The Horrifically Dystopian World of Software Engineering Interviews

There's another dimension to job interviews, the "black holes" ( Thanks for Submitting Your Résumé to This Black Hole : https://www.nytimes.com/2017/03/25/...ubmitting-your-resume-to-this-black-hole.html ).

One observation of mine, there are regular job boards like LinkedIn or Indeed, which offer the following options for applying to a job:
1) Easy apply on the job board itself, fill a few fields and you're done. If you don't get any reply, at least you didn't put much effort into it either.
2) Takes you to some other job board or employer site, but still an easy apply of a slightly different form. Again, at least it's not much hassle to apply.
3) The horror: takes you to some mega-corporate internal web application. You need to register with a mail address first and fill minutiae details about gender, race disability status etc. Then it starts a process that presumably takes several steps, each step consisting of a clunky web form with dozens of fields. Bugs all over the place, some fields won't focus, others won't accept your input, eventually after putting quite a struggle you fill them all and press submit, only to be "greeted" with an error message about a field not being validated and the rule is never clear on what you need to enter there, bonus points all your painfully completed fields are now reset to an brand new empty default. Irrationally you insist and make a personal goal to get the damn form accepted, hurray, it's accepted. Obviously it was only the first level out of a series of 10... sometimes you have to admit defeat and abandon the game but if you keep struggling eventually congratulations, you have submitted your resume.

#3 it's always, inevitably a black hole. Never ever got an answer back or an acknowledge that even a machine has processed my input.
 
Aquarians said:
"Start your own company" advice is exactly equivalent to "win the lottery". I'm surprised people still throw this in conversations.



Here you are correct. Size matters :D If you have 100s of guys competing for a single seat, it's gonna be tough. Still it doesn't have to be "Jewish problems" level of abuse but it seems it always goes there asymptotically.



It might be it's genuinely fine to work for you. I rarely change workplaces since as long as the current one provides a decent income (and in software they always did), I don't really expect much else in terms of compensation. My philosophy was that I'll never get wealthy through a job, that will and can only be the result of me starting a successful business, so a job is only there to support me until I hit it big. "Temporarily embarrassed millionaire" mentality, that is :D

Part of the reason it's easier to get a job at a smaller compay is still statistics: 2-3-10 competitors on a seat instead of 1000. Also it might be still size why smaller companies treat you as a human when you interview, even if you don't get hired. Can't approve or disprove for people working for giant companies instead of smaller ones, my personal preference is somewhere in the middle. I loathe mastodonts but small businesses also have a stability problem and may not last for long.
More...
Well I've won the lottery twice then, lucky me! Given that we have hundreds of thousands of successful businesses and a few dozen large lottery winners, it's not really an apt comparison. Not to mention the fact that although luck does play a part in businesses success, you have far more influence of the eventual success or failure of your business than a game of pure chance.

It really goes back to this fixation on the Googles of the world. If, when I say you should consider starting a business you think it has to turn out a Google or bust, then yeah, it's a lotto. If you're happy making a few million a year once established, then your odds are actually far better. If you're happy earning the equivalent of a Google software engineers salary, then I'd suggest that your odds are actually far higher of achieving that than your odds of getting a job as a Google software engineer.
 
I used to give those dystopian interview questions. I didn't read this whole thread but if anyone is interested in what and why:

Big Tech is desperately worried about "scale". Things need to be "scalable". It's the core of their mindset, and permeates into other parts of life. At the small end, the interview questions we ask are all about how to make code go faster, depending on the size of the input. (you may hear this by it's gangster name Big Oh). Typically, the interview question is something like "how would you find a a number in a sorted list?" Upon answering the question, the interviewer will then want the solution to scale: "How would you find a number in a sorted list if the list was was a bazillion elements long?" Whether or not the question has any relation to the job itself doesn't matter. It might be needed at some point in your 2-5 years at the company, so it get's asked at the beginning.

The "scale" needs to work as you zoom out from a single interview, because now you need to scale the interview process itself. "How do we interview the hundreds of thousands of candidates that apply here?" Also: "How do we have a consistent hiring bar?" And finally: "How do we make sure the interviewers aren't racist, sexist, biased, bigoted, or prejudiced?" All of these are part of "scaling" the hiring process. As a result, Big tech has centered around a small set of programming puzzles, that leave little room for the interviewer to inject any of their own questions, biases, or leanings into the process. Every interviewer asks from the same pool of completely neuter questions. There is no emotion, and no room for empathy, because that would be human, and humans don't scale. Since every question is the same, there is a huge amount of data to normalize the answers to, so that a completely objective measure can be taken. The interviewers themselves are normalized, showing if they have any leanings in the interview feedback, or if they recommendation is unusual.

I was one of those interviewers, and after sitting on both sides of the table, there really isn't a better alternative to date. It's the worst solution out there, except for all the other ones we tried.

On the flip side though, because the process is so bland, unfeeling, and formulaic, it's actually quite easy to get hired by FANG these days. You practice the interview questions ahead of time, effectively memorizing the answers. The company knows this, but there aren't that many interview questions, so they end up using them anyways. The requirements for the job used to be "be really smart", but it has become "be pretty smart, and grind away at coding for a few weeks before the interview".

I'm curious what the quant firms interview process is like, since I assume it would be a mix of trading and coding.
 
Pretty similar in quant space, same questions get asked. The number of puzzles in public domain is only that large. Anyone who vigorously prepares for puzzles and algorithms should pass that stage. What is different, though, is that puzzles are only part of the interview. A large portion is "stressing" interviewees through abrasive attitude, irrational behavior, and pressure. The difference is that quant traders sit inches apart all day every day, not in cubicles or comfortable conference rooms. The job entails constant pressure and stress, risk taking, with lots of money on the line. Weak personalities or those who can't deal with stress can't survive and that is tested at the very beginning of a quant career. It does make sense to test for stress management and behavioral traits as training someone who knows virtually nothing out of college costs a lot of money. The difference imo is that most excellent coders in IT jobs come already equipped with the tool set, whereas in financial trading new hires come with brain juice and survival instincts and not much else. All else must be taught and learned on the job for years. What I find really annoying though is when certain interviewers don't understand when they have a mid career or senior trader sitting in front of them. Posing puzzles to them is imo an affront and arrogant because it disrespects the skills they have already acquired. Does not happen often but it does occasionally happen. You would be surprised how unprepared and untrained some interviewers are but I bet this also happens in IT space. Much better to engage with experienced hires on a level respectful and worthy of them and actually talk about the traded products, approaches to trading and quant analysis, team structure, and corporate culture. Those are much more important aspects of the job for an experienced hire. Later on in my career I refused to even engage in puzzles on a number of occasions and still proceeded in the interview process. Like with everything in life that only works when one comes with a pedigree and valuable and marketable skill set.



Spaghetti Code said:
I used to give those dystopian interview questions. I didn't read this whole thread but if anyone is interested in what and why:

Big Tech is desperately worried about "scale". Things need to be "scalable". It's the core of their mindset, and permeates into other parts of life. At the small end, the interview questions we ask are all about how to make code go faster, depending on the size of the input. (you may hear this by it's gangster name Big Oh). Typically, the interview question is something like "how would you find a a number in a sorted list?" Upon answering the question, the interviewer will then want the solution to scale: "How would you find a number in a sorted list if the list was was a bazillion elements long?" Whether or not the question has any relation to the job itself doesn't matter. It might be needed at some point in your 2-5 years at the company, so it get's asked at the beginning.

The "scale" needs to work as you zoom out from a single interview, because now you need to scale the interview process itself. "How do we interview the hundreds of thousands of candidates that apply here?" Also: "How do we have a consistent hiring bar?" And finally: "How do we make sure the interviewers aren't racist, sexist, biased, bigoted, or prejudiced?" All of these are part of "scaling" the hiring process. As a result, Big tech has centered around a small set of programming puzzles, that leave little room for the interviewer to inject any of their own questions, biases, or leanings into the process. Every interviewer asks from the same pool of completely neuter questions. There is no emotion, and no room for empathy, because that would be human, and humans don't scale. Since every question is the same, there is a huge amount of data to normalize the answers to, so that a completely objective measure can be taken. The interviewers themselves are normalized, showing if they have any leanings in the interview feedback, or if they recommendation is unusual.

I was one of those interviewers, and after sitting on both sides of the table, there really isn't a better alternative to date. It's the worst solution out there, except for all the other ones we tried.

On the flip side though, because the process is so bland, unfeeling, and formulaic, it's actually quite easy to get hired by FANG these days. You practice the interview questions ahead of time, effectively memorizing the answers. The company knows this, but there aren't that many interview questions, so they end up using them anyways. The requirements for the job used to be "be really smart", but it has become "be pretty smart, and grind away at coding for a few weeks before the interview".

I'm curious what the quant firms interview process is like, since I assume it would be a mix of trading and coding.
More...
 
Last edited:
stochastix said:
you are absolutely perfectly correct. when I interviewed at chaseand bofa it was literally south Asians as far as they eye can see.. not even any other type of '
minorities'
More...

When I consulted for a major retailer in Minnesota, I asked him why he chose Tata Consultancy Services over much better consulting firms. His reply was that TCS will give you a senior data scientist at the same rate as a fresher and generally undercharge 25%. Of course, one competent US data scientist is worth about 10 offshore idiots.
 
Another example of abuse nonchalantly posted as normal and something candidates should get accustomed to: https://www.spakhm.com/p/how-to-interview-engineers

A lot of programmers reading this article seem to take it for "satire", when in fact it's pure evil. Comment on Hacker News that nails it:

"Even if it's satire — which I think it should be more clear when you write about such a delicate topic — I think it represents quite well the levels of "assholiness" the software industry is reaching these days. I think they can get away today with so many silly hoops because the supply of engineers is quite high these days, and I guess, the FAANG industry standard is the "fair" way to identify the best candidates.

Engineers now are putting up with all this because there is not much choice than to spend more time training for this kind of interviews.

No matter how much do they talk about talent, judgment, IQ or whatnot, the fact is the more you train, the better you get at this kind of interviews, independently of how capable in your day to day job you are as an engineer. Proof that training works is the whole emerging cottage industry around interview preparation that is thriving these days.

Perhaps this is a bubble that will burst at some point, but I'm afraid the bar of ridiculousness will keep raising. If any of the FAANG companies started requiring to complete a marathon in less than 5 hours, they will still find great candidates that will train hard to achieve it.

The rationalization after fact would be something like the ability to complete a marathon correlates well with the ability to deliver a project or whatever, but they are just selecting for engineers that will do whatever it takes to get the position, no matter how silly the requirement is."
 
To me, you have to be either an absolute genius or a total sucker to interview with someone you don't already know.
If you are not a genius, your time would be better spent on knowing more people.
 
Fresh horror story :cool:

"The computers rejecting your job application": https://www.bbc.com/news/business-55932977
And Hacker News thread: on it https://news.ycombinator.com/item?id=26063523

So we're already living in a world where getting a job is at the discretion of an "artificial intelligence". People report that unless you specifically write your CV in a way that pleases "the AI overlords", you never get over the initial screening process.

Isolated from HN comments:

1) Guy writes a manual CV, helped by human reviewers and gets absolutely no answers. (And rightfully so, like one guy who comments: "Anybody who still writes out their resume by hand is considered a luddite weirdo, and definitely not someone you'd want to have join your company.")

"I went to a top 40 finance program, interned at Morgan Stanley, a hedge fund, and a respectable incubator, but had a hard time getting replies to my job applications. [...] I had my resume looked at by everyone [HUMAN] I could get critiques and edits from. I wasn't getting answers from my applications."

2) But then same guy hires an "expert" in writing the very same CV, same content but in a way that "pleases" the AI masters:

"I found 5 people claiming to optimize resumes for algorithms for the sorts of jobs I wanted. [...] I picked the one who appeared the best. Not only did she have some personal proximity to the space, but she seemed to know all of the nuances of how to please the resume algos. It was the best $250 I ever spent.

Point is [...] it feels arbitrary but that's how the game works now if your a job seeker."
 
I have a great friend I grew up with who is one of those complete geniuses. He is not only a genius, but a people person who can talk to anyone about anything and has not a care in the world how smart someone else maybe. Anyway, he has worked for the government for years, but grew tired of various things in his world with computers and the government, so decided to find a job in the private sector figuring he would also see a huge pay increase and new challenges. He interviewed with 2 of the big tech companies and went through the interview stupidity of challenges and puzzles which he said he aced and had good conversations with the various interviewers, but ended up going back to his government job, with a hefty pay raise and he gets to hire the people he works with.
He never really gets ticked off, but he blew a gasket about the process. He insisted these were the stupidest processes he has ever experienced and figured they were more the result of a consultant group selling the process as the way to get the best people, and these companies like too many these days buy in to the b.s..
He has the opinion that smart is easy to check out, and the ability to work with others can be found out pretty quickly with background checks and there is no need for the endless interviews and challenges.
Of course where he works they get to do more in depth background checks if you want to work there than a private company can probably get away with, but I'm not sure he grasps that part of the equation.
 
You must log in or register to reply here.
Back
Top