5 ms·

IMHO the software engineering community has Stockholm’s syndrome around compensation (cue the doofy but not inaccurate jock/nerd dichotomy stereotype of not eve

IMHO the software engineering community has Stockholm’s syndrome around compensation (cue the doofy but not inaccurate jock/nerd dichotomy stereotype of not even 10 or 20 years ago). This shit is hard. Even relatively intermediate SWE jobs demand a level of fluid intelligence and capacity for detail within 1-2 sigma of the hardest jobs on Earth, and the more elite software jobs are among the most demanding jobs on Earth. This trend is accelerating as software “eats the world” or whatever we’re calling it now. Why shouldn’t we make a fuckload of money?

4gotunameagain 4y ago
> and the more elite software jobs are among the most demanding jobs on Earth. You need to get out of that bubble. Even being a professional violinist is like an order of magnitude harder than the most hardcore software job. And most of them are payed shit.
- benreesman 4y ago
  I have tremendous respect for serious concert musicians. I was never a professional musician but I was an extremely serious amateur musician for over a decade and my life is full of people who could trivially be professionals if, to your point, it paid better. But I find the contention that elite musicians are “an order of magnitude” more capable as concerns fluid intelligence and capacity for detail than the folks at (to name one example) DeepMind to be just silly. Now the guy composing those pieces after going deaf, yeah, that’s in this whole other world called “genius”. Playing violin well enough to be called a profession is an extremely demanding task both physically and intellectually, that’s been measured every which way, but let’s not get carried away.
  - Aeolun 4y ago
    > Now the guy composing those pieces after going deaf, yeah, that’s in this whole other world called “genius”. Isn’t this essentially mathematics? Not that that’s not still impressive :)
- wiseowise 4y ago
  > Even being a professional violinist is like an order of magnitude harder than the most hardcore software job. And you got to that conclusion how?
plonk 4y ago
> the more elite software jobs are among the most demanding jobs on Earth Programming is ridiculously easy compared to physics or math. The hardest CS problems are math, not engineering.
- BlargMcLarg 4y ago
  Not every programming job skips physics or math. If anything, the fact engine developers generally earn less or equal for a far harsher job than webdev proves the ridiculousness GP illustrates. People really need to get out of this webdev bubble.
  - plonk 4y ago
    The math involved in a game engine are miles away from actual math research or theoretical physics. > People really need to get out of this webdev bubble. I think you're also a victim of that bubble. If your point of reference is webdev, anything that involves basic algebra will look hard as hell, that's not a useful comparison.
SergeAx 4y ago
There is a funny conclusion from Pareto principle: in any arbitrary division 80% of people will think that they belong to 20%. I am doing tech interviews for I don't remember how long, and most of candidates not only are within 1-2 sigmas, but don't even have an idea what sigma is. But they still want to be compensated by an upper edge, and will even make a drama now and then. The last example was a candidate on a coding session, who almost failed first (warming, 5min tops) problem, was unable to cope with second problem and left the interview telling us that "your tasks are not related to real life". The twist is that second task was not only taken from our real life cases, but even has a real-life data collected from server logs.
dan-robertson 4y ago
To answer your last question: how much you’re paid has little to do with how hard you tried or how hard your job is. Secondly your vague statement about ‘within 1-2 sigma of the hardest jobs on Earth’ seems a little hard to interpret to me. If you say the hardest jobs are at 5 sigma (~2k people on earth have them) then 4 sigma gives 250k people and 3 sigma gives about 10 million people. (In the US the numbers give you 85 people, 9.5k people, and 400k people. 2 sigma gives 7 million people, about 2% of the population). So maybe if you say the hardest jobs in the US sit around a Z-score of 4 and programmers at high-paying companies sit around 2 your numbers make sense. But programmers at well paying companies being in the 98th %ile of income sounds plausible to me. It seems programmers are getting a reasonably good deal in that their income percentile is roughly corresponding to their percentile in whatever metric you’re thinking of (if their Z score is about 2). Other professions don’t seem to have this quality. (One assumption I’m making is that not everyone with Z>2 will be a programmer at some well-paying company).
- benreesman 4y ago
  Obviously to fit my point into what amounts to a tweet I’ve hand-waved a bit, and your less hand-wavy napkin math sounds more or less plausible. 5 standard deviations sounds like a lot for anything that can be called “jobs” plural: being an Olympic gold-medalist is a “job” in some sense but I think the idea of job has started to break down a bit there. Ditto Fields-caliber mathematicians, etc. I don’t think I’ve ever heard anyone seem all that surprised that senior but not world-famous attorneys often bring in 7 figures, but I routinely see people act not just surprised but almost indignant that senior software people sometimes do. I don’t think we strongly disagree as long as I’m clear that software pay seems to be growing roughly in line with the importance of software to life and commerce. My point was, why should anyone be surprised or even upset by this?
ricardobayes 4y ago
I don't think this true at all. Your intermediate SWE job will be adding functionality to a React page or writing REST endpoints in Python. You will reach this level with around 2 years of experience. I don't see this nearly as demanding as the 10+ year preparation to be a neurosurgeon, or to be a teacher. I think being an 'average' dev is pretty easy after all in the grand scheme of things.

4gotunameagain 4y ago

> and the more elite software jobs are among the most demanding jobs on Earth. You need to get out of that bubble. Even being a professional violinist is like an order of magnitude harder than the most hardcore software job. And most of them are payed shit.

benreesman 4y ago
I have tremendous respect for serious concert musicians. I was never a professional musician but I was an extremely serious amateur musician for over a decade and my life is full of people who could trivially be professionals if, to your point, it paid better. But I find the contention that elite musicians are “an order of magnitude” more capable as concerns fluid intelligence and capacity for detail than the folks at (to name one example) DeepMind to be just silly. Now the guy composing those pieces after going deaf, yeah, that’s in this whole other world called “genius”. Playing violin well enough to be called a profession is an extremely demanding task both physically and intellectually, that’s been measured every which way, but let’s not get carried away.
- Aeolun 4y ago
  > Now the guy composing those pieces after going deaf, yeah, that’s in this whole other world called “genius”. Isn’t this essentially mathematics? Not that that’s not still impressive :)
wiseowise 4y ago
> Even being a professional violinist is like an order of magnitude harder than the most hardcore software job. And you got to that conclusion how?

plonk 4y ago

> the more elite software jobs are among the most demanding jobs on Earth Programming is ridiculously easy compared to physics or math. The hardest CS problems are math, not engineering.

BlargMcLarg 4y ago
Not every programming job skips physics or math. If anything, the fact engine developers generally earn less or equal for a far harsher job than webdev proves the ridiculousness GP illustrates. People really need to get out of this webdev bubble.
- plonk 4y ago
  The math involved in a game engine are miles away from actual math research or theoretical physics. > People really need to get out of this webdev bubble. I think you're also a victim of that bubble. If your point of reference is webdev, anything that involves basic algebra will look hard as hell, that's not a useful comparison.

SergeAx 4y ago

There is a funny conclusion from Pareto principle: in any arbitrary division 80% of people will think that they belong to 20%. I am doing tech interviews for I don't remember how long, and most of candidates not only are within 1-2 sigmas, but don't even have an idea what sigma is. But they still want to be compensated by an upper edge, and will even make a drama now and then. The last example was a candidate on a coding session, who almost failed first (warming, 5min tops) problem, was unable to cope with second problem and left the interview telling us that "your tasks are not related to real life". The twist is that second task was not only taken from our real life cases, but even has a real-life data collected from server logs.

dan-robertson 4y ago

To answer your last question: how much you’re paid has little to do with how hard you tried or how hard your job is. Secondly your vague statement about ‘within 1-2 sigma of the hardest jobs on Earth’ seems a little hard to interpret to me. If you say the hardest jobs are at 5 sigma (~2k people on earth have them) then 4 sigma gives 250k people and 3 sigma gives about 10 million people. (In the US the numbers give you 85 people, 9.5k people, and 400k people. 2 sigma gives 7 million people, about 2% of the population). So maybe if you say the hardest jobs in the US sit around a Z-score of 4 and programmers at high-paying companies sit around 2 your numbers make sense. But programmers at well paying companies being in the 98th %ile of income sounds plausible to me. It seems programmers are getting a reasonably good deal in that their income percentile is roughly corresponding to their percentile in whatever metric you’re thinking of (if their Z score is about 2). Other professions don’t seem to have this quality. (One assumption I’m making is that not everyone with Z>2 will be a programmer at some well-paying company).

benreesman 4y ago
Obviously to fit my point into what amounts to a tweet I’ve hand-waved a bit, and your less hand-wavy napkin math sounds more or less plausible. 5 standard deviations sounds like a lot for anything that can be called “jobs” plural: being an Olympic gold-medalist is a “job” in some sense but I think the idea of job has started to break down a bit there. Ditto Fields-caliber mathematicians, etc. I don’t think I’ve ever heard anyone seem all that surprised that senior but not world-famous attorneys often bring in 7 figures, but I routinely see people act not just surprised but almost indignant that senior software people sometimes do. I don’t think we strongly disagree as long as I’m clear that software pay seems to be growing roughly in line with the importance of software to life and commerce. My point was, why should anyone be surprised or even upset by this?

ricardobayes 4y ago

I don't think this true at all. Your intermediate SWE job will be adding functionality to a React page or writing REST endpoints in Python. You will reach this level with around 2 years of experience. I don't see this nearly as demanding as the 10+ year preparation to be a neurosurgeon, or to be a teacher. I think being an 'average' dev is pretty easy after all in the grand scheme of things.