Also pushes the new business-friendly narrative that it's good for companies to pay their engineers less.

Well done, I hope the author's boss is proud.

/s

Ughh... reading this is ridiculous. Do we think bankers' salaries prevent them from staying at banks? Doesn't the first sentence answer the question of what really happened?

     > For the past year, Google's car project has been
     > a talent sieve, thanks to leadership changes,
     > strategy doubts, new startup dreams and rivals
     > luring self-driving technology experts. 

The people who left didn't STOP working on self driving cars. So they still love the work and are engaged by the dream of doing that. So theoretically - if Google had adapted to their needs & wants they could have retained the employees. Instead poor leadership drove them out the door.

I think a more typical response would be, "Hmmm. That wasn't so hard, let's see if they do it again next year." And when they don't -- that is typically when you walk.

That's anecdata. Not everyone is the same. I know friends who would quit in a heartbeat and pursue other passions (music, teaching, non-profit work, etc.) if they got F-you money.

I can pretty safely say a single $5M bonus would amount to an instant lifestyle change (namely less pressure to work) for me.

Two things to note: 1) this is different than winning a lottery, because with a lottery there is no expectation of a change in wages; 2) your first week of leisure is worth much, much more than your 50th, as is your 1st dollar worth more than your 5 millionth. There is some level at which you stop working. I think it's much higher than you realize.

I think I would keep working.

The important factor here isn't the money.

On Wall St, there's always a bigger bonus to aim for at the same firm next year. At Google, if you wanted to take your big bonus and turn it into more millions, you'd better try starting a new business.

2. If you work at Waymo you'll quit only if you're unhappy

1: http://paulgraham.com/submarine.html

It's hard to think of something Paul Graham wrote that has done more damage to discussions on HN than his piece about submarine PR pieces.

When people thought every published article was some propagandist's attempt to sway public opinion, they were wrong.

But the second view is a lot less wrong than the former.[1]

"Journalism is printing what someone else does not want printed: everything else is public relations." -- George Orwell.

[1] Didn't feel like following Asimov's pattern for the third line.

I think that your comment might be the damaging one because it attempts to delegitimize the opinion of another community member without any reasons given.

The usual path from "I don't like X" to "X is objectively false" to "no one would ever X in good faith" to "How much did they pay you for X, asshole" (not that you said that, but it's the flamewar end-state) never actually consults with reality. Since substantive discussion depends on looking beyond our pre-existing opinions, this is a problem for substantive discussion. Worse, it feels like being a champion of truth, so a lot of intense energy flows into it, so flamey passion goes up as substance goes down—a perilous combo.

I don't know what a full solution would look like, but at a minimum, HN comments on such issues—if they're to be good HN comments—need to show some sign of breaking out of that self-referential cycle. That's why we tell users that they're not allowed to accuse others of astroturfing or shillage without evidence, and always add "an opposing view is not evidence".

I don't know if this is a submarine article--it's impossible to know for sure, of course. It really, really reads like one though, and if pointing that out is unacceptable, then feel free to detach my comment and mark it as off-topic. I won't be offended.

By investing in this company you are "betting" that they will be able to generate "alpha" ( in the financial sense )

http://www.investopedia.com/terms/a/alpha.asp

Now it just comes across as tacky.

We also like that it means alpha-bet (Alpha is investment return above benchmark), which we strive for!

https://googleblog.blogspot.com/2015/08/google-alphabet.html

http://hooli.xyz/

If anything, the exodus at google shows clear lack of vision and direction from the management and executive teams. Engineers care about achieving something they can feel proud of and if the project lacks that, they move on.

This article is written in a way that suggests engineers should be paid less so that they remain loyal...you know, like slaves.

1. The highest paid executives in other industries tend to be at the apex of their careers, and new startups would be a status step down for them.

1. Empirically, executives do switch companies somewhat routinely.

Your last sentence presents a false dichotomy: either we're meant to reject the notion that flight money is a real phenomenon, or we're required to believe engineers shouldn't be compensated above slave levels. There's a huge excluded middle in that argument.

It's possible I've not seen them, but seems curiously absent in the discourse to me.

These are positive statements, not normative ones.

> new startups would be a status step down for them.

No, but highly paid executives can always go to a higher status company; similarly, if you were status-motivated but also a well placed mid career goog eng…

it's not clear to me you _needed_ a sweet bonus in order to switch to your startup, ya know?

>executives do switch companies somewhat routinely.

I think the theory of what I'm trying to advance is, what is good for the goose is good for the gander.

I personally feel I would be rather loyal to someone paying me fuck you money, but then again I'm well into my own incredible journey, so it's possible I'm now blind to status incentives relative to working at google.

The premise behind flight risk money is that people might not be interested in optimizing their salary by constantly switching jobs, especially because they might really like their current job. But a lot of people have a short list of jobs that they would be interested in if money weren't a factor; many employees (maybe most) aren't in their absolute ideal job, but are balancing money and work life. When you give those people lots of additional money, you change that balance.

This isn't complicated, right? A pretty large group of people (1) like their current job, and (2) would like to start or join a new company.

That's just not the same as what happens with executives. Executives are not waiting for their big break so that they can start a new thing.

To be charitable to you, maybe I'm just really underestimating how many people are chomping at the bit to gtfo;

having been self employed for "forever", it's hard for me to empathize: mid career A player at the goog probably could've made the switch much earlier, if leading a startup was what was in their heart of hearts.

Tesla's 'master plan' is pretty easy to understand externally and gives a clear path/vision forward. How Google plans to actually use their autonomous car tech is unclear from the outside (and I suspect also unclear on the inside).

Tesla's focus on manufacturing and using already sold cars to train their software was a really good plan. Google can't really compete with that even with their fleet of waymo mini cars - it also leaves the big problem of manufacturing cars unsolved (partnership? licensing?).

This seems especially tricky when any car manufacturer that plans to exist in a couple years is already building out their own software divisions with a focus on autonomy.

I'd suspect the engineers see that the management plan isn't going to win so they'd rather jump ship to someone that's doing a better job of it.

This seems to be a recurring trend with Google - get in early with something good, but then let others eventually catch up and beat you.

Say, for instance, that the best three self-driving technologies in the world, at the time consumer adoption starts, are Tesla, Google and Volvo. Every other manufacturer is going to rush to dump their own half-developed version and just licence Google's.

At the very least, Google should be able to make money by selling geographic data to all the different manufacturers. But I imagine you'll see Google self-driving tech all over the place.

Developers are creators, if they leave, they tend to leave because their dreams are getting squashed and the opportunity window is getting closed in an area they would love to innovate. If you are at google and don't see google releasing a self driving car, you will miss the window unless you jump ship to Tesla or GM or Audi. If you are into VR, maybe you would wanna jump ship to Oculus, the window to be part of history is very small.

Furthermore, you're charitably generalizing developers as people who leave because of their dreams and uncharitably generalizing executives as people who just get bored and want to turn the wheels of production at a whim.

I think the point that the other commenter and I are trying to make is what while you can defensibly claim you're not saying executives "aren't critical players", you're being dismissive of their overall contributions by painting them in a diminutive and uncharitable way.

You're essentially saying, "I'm not saying executives aren't critical players, they're just not the real creators of value...they leave because they get bored and have so much capital while developers leave because of a twinkle in their eye."

GP's opinion that management aren't 'creators' of course neglects a few things:

a) the role of management is necessarily one step removed from direct project output, so yes, they'll probably do less direct work on the project

b) that 'creators' can (and do) move in to management

c) there is no creativity involved in taking uncertain inputs and turning those into the direction & vision of a project

d) that shaping a project's output to maximize outcomes is something that can net a lot more value than adding another 'creator'

e) that 'creators' sometimes don't care about the project outcomes as much as the process

f) that 'creators' may not have experience in the various domains their product or service is ultimately being provided in, and thus may not understand or pick the right constraints to apply during their creative process.

...among many other possible misconceptions.

I've seen this in the education world myself, where teachers (who are similarly front-line workers) were skeptical of management and the viewpoints of those outside their domain being able to add value. 'But they're not educators' was the statement I heard a lot.

Why, just yesterday I made some suggestions that resulted in a 20% improvement in one of our API calls, and I didn't code a single line.

You underestimate the value of executives by not valuing their experience and ability to direct.

I imagine you'd rather live in a world ruled by programmer-kings who are the only people capable of creating value, while all the other peons toil away in the service of their creation with their inferior, bogus skillsets.

I say it's an important point because it was a big one for my own growth. I think many programmers (and many HN users) tend to feel that way, and not without reason. What's needed is patient explanation, not oppositional dismissiveness. Unfortunately the latter takes a lot less time.

1. Creating and Maintaining a vision for the future of the company based on industry experience, research and hypothesis.

2. Keeping all of the teams and players up to sync on that vision.

3. Driving the development of that vision from day to day and pivoting when needed.

So while they might not be "creating" code, it is important to note that code alone doesn't make a company and leadership is essential especially the larger a corp becomes.

And at the end of the day, what's going to make or break your business is how much product you're moving. Execs are hugely important for this, in addition to all the points you mentioned.

Sure, we roll our eyes at the nonsense language executives spout, and at the political games they play. And yet good executives create functional companies with a good corporate culture, where good engineering can take place. You might ask yourself: if everything is running smoothly, what do we need the executives for? Much like the IT guy who doesn't get appreciated for preventing disasters, the work executives do is invisible when done well.

Besides, the line between creation and facilitation is a blurry one. Without salespeople there is no budget for programmers to create things. Does that mean salespeople as non-creators are deserving of our disdain? Or should we just appreciate them for fullfilling a vital role within the company? Suppose that executives were non-creators (and they're not), why does it matter? They're nonetheless vital for the company's success.

the stop of that large money is what seems to have happened:

"In December, the car unit morphed into a standalone business called Waymo, and the system was replaced with a more uniform pay structure that treats all employees the same"

>Engineers care about achieving something they can feel proud of and if the project lacks that, they move on.

they achieved "milestones", got handsomely paid for that, after that the payment structure changed back to regular with no such payouts in the future, and the real achievement - fully autonomous car - is still long in the future with a lot of grunt unglorious work to be performed in order to reach that future ... While other companies invite you to make the same glorious "milestones" for even bigger money - sounds like no-brainer.

Google wanted that division to be like a startup - they seems to have got it right to all the details, including that exodus after cash-out on an "exit" with product barely reaching prototype stage and powerpoint of trillion dollar business. An expensive at that - 16x "valuations" growth applied to 4 years of an engineer's bonus and equity - i.e. something like 16 x $1M = $16M per engineer (nicely matches market going rate for AI and other hot aquihires. man! i should have made a career out of that GC hobby back in 2004-5 :) - i'd not be surprised if Ruth had heart attack upon seeing those numbers :)

Isn't that the argument behind Daniel Pink's book?

In retrospect, that's not at all what management was telling me. But I felt it. And I didn't want to go through another year like the one that led up to the bonus. And I didn't feel like climbing a career ladder (the way Microsoft at the time seemed to encourage). I just wanted to be a simple software engineer and build cool stuff.

I guess what I'm trying to say is anecdotally, actions (including big salaries / bonuses) can have unintended consequences.

Take IBM's recent initiative to kill remote work. Well, a predictable unintended consequence of that is a brain-drain. Those who aren't capable of finding other jobs will stay at IBM. Those who are capable will leave for greener (and remote-friendly) pastures.

Many top executives are on this track as well, which explains some of the crazed pay scales.

I chock this statement up alongside "A company exists solely to create profit". It has this raw, naked capitalist "that's the way the world works" jingoism that doesn't reflect that what motivates people so often entails much more than that.

Of course, in the age of welfare, its typically better to have your employees be slaves to the state than slaves to you so you dont have to pay to feed, clothe, and house them.

But for businesses where the market rate of their employees is above the wage to keep that person alive and housed, then slaves are a very attractive thing to have. Many people (~millions) have no moral opposition to slavery.

I would expect most executives like the power of being executives at established companies and managing lots of people, and they probably have fewer personal projects that they want to do. Maybe I'm wrong; I'm not an executive.

1. the regulatory environment in the US was relaxed

2. the cars were a big hit, and

3. there are articles about how far sighted Google was for letting the research project have these startup like returns.

Have onerous regulations been a problem for robocar development? If they were, I sort of think I would have heard of it, particularly here on HN where such vehicles are a common topic of conversation.

[1]https://www.wsj.com/articles/googles-self-driving-cars-hit-r... [2]https://www.texastribune.org/2013/02/21/googles-self-driving... [3]http://time.com/4152958/google-california-self-driving-car-r... [4]http://www.reuters.com/article/us-google-selfdrivingcar-idUS...

I didn't say they should not be relaxed, by the way.

I do think at some point there will be a big public fight about what we are willing to accept.

I have heard it said that even current self-driving tech is safer than humans. But I haven't looked at the stats myself.

Deployment is something else. California's rules for autonomous vehicle deployment are out.[1] There's a third party testing requirement and a three-year transition period during which autonomous vehicles can only be leased, not sold outright, and drivers require an autonomous vehicle certification. That's about what Volvo is doing with their test in Sweden.[2]

Nobody has technology good enough yet to turn loose on the general public. Google/Alphabet/Waymo seems to be getting close, but everybody else is at least two orders of magnitude worse on disconnects per mile.

[1] https://www.dmv.ca.gov/portal/wcm/connect/dbcf0f21-4085-47a1... [2] http://blog.caranddriver.com/meet-the-first-real-family-slat...

That's not to say the regulators were obstructionist or anything - but these things can be difficult even when there aren't any obstructions.

Multiple big players want to get into autonomous cars (and other vehicles -- this is the same space as military contracts for drones) and Google's executives were in total disarray. You'd think they'd lose out but instead they successfully hit some giant milestones by throwing enough cash at the problem.

I mean, if the staffers had been paid poorly, wouldn't they have quit without waiting for that 16x bonus that only comes after meeting the division goals and then waiting four years?

Lack of vision, lack of direction, failure to set expectations or execute a plan, better opportunities elsewhere, poor management... those are all reasons for brain drain.

Gave them too much money? That's absurd.

Honestly, it looks more like a propoganda message to pre-explain shrinking bonus money.

Might not be practical for a number of reasons. Just sounds interesting, without having put any thought into the idea beyond the time it took me to write this comment.

This is interesting because they were helping build the future. It's easy to see why people might not be excited or passionate about the next CRUD app but self driving cars seems like a passion project to me.

Bear in mind, many of them are still working on self-driving cars, they just went to work for other companies. If as you say, it's a passion project, they are going to go to whoever they think is best equipped to accomplish the vision. Especially since money is no longer an issue for them.

The main reason its a sieve is that they realized they could make more money by leaving. They saw examples with Otto and others.

A more accurate phrasing might be "One reason staffers quit google's car project? The company structured their pay so that an intermediate project milestone was treated like an IPO event."

It's not that the employees got rich so they no longer cared about money, it's that they basically got their maximum bonus from Google, so they were better off taking their talents elsewhere to get the next big payout.

Given that their CFO discussed these payouts on the earnings call of a company with $6.6 billion in operating expenses, it sounds like these were entering the $100million range. Makes sense given that Google was trying to recruit and motivate people to work on startup projects, while overall Google market cap wasn't going to grow exponentially the way that a 2011 facebook or uber might. It probably wasn't a bad idea, they just executed in a way that only incentivized employees to get partway done.

After hitting all those milestones and not seeing a car on the road, I can imagine that as a reason to be disgruntled and wanting to move to another company.

This just in, startup type incentives does work and also produces startup like effects including bouncing around more than they would like.

It may be been banal. Now that the visionary part of the program has been executed, and with development more or less on rails, it's possible there simply isn't anything exciting left for the top guys to work on. That and they're all extremely valuable to car companies who would rather not get left in the dust.