rupaul

The idea is that any one who masters some obscenely difficult area can easily be taught anything they are missing. Hell, you can tell them to pick up whatever you need them to know on the weekends. The book and the subject matter I suggested so happen to be brutally difficult. Anything they have to learn at your org will likely be trivial by comparison. Any one who learned this book on their own can easily walk in and out of any math PhD program (let alone other subjects).

I believe it's not really hard to check one's competency level in STEM fields. Suppose you're interviewing those with bachelor's degree in math. Intro to Abstract Algebra and Real Analysis are two of the core requirements for those in undergrad math. Admit all those with papers automatically. Place a super high bar of responsibility and expectations on the so called "self-taught" ones. For example, accept only those who can solve every single problem in Hatcher's Algebraic Topology cold. I doubt even 10% (numbers out of my ass) of the undergrads in the US are capable of this feat. Anyone who taught themselves this shit (gaping holes or not) can kick some serious ass. Sounds fair to me.