I really wish someone had explained this to me while I was in college.. I had ONE professor who'd actually worked in the real world who explained it, but all the other professors were lifetime academics who just parroted the same idiotic stuff about how important it is to get good grades.
Getting my first engineering job after I graduated was a nightmare.. I ended up having to go with a crappy internship that barely paid enough for me to live in a bedroom in a shared house, with a terrible boss who constantly threatened me and stressed me out. And that's with a 3.8 GPA.. When I told prospective employers that, they were like "What is that? Out of 4? 5? Is that good?"
A friend of mine in the HFT world told me “I know how hard some of your courses were. But you can’t highlight them on your cv because no one will believe you that they were that hard. And even if they did, they wouldn’t believe that those skills are transferable.”
Going to yet another job interview, I spoke to my interviewer who had a similar CS specalization but from another uni. He told me “my master was of baaad quality.” I responded with: “mine wasn’t, and my classmates all feel the same because we’ve learned practical skills and our teachers prided themselves in giving difficult yet practical courses.” He looked at me with disbelief.
Back in undergrad, I was talking to a hiring manager who was an alumni from my school that said if he will hire anyone no questions asked if he sees anyone with X course on their transcript/resume (listing the project from the course)
OP is not comparing a caltech grad to someone who contributed to open source. OP is comparing caltech grads to someone who studied algorithm interviews.
I agree. Degrees and accreditation are sort of gateway policies to conduct deliberate filtering and maintain a class based social hierarchy. There is some merit as it does take a brutal level of hard work and intelligence to get into caltech.
However, algorithms are the exact same type of filter, both are pretty much divorced from any relationship to the job at hand.
True, but it's probably not a worse predictor than most interviews.
A good interviewer can assess a candidate's competence even with crappy questions / question bank. A bad interviewer will often make wrong hiring decisions even if they asked good questions.
It seems to me that four years worth of coursework and exams provide a better assessment than a couple hours of on the spot performance.
If I had the choice between a Caltech grad who had all A's and no experience or a student from some random state school who had interned every summer and contributed to open source, I'd pick the latter. Many people are locked out from ever attending elite schools in the first place despite competence due to coming from poor backgrounds and underfunded highschool districts.