This. If you're nearly "perfectly" pricing risk on an individual level then you defeat the point of insurance which is to pool risk.
If my hypothetical cost over an N decade period is within a fraction of a percent of payouts in that time what do I gain by paying for insurance other than creating a principal-agent problem?
I can tell you, with absolute certainty, that before AI ~0 junior/mid level devs spent just 20% of their time programming. At least not at tech companies
I went into this field because I love programming. I didn't even know how well these jobs paid until my junior year of college when I got an internship at AWS. I constantly programmed and read programming texts in my spare time growing up, in college, and after work.
I love AI tools. I can have AI do the boring parts. I can even have to write polished, usable apps in languages that I don't know.
I miss being able to think so much about architecture, best practices, frameworks/languages, how to improve, etc.
https://www-cs-faculty.stanford.edu/~knuth/email.html
reply