Hacker News

points by allexsendev 2 days ago | 0 comments

This has happened with most of the fields - volunteers do a lot of work, make their fields/areas good, raise interest, bring ppl into it which usually brings money then which usually brings money chasers who then usually "optimize" the hell out of every metric for maximum gains.

Great OS volunteers still exist today, but the ratio of genuine field lover to money chaser is absurd now.

[–] cowlevel 2 days ago

Goodhart's law: whatever you measure will improve. And the measurement is money extraction.

[–] jonny2811 1 day ago

any measurement that becomes a target ceases to be a useful measurement*

[–] optimalsolver 1 day ago

That's not the definition of Goodhart's law. That's the opposite of Goodhart's law, in fact.

[–] cowlevel 1 day ago

It's exactly Goodhart's law. The measured thing improves.

[–] bigstrat2003 1 day ago

Goodhart's law is that the measured thing increases, not improves. Increasing may in fact be worsening.

[–] cowlevel 1 day ago

So if I measure accidents, accidents will increase?

[–] allexsendev 1 day ago

interrupting your thread bcs what the law states directly is less interesting than what it puts the emphasis on: if maxing out measurement becomes the goal, it isn't measuring what it was supposed to. Take a concrete example - original point of schools/academia was education. Eventually grades started measuring education, and now grades are the goal instead of education. ppl started optimizing for grades by any means possible -> grades aren't good measure of education anymore.

[–] palata 1 day ago

Is there a law that says "whenever something is profitable, people attracted to money will come for the money". And obviously, the best way to make money is not to build good tech.

[–] cowlevel 1 day ago

It's called capitalism, I think, but it's only one part of capitalism