularly shocking information and it makes one wonder if you think they're just doing it on purpose or something or that they have control over it.

[nine_zeros]

> completely freezing during an interview is a super negative signal.

I am sure you also go to your employers and tell them that asking super unnatural questions in a super unnatural pressure environment is also a super negative signal.

You do this right? Right?

[fragmede]

you don't?

[robofanatic]

Its subjective. I think "I don't know" is a honest and better answer than to pretend that you know and waste interviewer's time.

[snozolli]

in case there are any prospective interview candidates out there, completely freezing during an interview is a super negative signal.

Remember to take you Vyvanse so that you can stay focused, and make sure to throw in some beta blockers so that you stay calm during this toxic hellscape of modern interviewing.

[EGG_CREAM]

That’s pretty dumb. I want to work with people who say “I don’t know.”

Edit: to clarify, your advice is good, what you said isn’t dumb. That criteria is dumb, in my opinion. I don’t want my colleagues to spend a bunch of my time guessing on an answer I could easily lookup or find on a calculator.

[refulgentis]

up front: this situation is really messed up, this is a close to the opposite of how I got taught to do interviews at google (albeit, nearly a decade later)

conversely, it's reductive to compare it to a generic ban on saying "I don't know"

my first job was a startup I built, starting from being a waiter. There's a lot of people who don't even try, sort of reject the premise of engaging with the question.

Saying "I don't know", then looking at the interviewer, is about the least valuable interaction you can have with an interviewer, and there's a shocking # of people who do that: whether it be freezing up, some sort of implicit commentary on the question, or a form of performance art, whatever the implicit intent is, in 99% of cases, it doesn't serve you at all.

If I ask you how to fluffle nuffle the snorbknobs, you can at least come back with "I don't know, I may have misheard you: I've never heard of fluffle nuffle. Can you give me a hint?". If you're at the end of your rope with tech interviews: "Excuse me, those aren't sensical words, I'm worried you're having a stroke"

"I don't know" and silence is NPC stuff.

[EGG_CREAM]

Fair enough, I have also worked with people like that. On the other end, asking asinine questions with limitations id never have in real life is also NPC stuff. I would say if an interviewer is getting dead eyed “I don’t know” answers a lot, it says something about the questions they are asking as much as the quality of interviewees they are getting.

[kevmo314]

I agree, but as a candidate you don't have much control over the interviewer you get.

[EGG_CREAM]

Sorry, edited my comment to clarify. Your advice is good, I just don’t like that that’s how so many interviews work.

[IncreasePosts]

In the context of the interview, the "I don't know" is said in a situation where there is no coworker or resource to fall back on. These kinds of questions happen all the time, and the business will generally not accept "I didn't know the answer so I abandoned the whole project - sorry". If you don't know the answer, you work to find it, or you try to work around it.

[keiferski]

No, you want to work with people that say, "I don't know, but I think X is true, and here's how I'd find out if that's a correct assumption..."

[EGG_CREAM]

I really don’t, not when the question is something as simple as something I can type into a calculator.

[keiferski]

The purpose of that kind of question is not to get a piece of information, it's to evaluate how you go about solving problems that you don't know the answer to and/or don't have immediately obvious ways (like using a calculator) to find the solution.

[EGG_CREAM]

Your comment has me wondering what it is specifically I find so irksome about the interview question in the blog post. In one of my first interviews they gave me a fairly well know riddle and asked me to work out the answer on a white board in front of them. I had no problem with that, actually I found it pretty fun. There’s something about this question, I haven’t figured it out quite yet, that bothers me more than other analysis questions. Maybe it’s just the flavor the writer added to it, I’ll have to figure that out.

[withinboredom]

Yeah, you type it into a calculator. In what universe would I need to know this answer and NOT type it into a calculator? It isn't because I can't work it out, it's because I am a dumb human who makes mistakes and if I am trying to work out that exact number, it is because it matters. And if it matters, I don't want to fuck it up.

[keiferski]

Well I’m not that interviewer, but I’m pretty sure that getting the exact answer is not really the point of the exercise. Understanding how the candidate thinks, does mental calculations, etc. is the point, and anyone that is hostile to this basic exercise is probably too unpleasant to be a desirable coworker anyway.

[JojoFatsani]

I’d rather someone tell me they don’t know something than feed me a line of BS then show up to the job and not be able to do it. Not that leetcode is necessarily a true indicator of competency.

[harles]

> Sounds like the interviewer definitely came on strong …

Possibly, but I’m sure these aren’t exact quotes after 15 years and I expect they reflect how the author remembers feeling a lot more than the actual conversation.

[swatcoder]

Indeed. Those interviews measure whether you can dutifully perform the dance the organization wants you to.

But if that doesn't come naturally to them, "prospective interview candidates" might also consider if that's really a problem. There remain plenty of fantastic engineering opportunities that don't involve memorizing dance steps.

[kevmo314]

Very true. Unfortunately a lot of the junior engineers I've spoken with would be happy to have any job and can't afford to be picky anymore. Such is the state of the hiring market :(