[gozur88]

>If you use a resume as a litmus test for "smart programmers" as opposed to those faking it, you're undoubtably going to get a bunch of people who went to the best possible schools.

Yes, and that's exactly what happens at the most desirable companies.

>Maybe there's someone just as smart at your local state school who couldn't afford CMU who would bring a completely different mindset and style of thinking, but you'll never find them.

If you're really a stand-out, IQ-wise, you can go to the top tier schools even if you have no money. The people who get screwed are the people who are pretty smart but not exceptionally so.

[jamesroseman]

I'd wager there are more people who have snuck into top-tier universities than there are people who are there without paying/accruing massive debt. This is also ignoring the fact that Ivies and other top-tier universities are overwhelming homogenous.

More to the point, the thing everyone in a top-10 school has in common is that they're really good at looking good on paper. It says near nothing about what they'll contribute on a team, how they'll function with other people, or base programming skill. My only point is that if a company seeks diversity, they should seek it in all avenues.

[gozur88]

>I'd wager there are more people who have snuck into top-tier universities than there are people who are there without paying/accruing massive debt.

Probably. The people who go without paying are pretty desirable from the university's point of view.

>This is also ignoring the fact that Ivies and other top-tier universities are overwhelming homogenous.

I'm not a worshiper of diversity for its own sake. The benefits seem pretty theoretical, and there's a lot of evidence to suggest homogeneous populations function better.

>More to the point, the thing everyone in a top-10 school has in common is that they're really good at looking good on paper.

Being able to do things well on paper is pretty important in 21st century US. It's also important to hire people who can figure out what your customers (internal and external) want and put it on paper in a way that will make the sale.

>It says near nothing about what they'll contribute on a team, how they'll function with other people, or base programming skill.

That's all true. The problem is we don't really have any way to get a good picture that kind of stuff outside of hiring everyone and culling the ones we don't want. Pretty much every applicant to an ivy league school lists a nonprofit on their application these days, and 99% of them are bullshit.

>My only point is that if a company seeks diversity, they should seek it in all avenues.

I suppose, if that's your goal. Were I running a company I'd be mostly focused on profitability.

[jamesroseman]

I think maybe we just have an idealogical difference of opinion here, and that's fine.

It's my belief that if you take 10 similar people and give them a problem to solve, they'll all solve it the same way. Whereas if you take 10 different people and give them a problem to solve, they'll each individually be challenged to think differently and question their assumptions. If I ran a company I'd prefer the latter, because intuition tells me it will lead to the best answers (people questioning their opinions means thinking through whether they're as well founded as they hope, and teases out the problems with those opinions).

I think this innovation goes hand in hand with profitability. I'm not saying I wouldn't pick the creative driven innovative Harvard grad every time -- I'm saying I'd pick the creative driven innovative community college grad every time over the Harvard grad with the same background as everyone else in the company.

> Being able to do things well on paper is pretty important in 21st century US.

Except when it comes time to actually do work, it's not. The overlap in relevant skills between convincing people you're a good engineer and being a good engineer is very small. One is mostly about leveraging your background and social engineering, the other is about how you solve problems. While one could argue that being good on paper is good indicator of drive, I'd counter that a candidate with drive but without skills is less preferable than a candidate with skills.