I gave the feedback at one Google interview that they should send Google employees through to see how many get hired. Good to see they basically tried that.<p>The conclusion at the end that bringing someone on board is the ideal method is true I'm sure, but even that runs into the issue that employee evaluation is an even worse situation than the interview process.<p>You can openly see some managers panic when they realize they have no idea what their employees have been doing for the last 6-12 months when they're asked to provide feedback.
2 anecdotes ...<p>1) The worst interview I ever had (BY FAR) was at Google--disrespectful people, no respect for time, I could go on and on. And I went back to try again to get that money showered on me. Worth it in the long run.<p>2) Their new system for "performance management" is a hoax. Just like at all other places, it "documents" what you should do so they can fire you more easily with unspoken rules and all sorts of arbitrary causes as well. A friend literally hit EVERY pre-agreed target and still got pushed out for "not delivering".
> I gave the feedback at one Google interview that they should send Google employees through to see how many get hired. Good to see they basically tried that.<p>They did, but not with the intention of doing anything about the problem.<p>This is a question of <i>reliability</i>, the conceptual 'correlation' of a measurement instrument with itself when measuring the same thing.<p>Reliability is one of two major concepts in psychometrics, the other being <i>validity</i>, the conceptual correlation between a measurement instrument and that part of reality that you're hoping to measure.<p>The question behind validity is "I want to know X; if I measure Y, how <i>helpful</i> will that be?". And the question behind reliability is "if I measure Z, how <i>accurate</i> will that measurement be?"<p><a href="https://en.wikipedia.org/wiki/Reliability_(statistics)" rel="nofollow">https://en.wikipedia.org/wiki/Reliability_(statistics)</a><p><a href="https://en.wikipedia.org/wiki/Construct_validity" rel="nofollow">https://en.wikipedia.org/wiki/Construct_validity</a><p>Yegge calls out both concepts explicitly, though not by name, in this essay:<p>>> The outcomes from interviewing are statistically terrible. Google did wave upon wave of analysis over the years, and all the results were incredibly depressing.<p>>> [reliability] To name just a few off the top of my head: interviewers barely agreed with each other. Put the same candidate in front of two of our sharpest people and you’d routinely get a confident “strong hire” from one and a flat “no” from the other.<p>>> [validity, though the 'problem' here is strongly confounded by a restriction of range issue] And once people were actually on the job, their interview scores told you next to nothing about how they’d do<p>>> [reliability] Hell, some of our star performers failed their Google interviews four or five times, finally got in after 2+ years...<p>>> [validity] ...and then outshone everyone else.<p>The discussion of how interviewing outcomes are statistically terrible would benefit from naming the ways in which they're statistically terrible. Knowing the problem you have is an important step toward solving it.<p>(And as a side note, the last I heard from Google, you're not allowed to interview more often than once a year. Interviewing five times in two years would seem to violate that policy.)<p>It is a basic theorem that the validity of any instrument is bounded above by the square root of the reliability. It isn't possible for an unreliable instrument to be tightly correlated to reality, because it is, by definition, not tightly correlated with anything. That's what it means to be unreliable.<p>Thus, any company that wanted its hiring process to be <i>good</i> would necessarily be extremely concerned with making that process <i>accurate</i>; you need to come to the same decision when you assess the same person. This is something that interviews cannot achieve except at extreme cost. You'd need far more than five interviews to get a reliable assessment from them, despite the claim in this essay that "any more than four interviews and you're just playin' with your food". Of course, the Google interviews aren't supposed to be reliable anyway, so in that sense the claim is probably accurate.<p>The prescription Yegge offers is valid. Multi-month work assessments will give you a strong, reliable, and valid signal. They're also very expensive.<p>Another thing the essay completely glosses over is that this problem has been recognized for a long time, and we already know how to do assessments that are reliable, valid, and cheap to perform. They're called standardized tests.
At least historically, Google prioritized not hiring bad candidates over hiring good candidates. So it was neither a priority for interviews to be consistent (for good candidates) or for employees to be able to consistently pass interviews.
Serious question, tell me what you think of using IQ tests to hire SWEs? Should we just do that instead?
The gold standard in hiring qualification is work-sample testing. It works fine. You do not need to "make hiring a profit center" or "provisionally hire" or do internships. Work samples done correctly demand less time from candidates than interviews and scale better than interviews. They are standardizable and iterable.<p>What I feel like I'm reading here is someone who has been poisoned by FAANG hiring practices --- and they are terrible --- and has missed most of the work that's been done (outside of Google's admirable work in debunking their own processes).<p>I appreciate the "kitchen confidential" here, but with respect to Yegge, I think he's been working at the Olive Garden this whole time. Go stage at Gramercy Tavern! They're working at a different scale, yes, but you'll at least get a different perspective on the "gold standard".
> One day, the recruiters gave us a special round of packets to review. In these special packets, we were able to read the interviewer notes and candidate responses. All personal details were stripped out, and we were told it was a “calibration exercise.” We had to do our regular voting job with these special packets, and see how it went. I think we may have assumed they were from another site, since cross-site calibration was common.
Our group did our job, and voted not to hire about 2/3 of the packets. This was about par for the course.
But surprise surprise, this time, those were our own packets from when we had all interviewed at Google. The recruiters had tricked us into reviewing our own interview packets, and we had voted not to hire most of our own group.
For that brief moment, we all had a glimpse into how utterly broken our process was. The people-team had rubbed our noses in it.<p>Or maybe the company changed in the 10 years or so since everyone in that room was hired and the employee needed 10 years ago is not the same as the ones needed now?
Seconded, stop the theatrics and gatekeeping and let's keep a growth mindset while training / retraining those who 'pass the buck' overtime. At least everyone can get skills and talented outliers will find themselves with more structure and collectively we'll produce better outcomes for more engineers at multiple levels of experience.
I couldn't find the text of this joke, attributed to Dirac. I'll paraphrase.<p>A man walks into a pet store. There's a parrot for $100, it says this parrot speaks perfect English . The one next to it is $1,000, and says, This parrot speaks 12 languages fluently.<p>Then there's a bedraggled looking, droopy, parrot, and its label simply says One Million Dollars.<p>Does it sing opera and has successfully run for President? the man asks with a sneer.<p>This parrot, says the store owner, _thinks_.<p>That's what this entire post is about - how to evaluate people with a series of attributes, score, correlate, blah blah blah.<p>Hire them, see if they think. If they don't, fire them. It's cheaper than this credential/signal rigmarole, most of which is about CYA legal b+llsh1t. Yes, it's a simplistic strategy and it doesn't work for Shoogle, Banthropic, Goober, whatever. You know what, boo f*cking hoo. You're a trillion dollar company, suck it up. You have a zombie horde at your doors and you're just upset the "true gems" are hard for you to spot amongst the slavering masses. You're going to heartlessly lay them off anyway in a few years. You SHOULD feel this pain and anguish of having to sort through them, constantly regretting all your choices. That's the only way to have balance in the Universe.
Our hiring dis-function is because there a lot of people that dislike conflict and firing someone.
There’s an old Malcom Gladwell podcast episode, I think the show was Revisionist History, where he says he’s an interview nihilist. As long as the person seems reasonably capable, and can probably do a bit of what you need, hire them. Interviews are so hard to get right that what you’re saying ends up being most effective.<p>Edit: Didn’t link it initially because I thought it would be hard to find. Turns out it’s not. <a href="https://www.pushkin.fm/podcasts/revisionist-history/hamlet-was-wrong" rel="nofollow">https://www.pushkin.fm/podcasts/revisionist-history/hamlet-w...</a>
[dead]
His idea has some merit but will require the old system to completely crash out before anything new will be considered and I'm not sure if it will crash or just keep limping along. If it really does crash out hopefully we will see multiple new strategies emerge as there are many possible options once the current one is off the table.
The current interview process isn’t meant to get the best talent, its intention is to give managers on visas an easy way to bring in other visa workers (via sharing questions and scaling difficulty for no-visa applicants)
The most talent dense place I've worked had a dead simple process - two one hour chats, one with your potential manager/team and one with the CTO or CEO. If things didn't work out, well you got sent away, probably happened twice per month. There was a particular meeting the CTO/CEO used and if you saw someone meeting with them there on a Friday afternoon, you would not see that person on Monday.<p>The place was not big - never got much beyond 100 engineers, but produced dozens of founders, VPs, GMs etc at well known companies, as well as engineers with very notable OS projects and lots of high-placed engineering in FAANG companies.
Steve Yegge is one of my very favorite authors I love his work. But lots of things to say here.<p>FIRST - is that before you get to campfire anyone, you have to have done some sort of interview process to boil it down to the one person to do the campfire - so how does that work eh?<p>SECOND - campfire is deeply invasive to the candidate's life and time.<p>THIRD - you have to pretty damn sure someone is a hire before doing a campfire. You CANNOT do campfire as an evaluation step, after which there are more interviews.<p>FOURTH - this is effectively just a really really long version of the take home work test which is absolute bullshit.<p>FIFTH - there's STILL no science to the campfire. Don't give anyone a fucking test if you don't actually know how to scientifically evaluate the results. And campfire does NOT result in a scientific outcome, it still results in an arbitrary opinion.<p>SIXTH - any company that wants me to do a campfire - to commit days or even weeks of work as part of them trying to decide to offer me a job - can fuck off. Sorry, the party got spoiled by all the other companies who asked me to do something as part of the interview process and then either ghosted me or gave me some bullshit outcome like "they didn't like your work".<p>I can tell you how to recruit people and it does not require campfire.<p>You TALK to people about software development - you engage them in extended conversation about what they have done, what they know, what their interests are, what they have built, what projects they worked on, what went right, what went wrong. You look for people who have BUILT STUFF - this was true before AI and is 100X more true now - anyone who has not built anything today is not worth employing and anyone who has built something must be able to talk about it in depth. This interview processes worked before AI and it works after AI. And finally, you accept the limitations of recruiting which is that people are people and you won't find out how well someone performs until they have been on the job six months - live with it.<p>Sorry Steve - I love your work but I'd never work for any company that wanted me to do a stupid campfire because they don't know how to actually work out if I can do the job or not.
This sandwich is huge but the bread layers (top and bottom) are moldy, love the empathetic writing style though. Agree on your TALK paragraph - mostly - as well. Make sure to consider juniors in that lumping too though, we can't have systemic failure over time. AI-outcomes are not guaranteed.
And good riddance.