7 ms·

My theory is that while work is dull, people are not overqualified. They need smart people not because logic they write is complex, but because navigating total

My theory is that while work is dull, people are not overqualified. They need smart people not because logic they write is complex, but because navigating total chaos of internal system complexity needs smartness. If they could refactor and get internal systems in order they could lower qualification bar, but that refactoring requires super duper overqualified people, etc.

titanomachy 5y ago
I think this might be about right. Google's internal tools were built by very smart engineers for use by other very smart engineers. (I'm thinking especially of the Producers Framework as I write this, but there are other examples.) I've seen plenty of intelligent people struggle to get their heads around Google's abstractions. The tools are powerful enough, though, that once you get it you can pretty easily build stuff that scales to ridiculous levels. It could be that better design will make these kinds of tools more accessible in the future.
- onekorg 5y ago
  Producers Framework is awesome. Took me a while to wrap my head around writing code that way. Is there anything open source that is similar to it?
  - titanomachy 5y ago
    Not that I know of. I don’t think it’s user-friendly enough to succeed as an open-source tool.
notJim 5y ago
I think that's about right, and the "mess" isn't just code, but organizational. You might be hired to maintain some simple widget, and then 1–2 years later be writing a large scale job processing service, or designing a new front-end rendering system, or any other number of high-impact, challenging projects. Theoretically, these could be different sets of people (like techs vs engineers or something), but since companies are constantly re-orging and don't know what they need, it's the same people doing both. I have to say however, that for engineers this is a massively good thing, both in terms of being able to work on lots of different stuff, and the compensation implications.
bigmutant 5y ago
FAANGs/large companies in general have a ton of accidental complexity. A lot of the work devolves into "make this work in our complicated environment". Its just more pronounced at FAANG due to the scale.
oreally 5y ago
> They need smart people not because logic they write is complex, but because navigating total chaos of internal system complexity needs smartness. Careful, that's the FANG SWE ego speaking. What you describe isn't smartness but a willingness and patience to put up with lots of bloat.
- sanxiyn 5y ago
  I mean, yes, it's not only intelligence, but also willingness to put up with bullshit, but "isn't smartness" is going too far. It IS taxing on brain.
- bawolff 5y ago
  Much "smartness" is a willingness to work hard, being able to read the room and being willing to put up with a lot of BS. If you think FAANG is bad, just look at academia.
- summerlight 5y ago
  Many people (including this comment) severely underestimate the sheer volume of information processed everyday by those staff level engineers in big techs. Imagine that each day you need to review 1~2 design docs, read and reply tens of mails, attend at least 3 meetings and make non-negligible engineering decisions each, keep the lights on, manage and mentor some junior engineers. And write some code and design docs for your own projects upon all the crazy stacks of technical and procedural complexity. Time is limited resource, so you cannot really have enough time to understand everything needed as scopes expand. You will eventually need to embrace ambiguities and uncertainties stemming from the complexity. And this needs more than just willingness and patience. Perhaps having some level of smartness can help, but this alone might not be sufficient though.
- blaser-waffle 5y ago
  I'm not at a FAANG but am at a F500. We're paid well because we have to shift gears a lot, learn a lot, fake-it-till-you-make it, and put up with oodles and oodles of BS. From a technical standpoint a 16 year old could do my job. We're getting paid for patience and the ability to "code switch" in the socio-linguistic sense.
wolfretcrap 5y ago
My theory is FANGS pay more because it's a casino like setup. Yes, their employees are making lot of money but where her money ends up going? A large part of that ends up paying rent in a region where there are many others making similar wages like you, and trying to outcompete you out of the best property you can choose to live. Mostly people at headquarters or major offices get paid large salaries. Executives make decisions where to open office, they've upphand on the info and have access to money, they buy the land in and around the region where office will be setup, where their employees will want to live. That way half the money they pay ends right back into pocket of executives while FANG employees have statisfaction of making highest engineering wages. It doesn't matter how much you make, if you still find yourself doing the same thing you were doing 5 years ago, perhaps, all that money didn't bring you much. Now, there are exceptions to this where people saved up their salary and launched another business into huge success, yea those benefit the most. But most of the FANG employees will find themselves doing similar job after 5 years, so... How exactly that extra money has changed your life? Are you driving Buggati or Bentley to your office? Do you have celebrities (male/female -depending on your orientation) caressing your hair while you hack on the code? Did you start a school to teach underprivileged kids? As we can see most of these luxuries are still not available to majority FAANG employees regardless of how much they make. It's the money on paper, it doesn't buy much power. To have real power in society, having more money than your peers is not necessary - what you need is vastly more money and influence over other people.
- comp_throw7 5y ago
  I mean, FAANG compensation lets you trivially retire (or achieve financial independence) by ~40 (if you're single), and that's assuming you're budgeting for 200k/year with an extremely conservative withdrawal rate (3%). If you're willing to resettle somewhere cheaper and live a somewhat less extravagant lifestyle, you could do it by your early 30s no problem. And, again, this is if you're single, i.e. no second income, paying a single-filer's tax rate, and no benefit of cheaper housing by having someone to split the expense with, etc. A partner earning any real amount of money speeds this up dramatically. (Kids do cost money, if you have them, but they don't cost enough money to offset having a second SWE income, even if you pay for extremely expensive daycare, unless you have like five.) I don't know about you, but having "fuck you" money after a decade in the workforce, or "no, really, fuck you" money after 2 decades is a pretty good deal. It is, in fact, much easier to make this happen working at a FAANG (or comparable) than elsewhere. People like to complain about the cost of living, and yes, those complaints are valid, but working at a FAANG it's entirely possible to save more after taxes and expenses than software engineers in other parts of the country make _before_ taxes; trying to equivocate between those situations is really quite silly. The first is in a dramatically better financial position than the second.
fn1 5y ago
> but that refactoring requires super duper overqualified people, etc Yes, and it needs super duper overqualified managers to understand that the super duper overqualified engineers actually are on to something with their suggestions.

titanomachy 5y ago

I think this might be about right. Google's internal tools were built by very smart engineers for use by other very smart engineers. (I'm thinking especially of the Producers Framework as I write this, but there are other examples.) I've seen plenty of intelligent people struggle to get their heads around Google's abstractions. The tools are powerful enough, though, that once you get it you can pretty easily build stuff that scales to ridiculous levels. It could be that better design will make these kinds of tools more accessible in the future.

onekorg 5y ago
Producers Framework is awesome. Took me a while to wrap my head around writing code that way. Is there anything open source that is similar to it?
- titanomachy 5y ago
  Not that I know of. I don’t think it’s user-friendly enough to succeed as an open-source tool.

notJim 5y ago

I think that's about right, and the "mess" isn't just code, but organizational. You might be hired to maintain some simple widget, and then 1–2 years later be writing a large scale job processing service, or designing a new front-end rendering system, or any other number of high-impact, challenging projects. Theoretically, these could be different sets of people (like techs vs engineers or something), but since companies are constantly re-orging and don't know what they need, it's the same people doing both. I have to say however, that for engineers this is a massively good thing, both in terms of being able to work on lots of different stuff, and the compensation implications.

bigmutant 5y ago

FAANGs/large companies in general have a ton of accidental complexity. A lot of the work devolves into "make this work in our complicated environment". Its just more pronounced at FAANG due to the scale.

oreally 5y ago

> They need smart people not because logic they write is complex, but because navigating total chaos of internal system complexity needs smartness. Careful, that's the FANG SWE ego speaking. What you describe isn't smartness but a willingness and patience to put up with lots of bloat.

sanxiyn 5y ago
I mean, yes, it's not only intelligence, but also willingness to put up with bullshit, but "isn't smartness" is going too far. It IS taxing on brain.
bawolff 5y ago
Much "smartness" is a willingness to work hard, being able to read the room and being willing to put up with a lot of BS. If you think FAANG is bad, just look at academia.
summerlight 5y ago
Many people (including this comment) severely underestimate the sheer volume of information processed everyday by those staff level engineers in big techs. Imagine that each day you need to review 1~2 design docs, read and reply tens of mails, attend at least 3 meetings and make non-negligible engineering decisions each, keep the lights on, manage and mentor some junior engineers. And write some code and design docs for your own projects upon all the crazy stacks of technical and procedural complexity. Time is limited resource, so you cannot really have enough time to understand everything needed as scopes expand. You will eventually need to embrace ambiguities and uncertainties stemming from the complexity. And this needs more than just willingness and patience. Perhaps having some level of smartness can help, but this alone might not be sufficient though.
blaser-waffle 5y ago
I'm not at a FAANG but am at a F500. We're paid well because we have to shift gears a lot, learn a lot, fake-it-till-you-make it, and put up with oodles and oodles of BS. From a technical standpoint a 16 year old could do my job. We're getting paid for patience and the ability to "code switch" in the socio-linguistic sense.

wolfretcrap 5y ago

My theory is FANGS pay more because it's a casino like setup. Yes, their employees are making lot of money but where her money ends up going? A large part of that ends up paying rent in a region where there are many others making similar wages like you, and trying to outcompete you out of the best property you can choose to live. Mostly people at headquarters or major offices get paid large salaries. Executives make decisions where to open office, they've upphand on the info and have access to money, they buy the land in and around the region where office will be setup, where their employees will want to live. That way half the money they pay ends right back into pocket of executives while FANG employees have statisfaction of making highest engineering wages. It doesn't matter how much you make, if you still find yourself doing the same thing you were doing 5 years ago, perhaps, all that money didn't bring you much. Now, there are exceptions to this where people saved up their salary and launched another business into huge success, yea those benefit the most. But most of the FANG employees will find themselves doing similar job after 5 years, so... How exactly that extra money has changed your life? Are you driving Buggati or Bentley to your office? Do you have celebrities (male/female -depending on your orientation) caressing your hair while you hack on the code? Did you start a school to teach underprivileged kids? As we can see most of these luxuries are still not available to majority FAANG employees regardless of how much they make. It's the money on paper, it doesn't buy much power. To have real power in society, having more money than your peers is not necessary - what you need is vastly more money and influence over other people.

comp_throw7 5y ago
I mean, FAANG compensation lets you trivially retire (or achieve financial independence) by ~40 (if you're single), and that's assuming you're budgeting for 200k/year with an extremely conservative withdrawal rate (3%). If you're willing to resettle somewhere cheaper and live a somewhat less extravagant lifestyle, you could do it by your early 30s no problem. And, again, this is if you're single, i.e. no second income, paying a single-filer's tax rate, and no benefit of cheaper housing by having someone to split the expense with, etc. A partner earning any real amount of money speeds this up dramatically. (Kids do cost money, if you have them, but they don't cost enough money to offset having a second SWE income, even if you pay for extremely expensive daycare, unless you have like five.) I don't know about you, but having "fuck you" money after a decade in the workforce, or "no, really, fuck you" money after 2 decades is a pretty good deal. It is, in fact, much easier to make this happen working at a FAANG (or comparable) than elsewhere. People like to complain about the cost of living, and yes, those complaints are valid, but working at a FAANG it's entirely possible to save more after taxes and expenses than software engineers in other parts of the country make _before_ taxes; trying to equivocate between those situations is really quite silly. The first is in a dramatically better financial position than the second.

fn1 5y ago

> but that refactoring requires super duper overqualified people, etc Yes, and it needs super duper overqualified managers to understand that the super duper overqualified engineers actually are on to something with their suggestions.