Hacker News Reader

Comment by dustingetz | Hacker News Reader

Main assumption here is that a top 1% developer costs top 1% salary, but that is probably not correlated at all, bc the real way to make money is to attach yourself to the largest possible business and then climb, and the real way to get good is to hack a lot on the most interesting unsolved problems ... these are not just orthogonal but possibly even diametrically opposed

View on HN

> bc the real way to make money is to attach yourself you the largest possible business and then climb

I don't know how you get that idea unless you are talking about the handful of C-level positions at tech giants. Otherwise switching your job every few years and increasing salary/position each time is generally the more profitable path than staying at a single company and waiting for promotion.

I think the question is also, “good in what sense?”

The really valuable skill (from a dollars & cents perspective) is to be able to provide business value to _the business you’re employed at_. That might not transfer anywhere else.

Eg I doubt my skills apply to a young startup- I’m a specialist (ML research), they need a generalist.

If I work at Google on search & get really good at solving political problems related to search, that’s a super valuable skill set _to Google_ but probably not to anyone else. I’d probably get promoted and/or get big bonuses. It also probably means that, _ceteris paribus_, I’m a worse coder, as I’m spending less time focusing on tough coding problems.

Yes, the author is assuming for the sake of argument that salary is dictated by talent. But if talent is distributed on a normal distribution and compensation is fat-tailed, that might mean that there is a non-linear relation between talent and compensation, or it might mean that other factors entirely are at play. My guess is that the compensation graph shown is skewed by the presence of executive salaries, which have been shown to be uncorrelated to corporate financial performance, suggesting talent is not a significant factor at the fat end of the distribution; more likely, the fat tail is a social network effect. To be fair, the author concedes this possibility in a footnote.

Top 1% developers definitely average about top 1% salary. High-variance returns like startup stock luck don’t count.

If you’re not paying your devs substantially more than industry average (like several times more) they’re probably not that abnormally good.

or maybe they work at a university for dirt outside the USA

Wouldn't the real way to make money to be a founder/cofounder at the next Google/Facebook/Amazon/whatever? Which I'm guessing would also not require being a great programmer.

Reply on news.ycombinator.com