The categorization is interesting albeit deeply ungrounded in any real rigor and seems of a piece with one of his other recent essays, in which he developed a psychoanalytic theory of the various kinds of "haters" and "losers."
Further, I wish Paul Graham would try to convey his ideas with less condescension and smugness. There's a sense in which he maligns large swaths of humanity as somehow defective or worthy of shame. Certainly the term "idiots" doesn't help.
Further there's an essentialism and determinism that's sort of disturbing (labeling preschoolers as sheep is kind of messed up) and lacking in empathy.
Finally I suppose this is obvious, but I'm guessing Graham situates himself as a paragon of fierce independent-minded thinking and courage. It's rather easier to do that when you're absurdly independently wealthy. Thinking through the courageous stand countless people are taking even right now around the world, risking life and limb, just makes this feel a bit like a grievance-laden tempest in a teapot.
Yikes. You make some pretty strong claims that are not credible on their face.
1. "Not adaptive in any situation..."
Putting aside the many problems with evolutionary psychological explanations (just-so theory, underdetermination, so-called disjunction and grain problems), there's actually an very strong argument to OCD's adaptive role, both at the individual (threat response) and group. I think the group argument is most compelling, as various degrees of neuroticism have very high upside for risk management over time.
Or take depression for example. It can serve, theoretically, to reduce risk of conflict and death when social hierarchies might be in flux, it's a way to honestly signal a problem to ones group, it could be a mechanism to accurately try and signal a problem to oneself like physical pain does, its been theorized to potentially reduce risk of infection, etc.
2. These genes very clearly do not weed themselves out of the gene pool. In fact, mental illness has been on the rise, probably mostly because neurodiversity has been increasingly pathologized. The social construct in which these are considered disordered is hugely important.
3. Your comment about that empathy snp is extreme genetic essentialsm and determinism. Moreover, it's a single snp. I don't know of a single researcher who'd claim that something as complicated as empathy is either toast or not toast from a single snp.
They do in theory but there are countless excel spreadsheets in circulation with a full list of companies way in advance of demo day. If you're not plugged in and arrive at demo day, many VCs have reached out to companies way before you see them present.
The notion of control over one's life is elusive. Our futures can be wrought by trauma, poverty and the lottery that is our DNA, and by extension, our brain. Given the randomness/contingency Graham acknowledges explains his and others' success, and by implication the various dependencies en route to that success, I'd just ask for a bit more compassion for those, whether through birth or circumstance, find themselves constitutionally unable to make the most of their lives in the sense you probably mean. Something as simple as a deficit in executive function can wreak havoc on one's ability to self-motivate, just as an example.
I think there's a bit of cruelty in so casually reducing people to "losers." Taking what must be a complex set of personal histories and circumstances that lead people to that behavior and framing it as a matter of winner vs loser seems to border on smug, especially when the writer is clearly the former.