And in general, stock-based compensation significantly skews the numbers, such that they are rarely reflective of what people get in their pay-packets each month.
Stock-based compensation for a public company with a good track record of growth and stability is not much different than cash, it's just on a different timer than 2x/month. It should skew the numbers.
"I interviewed at Airbnb, Facebook, Lyft, Uber, and 3 other companies. I received offers from Facebook, Lyft, Uber, and 2 other companies. First year total compensation ranged from 380k~500k from both public and private companies, 2nd year comp was around 400k for most of them. The offer numbers correlated with my interview performance, and from my research, it seems like some companies offer standardized offers to every candidate. Surprisingly the private companies did not beat out the public companies in equity package.
Breakdown was around 180~210 base, 500~800 equity, and 0~100 sign-on." [emphasis added]
They all can exist without it, and they all benefit from it. I see your aim in singling out shopping, but let’s not pretend that other services don’t provide signal to influencing the outcomes on shopping or flight purchase or whatever.
Not really, professional sports salaries often work the same way- teams who have a superstar can't afford to have the star play for the competition even if they could win without the star playing for their team.
76ers signing Al Horford was probably as much about getting him off of the Celtics (because he's historically given Embiid problems in the playoffs) as it was about actually getting him.
Yeah this is one of the biggest urban legends in sports. It's just not a thing.
A team won't pay a player just to keep them from going to their division rival or whatnot, it's a nice fairy tail for the ESPN talking heads to keep repeating and it generates a nice bit of juicy controversy, but it just isn't a thing.
Besides it doesn't pass the smell test. If you don't value the player or you think your chances of winning are better by spending that money elsewhere, why would it be a net benefit to your team to keep a player you don't want to pay just to stop him from going to another team? Do you think the Dallas Cowboys are really thinking to themselves "Man, we don't really value RBs like that, but we really got to stop Zeke from going to the Eagles"? Nonsense!
Likewise if Google really doesn't want an employee going somewhere else to put a damper on their dominance, if it really was the goal of that employee to prevent Google's dominance, don't you think there are infinitely many other avenues that person could take? If you are paying for the knowledge, what is to stop this individual from creating a pseudonym and committing to an open source AI library on a separate laptop on their free time? The answer is nothing, and in fact that happens all the time, or the employee is busy creating their side business unbeknownst to the company, leaves, and waits out their non-compete.
It's a silly conspiracy to think smart people will willfully stand by getting paid to do busy work and not pick up on what's really going on, and it doesn't make sense.
> what is to stop this individual from creating a pseudonym and committing to an open source AI library on a separate laptop on their free time
For one, the fact that this would break their work contracts. And I'm pretty sure Google can afford better lawyers that the average dev (even a well paid one).
Not to mention, this is especially true for leagues that have a salary cap. So that removes the major North American sports from the picture.
It just doesn't make sense to waste your cap space on a player you don't actually want to play. If you sign someone, you want to integrate them into your gameplan or at least warm the bench. And if they're a benchwarmer, you weren't signing them just to prevent others from doing so.
Sports isn't an equation. The human elements by far outweigh the "on paper" strength of the team. Also you never know who is going to get injured or when, so teams will stack their roster. The only thing that stops the biggest teams from having all awesome players is financials (either salary cap/financial fair play, or simple economics)
Look at how the Patriots succeed, for example. On paper, they are nowhere near the best NFL team. But they contend for championships almost every year.
The only example I can think of recently is Kimi Raikkonen in F1. Ferrari paid him his usual salary after removing him from the first team (each team in F1 only has two seats, then sometimes there is a reserve and an additional test driver role, mostly for the simulator). The agreement was they would continue to pay him his regular F1 driver salary, as long as he didn't race for any other F1 team.
He spent a year doing rally, did a couple of NASCAR races, snowmobile, motocross, and came back the following year for Renault.
Funnily enough, a few years later he returned as a first team driver for Ferrari for a few additional years and now races for Alfa Romeo.
His main thing is he hates promotional appearances (stuff like commercials and autograph signings) and restrictions (he has a private motocross track and shop, and he refuses to race for any team that won't let him ride motorcycles and snowmobiles and whatever the F else he wants to do).
I don’t know much about football but it seems to me that if this is true, then the metrics being tracked on paper are simply subpar. Probably all sports are an equation. Moneyball beinf the obvious example of this. If certain teams consistently win, then it’s not just an equation, but one with a theoretically relatively low variance model.
Tell that to Billy Beane, general manager of the Oakland A's, who used statistical analysis and team interaction models (i.e. "equations") to turn propel the team into the top ranks with salary expenses of only 1/3th of the other top teams. The story was described in the book Moneyball and dramatized in the film of the same name.
This is a common misconception. Beane did use statistics, but the central point of Moneyball was how he found market inefficiencies and exploited them. Other teams were using statistics as well, but it was Beane who applied market principles.
That's basically what early Michael Lewis books were about: exploiting market inefficiencies.
(Agree with your larger point about sports, however.)
>Not really. Those salaries are there to keep these smart people from doing something world-changing
I think it is more due to the fact there is still a relatively few number of people with a decent amount of experience in this field. Didn't OpenAI, a couple of few years ago, poach one of Google's employees for something like a million in compensation over x years?
Edit: yeah:
>Tax filings by the Elon Musk-founded research lab OpenAI revealed that its best researchers are paid top dollar for their expertise. In 2016 alone, OpenAI paid Ilya Sutskever, its top researcher, an impressive salary of $1.9 million.
AI is expected to be the next revolutionary step, so with losses at Deepmind pilling up to 1billion pound [1], it is hard to imagine that Alphabet can simply tell their shareholders "We just want to park these smart people here so that they don't do great things elsewhere" and that some (or all) of the 'smart' minds are driven by profit maximization at all times.
> For Google, the less productive they are, the better.
So Google spends a tonne of resources developing Tensor Flow only to OSS it, commoditizing AI and making it more accessible than ever, gives away lots of $$$ on AI competitions encouraging more people to get into AI/ML, acquires Kaggle, gives away tonnes of public datasets for anyone to use, just so they can hire new AI/ML staff to pay them obscene amounts of money to do nothing?
How did this line of thought become prevalent here?
You’re underestimating how hard it is to grow the profit of a trillion dollar company by 20-30% in a year. It takes lot of innovation (both in improving its properties and ad targeting in Google’s case for example)
1% acceptance rate of new hires — they have to compensate those they hire for making it through the gauntlet, because if the process works, they’ve just hired the best of the best and the best of the best would qualify for some ridiculous salaries given their likely high opportunity costs (one would think a duo or trio of Google engineers would make for a compelling startup team potentially worth millions so paying high figure salaries as guaranteed income could be a good compromise).
1% acceptance rate of new hires — they have to compensate those they hire for making it through the gauntlet
But there are loads of companies, universities, whatever making that claim. If you get 100 applicants and 90 are just spamming their resumes everywhere or are recruiters doing it, 9 go elsewhere or stay put for whatever reason, and you get 1 then you get that 1% but it doesn’t really mean anything.
The median salary at google is 246,804. It has about 100k employees. That means 50% make more than that which is about 50k people. I guess they have 700 employee who make about 581k. It seems the are on the right side of the curve.
Yes, it does, median is the value defined as 50% of the set
The "median" is the "middle" value in the list of numbers. To find the median, your numbers have to be listed in numerical order from smallest to largest, so you may have to rewrite your list before you can find the median.
Yeah when you consider the total comp, 400k doesn't sound crazy if there is stock, sign-on bonus, retirement/pension/401k, expected bonus, insurance (health/dental/life/whatever isn't covered by the UK NHS), etc.
This might be news to people not familiar with high tech job salaries, likely the Dice audience. So as general news it’s quite valid. On HN I doubt anyone is surprised by this.
Dice is one of the oldest tech focused job boards. 99% of the jobs posted there are for tech contractors (business analysts, it admin, software developers) posted by staffing agencies or Indian outsourcers like Tata who need to hire contractors. Pay for these jobs are between $25-$100 hr.
It's unusual in the context of Britain, I’d say, usually there's quite a well defined pay scale, with grades of worker, skilled worker, management, senior management, etc. It's all very class based. The exception is for self-employed contractors, which can be higher. But in general the notion of non-management salaried workers getting paid £300k upwards would be very surprising for a lot of people.
Yes. Outside of the US, salaries for non-managers rarely break $100k. Switzerland is probably an exception, but that's just because the average salary in Switzerland is really high, as are living costs.
He said £300,000 not $300,000. That'd be about $364,000.
I think you'll find such salaries for individual contributors are surprising everywhere except the Bay Area. Britain isn't special in this regard - this sort of story is merely highlighting the fact that wages in the UK are catching up to the (extremely high) levels found in Mountain View.
Why do you think this is? I fully invite speculation. It's... interesting. If paying tech professionals more is the way to win, then it's surprising the market doesn't eventually move in this direction.
I can think of a few possibilities. One is that tech in the UK is dissimilar to Silicon Valley. Perhaps tech in UK industry is much more of a supporting role, valuable and important to do well, but not a differentiator that separates market winners from losers (unlike Silicon Valley, where you have existential wars going on between massive multi billion dollar corporations where hiring engineers can determine the winner who takes all). I'm kind of begging the question here, though (it isn't cause it isn't).
Another possibility is flat out collusion. This clearly happened in Silicon Valley, with the no poach wage collusion. Once that cartel broke up (partly due to external pressures from participants who refused to join the cartel, and partly from a large anti-trust action), tech salaries did skyrocket.
Pure, flat out cultural resistance can also play a big role (going back to my first point). Maybe class-based behavior patterns are just too ingrained. I think market oriented people often underestimate just how stubborn deeply ingrained structural behavior patterns can be. There's something to the claim that there are people who would rather rule in hell than serve in heaven, so to speak, and I suppose there could be a class of people who would rather ensure that the UK never develop a competitive tech industry than drop a rung in the job hierarchy (or even just share a spot on the perch). In this sense, tech in there UK can develop only as far as possible within the boundaries of certain tech workers never being paid more a designated set of other classes of workers.
I don't think that's how these calculations should be done.
Let's assume that all of their estimates are correct, in most European countries the difference between what you cost an employee and what you take home is often more than a factor of two. Now I'm not quite sure about the UK, but I know about France and Italy and in both those cases if you spend 100 your employee takes home around 40.
If a few people making $500,000 come up with breakthrough ideas that lead to AGI, then Google would be able to easily recoup their entire DeepMind payroll many times over. Because, how much would that really be worth? $100 billion? $1 trillion?
Doesn't this include all the costs of an employee, like health benefits, insurance, and amortizing things like office space, HR and company services...
There's really only one interesting fact in this story for the HN crowd (who already know about tech salaries at the top), and it's this:
Pay is basically doubling every year ... Last year, the company earned £102 million (admittedly up from £54 million) a year earlier.
So both pay/headcount and revenue doubled.
Tech firms hiring isn't news. What shocked me here is that DeepMind makes £100 million pounds a year off anything. What are they doing, exactly? I thought they didn't have any products. Is this based on some kind of funny inter-firm accounting where the main Google HQ pays them IP licensing fees? Or do they have lines of business they aren't talking about much?
68 comments
[ 1.8 ms ] story [ 120 ms ] thread> Either way, the implication is that DeepMind pays an average of at least £478,— (U.S. $581,000) per head.
https://www.reddit.com/r/cscareerquestions/comments/cdp9yz/a...
"I interviewed at Airbnb, Facebook, Lyft, Uber, and 3 other companies. I received offers from Facebook, Lyft, Uber, and 2 other companies. First year total compensation ranged from 380k~500k from both public and private companies, 2nd year comp was around 400k for most of them. The offer numbers correlated with my interview performance, and from my research, it seems like some companies offer standardized offers to every candidate. Surprisingly the private companies did not beat out the public companies in equity package.
Breakdown was around 180~210 base, 500~800 equity, and 0~100 sign-on." [emphasis added]
Not really. Those salaries are there to keep these smart people from doing something world-changing that might contest Google's Internet domination.
For Google, the less productive they are, the better.
If I were google, I would keep investing in improving search, and that would obviously include hiring machine learning and AI experts.
A team won't pay a player just to keep them from going to their division rival or whatnot, it's a nice fairy tail for the ESPN talking heads to keep repeating and it generates a nice bit of juicy controversy, but it just isn't a thing.
Besides it doesn't pass the smell test. If you don't value the player or you think your chances of winning are better by spending that money elsewhere, why would it be a net benefit to your team to keep a player you don't want to pay just to stop him from going to another team? Do you think the Dallas Cowboys are really thinking to themselves "Man, we don't really value RBs like that, but we really got to stop Zeke from going to the Eagles"? Nonsense!
Likewise if Google really doesn't want an employee going somewhere else to put a damper on their dominance, if it really was the goal of that employee to prevent Google's dominance, don't you think there are infinitely many other avenues that person could take? If you are paying for the knowledge, what is to stop this individual from creating a pseudonym and committing to an open source AI library on a separate laptop on their free time? The answer is nothing, and in fact that happens all the time, or the employee is busy creating their side business unbeknownst to the company, leaves, and waits out their non-compete.
It's a silly conspiracy to think smart people will willfully stand by getting paid to do busy work and not pick up on what's really going on, and it doesn't make sense.
For one, the fact that this would break their work contracts. And I'm pretty sure Google can afford better lawyers that the average dev (even a well paid one).
It just doesn't make sense to waste your cap space on a player you don't actually want to play. If you sign someone, you want to integrate them into your gameplan or at least warm the bench. And if they're a benchwarmer, you weren't signing them just to prevent others from doing so.
Look at how the Patriots succeed, for example. On paper, they are nowhere near the best NFL team. But they contend for championships almost every year.
The only example I can think of recently is Kimi Raikkonen in F1. Ferrari paid him his usual salary after removing him from the first team (each team in F1 only has two seats, then sometimes there is a reserve and an additional test driver role, mostly for the simulator). The agreement was they would continue to pay him his regular F1 driver salary, as long as he didn't race for any other F1 team.
He spent a year doing rally, did a couple of NASCAR races, snowmobile, motocross, and came back the following year for Renault.
Funnily enough, a few years later he returned as a first team driver for Ferrari for a few additional years and now races for Alfa Romeo.
His main thing is he hates promotional appearances (stuff like commercials and autograph signings) and restrictions (he has a private motocross track and shop, and he refuses to race for any team that won't let him ride motorcycles and snowmobiles and whatever the F else he wants to do).
Tell that to Billy Beane, general manager of the Oakland A's, who used statistical analysis and team interaction models (i.e. "equations") to turn propel the team into the top ranks with salary expenses of only 1/3th of the other top teams. The story was described in the book Moneyball and dramatized in the film of the same name.
That's basically what early Michael Lewis books were about: exploiting market inefficiencies.
(Agree with your larger point about sports, however.)
Something something deflated balls cheating deal with satan something
I think it is more due to the fact there is still a relatively few number of people with a decent amount of experience in this field. Didn't OpenAI, a couple of few years ago, poach one of Google's employees for something like a million in compensation over x years?
Edit: yeah:
>Tax filings by the Elon Musk-founded research lab OpenAI revealed that its best researchers are paid top dollar for their expertise. In 2016 alone, OpenAI paid Ilya Sutskever, its top researcher, an impressive salary of $1.9 million.
https://www.teslarati.com/elon-musk-openai-paid-top-research...
[1] https://www.ft.com/content/1fc6db8a-b919-11e9-96bd-8e884d3ea...
So Google spends a tonne of resources developing Tensor Flow only to OSS it, commoditizing AI and making it more accessible than ever, gives away lots of $$$ on AI competitions encouraging more people to get into AI/ML, acquires Kaggle, gives away tonnes of public datasets for anyone to use, just so they can hire new AI/ML staff to pay them obscene amounts of money to do nothing?
How did this line of thought become prevalent here?
Pretty much. If somebody somewhere invents their own AI framework and uses it to build a better ad targeting system, then Google is toast.
This way the risk is taken care of. (Or at least it is from the management point of view.)
But there are loads of companies, universities, whatever making that claim. If you get 100 applicants and 90 are just spamming their resumes everywhere or are recruiters doing it, 9 go elsewhere or stay put for whatever reason, and you get 1 then you get that 1% but it doesn’t really mean anything.
>It has about 100k employees. That means 50% make more than that which is about 50k people.
No it doesn't, a typical business has far more people making minimum wage than it does making millions.
The "median" is the "middle" value in the list of numbers. To find the median, your numbers have to be listed in numerical order from smallest to largest, so you may have to rewrite your list before you can find the median.
I think you'll find such salaries for individual contributors are surprising everywhere except the Bay Area. Britain isn't special in this regard - this sort of story is merely highlighting the fact that wages in the UK are catching up to the (extremely high) levels found in Mountain View.
I can think of a few possibilities. One is that tech in the UK is dissimilar to Silicon Valley. Perhaps tech in UK industry is much more of a supporting role, valuable and important to do well, but not a differentiator that separates market winners from losers (unlike Silicon Valley, where you have existential wars going on between massive multi billion dollar corporations where hiring engineers can determine the winner who takes all). I'm kind of begging the question here, though (it isn't cause it isn't).
Another possibility is flat out collusion. This clearly happened in Silicon Valley, with the no poach wage collusion. Once that cartel broke up (partly due to external pressures from participants who refused to join the cartel, and partly from a large anti-trust action), tech salaries did skyrocket.
Pure, flat out cultural resistance can also play a big role (going back to my first point). Maybe class-based behavior patterns are just too ingrained. I think market oriented people often underestimate just how stubborn deeply ingrained structural behavior patterns can be. There's something to the claim that there are people who would rather rule in hell than serve in heaven, so to speak, and I suppose there could be a class of people who would rather ensure that the UK never develop a competitive tech industry than drop a rung in the job hierarchy (or even just share a spot on the perch). In this sense, tech in there UK can develop only as far as possible within the boundaries of certain tech workers never being paid more a designated set of other classes of workers.
Let's assume that all of their estimates are correct, in most European countries the difference between what you cost an employee and what you take home is often more than a factor of two. Now I'm not quite sure about the UK, but I know about France and Italy and in both those cases if you spend 100 your employee takes home around 40.
https://www.top500.org/news/ai-engineers-commanding-six-figu...
Pay is basically doubling every year ... Last year, the company earned £102 million (admittedly up from £54 million) a year earlier.
So both pay/headcount and revenue doubled.
Tech firms hiring isn't news. What shocked me here is that DeepMind makes £100 million pounds a year off anything. What are they doing, exactly? I thought they didn't have any products. Is this based on some kind of funny inter-firm accounting where the main Google HQ pays them IP licensing fees? Or do they have lines of business they aren't talking about much?
Anyone know?