3 ms·
Suppose the thesis is true -- the job market has undervalued workers who can't write JS. Why would the obvious arbitrage persist?
by dogruck 9y ago
Suppose the thesis is true -- the job market has undervalued workers who can't write JS.
Why would the obvious arbitrage persist?
- sidlls 9y agoNot that I agree that "HTML + CSS" only workers are improperly valued less than JS coders, but one reason that might explain if this were the case could be because the technology job market isn't as rational or merit-based as some assume. Instead it's festooned with the same cult-of-X (where "X" is "technology", "framework", etc.) bandwagoning effect many practitioners of programming mock other fields for having. There's no reason, as I see it, to think that an "arbitrage" of this sort couldn't persist for some time (years/decades) in this industry: it's too insular and full of itself in many respects.
- dogruck 9y agoYou certainly could be right. But what's your conviction level? Is it as high as "earth's gravitational pull is real" or is it lower? If it's high, can't you think of some way of capturing the arbitrage? Perhaps starting a company that employs HTML+CSS wizards?
- sidlls 9y agoConviction level for which? For the state of the industry comment: quite high. Developers think too highly of themselves, their ability to reason (in general), and assume unjustifiably that this industry is somehow more merit-based and rational than others. For the particular arbitrage you describe: next to none. I know enough JavaScript and enough webdevs to know I don't care to know any more. I'll stick with data engineering, computational research, and systems programming, thanks very much.