Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Maybe, but there's more to it.

(1) inconclusive means we may not have looked hard enough, or

(2) the effect may be very very small, as in not distinguishable from zero, or

(3) there are alternative models of learning. People could do bayesian updating, but they may also employ thresholds, toggle between belief systems and so on.

Summary: it's very complicated, lack of evidence suggests caution.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: