I want to specifically address 1) Sam Altman and 2) Chris Fry's respective points about the problems with regards to models that align with the candidate more directly (like ours):
1) Much respect for Sam, but he's dead wrong with respect to the 'negative selection' problem - yes, good people have no problem finding work, but the key problem is that the opportunity cost remains phenomenally high for suboptimal decision making. We have actually worked with outstanding engineers who are at YC portfolio companies, simply because they wouldn't have known how to get what they're worth otherwise and push for more. And Sam is missing the point entirely with regards to worth - people in the industry are not getting paid based on their merit - not at all. The gender wage gap is perhaps the most stark example of this, but we see it, starkly, along many other demographic slices as well.
2) With respect to Chris Fry's comments - I was actually in Twitter's eng org when Chris raised the internal engineering referral bonus from 2.5k to 10k because the company wasn't getting the volume of quality people it wanted. Chris is a really great guy, but I find his point about "[at] Twitter, you get the best résumés on your desk already [via recruiting department, referrals, etc]" somewhat misleading - there's no way he would have raised the referral bonus as sharply as he did if he really felt that. In fact, there wouldn't have to be an referral bonus structure at all. Twitter is wonderful, and I loved my time there, but even there we weren't getting all the high-quality people we wanted.
[#plug: check out http://OfferLetter.io - we all deserve to get what we're worth]