Secondly, if it's a structural issue, creating a new startup based around technical-recruiting is absurdly pointless. If it's structural, your startup is going to accomplish _absolutely nothing_, because it's just going to keep sending in safe candidates, in favor of qualified candidates.
However I've read posts where you disagree with the same exact sentiment: credentials don't tell even the majority of the story of whether someone is qualified.
By no means does she think that the only good engineers are credentialed.
2. Bring in 10 ugly on paper candidates, two get an offer, and one accepts - expect a stern talking to about your performance.
Imagine you're a Stanford alum (even dropout) working at Google. If someone from University of [current bad state] comes across your desk with their main work experience being enterprise consulting in Java, odds are you're going to pass and look at the next person.
Quickly filtering by these "important" things seems to make some sense except for the fact that none of these factors tend to be indicators of success or failure. You may be saving yourself some time but you're missing some good candidates too.
I went to an absolute shit school (like, bottom 150 of world ranked schools) an I had interviews that lead to offers with Google, Facebook, Microsoft, Intel, among others, when I graduated. People from my graduating class are working at many of these companies, i.e. with the same background as me.
So clearly it can't be a horrible problem if they seem happy to employ people from my no-name, considered harmful, school.
Who knows, maybe every single one of us got extremely lucky, or some engineering managers just happen to know it doesn't take a M.S. from MIT to know how to write good software.
My school is in the 151-200 ranking on the Shanghai list, I don't know if that's from the bottom or what.
Although the company I'm at right now should be considered cool (or at least, to me).
Edit: Btw I upvoted because I believe we should all try to refrain from making judgments based on incomplete information, WHEN we have the data in front of us but decide not to go through it all. I do this a lot.
But people do this all the time. When people argue in person, they spend most of their time not listening to the other side, but trying to formulate a reason the other side is wrong.
It's not the right thing to do, and I know that, and it's why I put "I'm really sorry" at the start.