Hacker News Reader

Comment by ritzaco | Hacker News Reader

I guess most HN readers are aware of it by now, but if anyone isn't this is basically the idea of the Lindy Effect [0] -- counter-intuitively, in many cases age is a good predictor of future life span.

[0] https://en.wikipedia.org/wiki/Lindy_effect

View on HN

> ounter-intuitively, in many cases age is a good predictor of future life span.

So people making swords was a job for 30 000 years but for some reason the Lindy Effect does not work with that one? That kind of logic does not work at all with all the jobs that disappeared in the 19th and 20th century.

The logic worked for 29999 years, which is the whole point. It's like the German Tank Problem[0], but for lifespans.

[0]: https://en.wikipedia.org/wiki/German_tank_problem

Weapon smiths might be gone but weapons manufacturing is as big or even bigger than ever in history. So until humanity does a fundamental shift and stops all armed conflicts, it is as safe job as any.

> Weapon smiths might be gone but weapons manufacturing is as big or even bigger than ever in history

That's called moving goalposts

that's exactly why it's counter-intuitive! Obviously it can't be right forever, but at any given moment if you need predict the minimum future life span of a book, framework, language, job, country, whatever and you have very limited information it turns out that just predicting the thing's age will often be a reasonable prediction -- probability is hard!

A "good predictor" doesn't need 100% accuracy.

Rather looks like 10% accuracy in this case

Reply on news.ycombinator.com