While I basically agree with you (I also had an A GPA, so not sour grapes) what the Google guy actually said is that GPA's don't matter after you've been out of school after a few years. For recent grads they have some correlation. It certainly isn't perfect, but with recent grads its harder to pick the best people because they have little work experience to go by. Methinks that would be a good reason for the old summer internships that seem so rare these days - it gives you a chance to find the good new talent.
This approach provides the interviewer with a lot more insight than casual chats, including helping you to find those people who are really capable but aren't good conversationalists.
That sounds like a good approach, but how do you know that it produces good results? Gut feeling or even the fact that some people you hired worked out well is not sufficient, because you really have to compare your hiring decisions to the counterfactuals. What would have happened if you hired person A instead of person B? In the real world experiments are impractical, but you can correlate how well someone does at the company with how well they did with various interview approaches. To their credit, it seems like this is what Google is doing.
Also what's described in the NYT article seems closer to the specific type of chat approach that SnapperHead and I use (perhaps a little more structured) than what you describe:
you’re not giving someone a hypothetical, but you’re starting with a question like, “Give me an example of a time when you solved an analytically difficult problem.” The interesting thing about the behavioral interview is that when you ask somebody to speak to their own experience, and you drill into that, you get two kinds of information. One is you get to see how they actually interacted in a real-world situation, and the valuable “meta” information you get about the candidate is a sense of what they consider to be difficult.
They often waive their official GPA requirements if you worked in the field while earning your degree. 25-30 hours a week as a programmer while going to college full time and most corps won't care whether your GPA was 2.5 or 3.5 when applying for a development job.
Not a bad approach. Several years into my BS I switched from full-time student to full-time employment and part-time student. My grades went down, but I actually learned more in my classes because I saw the applications. It also cured me of the suspicion that classes only taught ivory tower nonsense.
I've never applied to Google because I'd heard enough about the interview process to realize it was mostly an unintentional way of asking if you're a recent graduate with a mind uncluttered by practical on-the-job knowledge, so you can focus on algorithms and brainteasers that have very few real world applications (and none in the job you're applying for.)
Try Netflix. According to a recent Slashdot post, they prefer hiring people who've spent a few years at Google learning their trade.
I find what Netflix does very interesting - effects of scale can be serious. They also take reliability very seriously, as some people deprived of a promised premiere can be dangerous:) As is so typical for the "ooh, ahh" evaluation of tech, little heed is paid to Netflix because they're selling movies, never mind that the tech is the magic behind the service. Yet Facebook get loads of "oohs" and "ahhs". The best tech is usually the tech that gets noticed least. Nobody thinks much of running the faucet, yet you're dealing with a tech that's been a key factor in civilizations since, uh, since there have been civilizations (and probably before).
Best not to rely on any one criterion. Personal projects are a positive indicator, but lack of them shouldn't be a show stopper. I've known some very good people, who are very interested in their work, who wouldn't have anything to do with the work when they're not on the job. Some of them even have lives (or so I've heard).
I have found the best indicator in the world is to just casually bullshit about technology. You can very quickly find someones strengths, weaknesses and if they are full of it. In a casual chat, people let their guard down and you get a look in.
That's the technique that works best for me when I interview people. Typically I'll ask them to pick something on their resume to chat about. I expect the interviewee not to be happy about discussing everything on their resume, because they all contain some some exaggerations (hell, you should add some because everybody does). However, if you can't come up with anything that was interesting and challenging, and that you're comfortable talking about, you're probably a fake. Some people are even shy about what they think is tooting their horn, and I encourage them to open up (others you have to shut up).
However, I'm also convinced there is no one magic formula for hiring people. Different techniques work well for different interviewers (and interviewees). The best approach is to have a candidate interviewed by a number of people w/ different approaches.
It's nice to see a large company try to objectively evaluate its hiring process and express some self doubt. All to often the hiring process at a company is assumed to be good because the company is successful, which is an obvious fallacy since many factors contribute to a company's success. In fact I wouldn't hire anyone who didn't immediately question such an assumption:)
All too often the hiring process at a company, or the admissions process at a university, is treated as though it were created with some magical special sauce, when in fact it does little more than reinforce some (often unstated) prejudices. It's especially troubling coming from organizations that supposedly value rational and scientific analysis.
You have to admit, for all the Facebook bashing that happens, the fact that hacks, break-ins, and bugs of this nature are so uncommon, given that they're dynamically managing a userbase of a billion people, is an impressive task.
I have to admit no such thing. First, there are a billion accounts, not a billion users (many users have multiple accounts), and many accounts are largely dormant. FB loves to hype their numbers. Second, there are hundreds of millions of bank accounts in the world, many of them now accessible online. Financial networks have been around since the 60's and have gotten much more sophisticated. While not perfect, they're incredibly more reliable than FB, otherwise we'd all be keeping money in mattresses. People go "ooh, ahh" over FB not because their technology is impressive, but because it's a cool new thing from Silicon Valley. Best renamed Silly Valley to keep up with its trends, it's gone from selling amazing tech (e.g. the first IC's) to selling amazing hype.
FB is not up 24/7. It sometimes goes down for hours at a time (second hand info as I don't use it myself).
Furthermore, and rather obviously if you understand that not every passing snark is meant to be completely literal, my point was that they don't do very thorough testing before going live. I have no idea why anyone would be impressed by most of FB's "technology". They're hardly so bleeding edge that they can be forgiven such flakiness as an inevitable part of new technology. As a contrasting example, financial networks have been around since the 1960's, and have kept getting more sophisticated. While not infallible, anybody who deployed code that lost $1B in transit would probably be shot rather than fired. It's FB's sort of flaky toys that makes many people think that software is an inherently unreliable scam.
Good idea. There's something interesting about Tor I didn't realize before reading the the Wikipedia article:
Originally sponsored by the U.S. Naval Research Laboratory... As of 2012, 80% of the Tor Project's $2M annual budget comes from the United States government, with the Swedish government
Yet the NSA takes Tor as a "definitely track this". Fact is stranger than fiction.
The urban legend is that Gore claimed he invented the Internet. What he actually said was poorly phrased, and typical of a politician, but no different from Eisenhower saying he took the initiative on creating the Interstate Highway system. Even Vint Cerf and Newt Gingrich have said the the urban legend is silly.
It's not that I'm a great Gore defender, or even that I mind people using the urban legend as a joke, but it's going too far for the aptly pseudonymous Hognoxious to use it as support for his resentment of Americans.
Nice anti-Wikipedia rant, but I already pointed out that the OP was simply mistaken and the article does mention Baby. Don't let that get in the way of a good tirade though. Speaking of facts, do you have any that demonstrate that Wikipedia is less accurate than highly regarded encyclopedias like Britannica?
Test cases? We're talking about Facebook - the company that often tests software by just going live with it. Some people call this rapid development, but I call it sloppy garbage.
Yet many Americans do believe that the US invented everything and can often recall names and dates to back this up.
Concrete examples?
Better yet, stats or studies. You can always cite anecdotes of a few people with an absurd misunderstanding of something, but inferring too much from that may be a matter of confirmation bias, or worse yet, over-generalization.
Slashdot Mirror
In an ideal world you'd be right, but in the real world the experiments you advocate aren't practical.
While I basically agree with you (I also had an A GPA, so not sour grapes) what the Google guy actually said is that GPA's don't matter after you've been out of school after a few years. For recent grads they have some correlation. It certainly isn't perfect, but with recent grads its harder to pick the best people because they have little work experience to go by. Methinks that would be a good reason for the old summer internships that seem so rare these days - it gives you a chance to find the good new talent.
This approach provides the interviewer with a lot more insight than casual chats, including helping you to find those people who are really capable but aren't good conversationalists.
That sounds like a good approach, but how do you know that it produces good results? Gut feeling or even the fact that some people you hired worked out well is not sufficient, because you really have to compare your hiring decisions to the counterfactuals. What would have happened if you hired person A instead of person B? In the real world experiments are impractical, but you can correlate how well someone does at the company with how well they did with various interview approaches. To their credit, it seems like this is what Google is doing.
Also what's described in the NYT article seems closer to the specific type of chat approach that SnapperHead and I use (perhaps a little more structured) than what you describe:
you’re not giving someone a hypothetical, but you’re starting with a question like, “Give me an example of a time when you solved an analytically difficult problem.” The interesting thing about the behavioral interview is that when you ask somebody to speak to their own experience, and you drill into that, you get two kinds of information. One is you get to see how they actually interacted in a real-world situation, and the valuable “meta” information you get about the candidate is a sense of what they consider to be difficult.
They often waive their official GPA requirements if you worked in the field while earning your degree. 25-30 hours a week as a programmer while going to college full time and most corps won't care whether your GPA was 2.5 or 3.5 when applying for a development job.
Not a bad approach. Several years into my BS I switched from full-time student to full-time employment and part-time student. My grades went down, but I actually learned more in my classes because I saw the applications. It also cured me of the suspicion that classes only taught ivory tower nonsense.
I've never applied to Google because I'd heard enough about the interview process to realize it was mostly an unintentional way of asking if you're a recent graduate with a mind uncluttered by practical on-the-job knowledge, so you can focus on algorithms and brainteasers that have very few real world applications (and none in the job you're applying for.)
Try Netflix. According to a recent Slashdot post, they prefer hiring people who've spent a few years at Google learning their trade.
I find what Netflix does very interesting - effects of scale can be serious. They also take reliability very seriously, as some people deprived of a promised premiere can be dangerous :) As is so typical for the "ooh, ahh" evaluation of tech, little heed is paid to Netflix because they're selling movies, never mind that the tech is the magic behind the service. Yet Facebook get loads of "oohs" and "ahhs". The best tech is usually the tech that gets noticed least. Nobody thinks much of running the faucet, yet you're dealing with a tech that's been a key factor in civilizations since, uh, since there have been civilizations (and probably before).
For Google interviews the answer to "run of the mill" brain teasers should be "Hang on while I Google it" ;).
And if they say no, ask them if it's better to use Bing instead.
When I clicked through to RTFA, I found that the posts all date from 2009, and the article itself is dated 10/29/09. So maybe this is old news, folks?
That's the Gawker article cited as background. The NYT article is from 3 days ago.
Catch 22: they need to find a good way to hire HR people who are good at hiring.
Best not to rely on any one criterion. Personal projects are a positive indicator, but lack of them shouldn't be a show stopper. I've known some very good people, who are very interested in their work, who wouldn't have anything to do with the work when they're not on the job. Some of them even have lives (or so I've heard).
I have found the best indicator in the world is to just casually bullshit about technology. You can very quickly find someones strengths, weaknesses and if they are full of it. In a casual chat, people let their guard down and you get a look in.
That's the technique that works best for me when I interview people. Typically I'll ask them to pick something on their resume to chat about. I expect the interviewee not to be happy about discussing everything on their resume, because they all contain some some exaggerations (hell, you should add some because everybody does). However, if you can't come up with anything that was interesting and challenging, and that you're comfortable talking about, you're probably a fake. Some people are even shy about what they think is tooting their horn, and I encourage them to open up (others you have to shut up).
However, I'm also convinced there is no one magic formula for hiring people. Different techniques work well for different interviewers (and interviewees). The best approach is to have a candidate interviewed by a number of people w/ different approaches.
It's nice to see a large company try to objectively evaluate its hiring process and express some self doubt. All to often the hiring process at a company is assumed to be good because the company is successful, which is an obvious fallacy since many factors contribute to a company's success. In fact I wouldn't hire anyone who didn't immediately question such an assumption :)
All too often the hiring process at a company, or the admissions process at a university, is treated as though it were created with some magical special sauce, when in fact it does little more than reinforce some (often unstated) prejudices. It's especially troubling coming from organizations that supposedly value rational and scientific analysis.
I think such people should have appellations akin to those of ancient Greek Heroes
Given ancient Greek tastes in intimacy, that would be especially appropriate for Turing. Shame the British government didn't see it that way.
You have to admit, for all the Facebook bashing that happens, the fact that hacks, break-ins, and bugs of this nature are so uncommon, given that they're dynamically managing a userbase of a billion people, is an impressive task.
I have to admit no such thing. First, there are a billion accounts, not a billion users (many users have multiple accounts), and many accounts are largely dormant. FB loves to hype their numbers. Second, there are hundreds of millions of bank accounts in the world, many of them now accessible online. Financial networks have been around since the 60's and have gotten much more sophisticated. While not perfect, they're incredibly more reliable than FB, otherwise we'd all be keeping money in mattresses. People go "ooh, ahh" over FB not because their technology is impressive, but because it's a cool new thing from Silicon Valley. Best renamed Silly Valley to keep up with its trends, it's gone from selling amazing tech (e.g. the first IC's) to selling amazing hype.
P.S. This comment below has some good insight.
Ebno's law: you can write bad code in any language, but some languages make it easier than others.
FB is not up 24/7. It sometimes goes down for hours at a time (second hand info as I don't use it myself).
Furthermore, and rather obviously if you understand that not every passing snark is meant to be completely literal, my point was that they don't do very thorough testing before going live. I have no idea why anyone would be impressed by most of FB's "technology". They're hardly so bleeding edge that they can be forgiven such flakiness as an inevitable part of new technology. As a contrasting example, financial networks have been around since the 1960's, and have kept getting more sophisticated. While not infallible, anybody who deployed code that lost $1B in transit would probably be shot rather than fired. It's FB's sort of flaky toys that makes many people think that software is an inherently unreliable scam.
The all knowing market also brought us the tulip bulb bubble, and that invisible hand is reaching for your wallet.
Good idea. There's something interesting about Tor I didn't realize before reading the the Wikipedia article:
Originally sponsored by the U.S. Naval Research Laboratory ... As of 2012, 80% of the Tor Project's $2M annual budget comes from the United States government, with the Swedish government
Yet the NSA takes Tor as a "definitely track this". Fact is stranger than fiction.
The urban legend is that Gore claimed he invented the Internet. What he actually said was poorly phrased, and typical of a politician, but no different from Eisenhower saying he took the initiative on creating the Interstate Highway system. Even Vint Cerf and Newt Gingrich have said the the urban legend is silly.
It's not that I'm a great Gore defender, or even that I mind people using the urban legend as a joke, but it's going too far for the aptly pseudonymous Hognoxious to use it as support for his resentment of Americans.
Let's hope it also lays to rest the myth that we only have computers because of NASA.
Sounds like you're creating myths about myths.
Nice anti-Wikipedia rant, but I already pointed out that the OP was simply mistaken and the article does mention Baby. Don't let that get in the way of a good tirade though. Speaking of facts, do you have any that demonstrate that Wikipedia is less accurate than highly regarded encyclopedias like Britannica?
Test cases? We're talking about Facebook - the company that often tests software by just going live with it. Some people call this rapid development, but I call it sloppy garbage.
You're not "smug and superior". You're full of ... reason.
It's not an either/or situation - they make a great combo.
Early computers were large and delicate. Not a good combination if it needs to be shipped across the Atlantic.
If you can ship it by road or rail without problems, you can ship it by sea.
Though the "not invented here" factor probably had more to do with it.
Evidence? Or are you just indulging your prejudices again?
Yet many Americans do believe that the US invented everything and can often recall names and dates to back this up.
Concrete examples?
Better yet, stats or studies. You can always cite anecdotes of a few people with an absurd misunderstanding of something, but inferring too much from that may be a matter of confirmation bias, or worse yet, over-generalization.