>is still the best way the human race has so far discovered to learn and advance
The way I see it - it's really successful in some domains and is worse than useless in others - because it rides on credibility it acquired in the domains it is effective in, it creates false confidence in results it produces in domains it doesn't fit.
The way I see it - it's really successful in some domains and is worse than useless in others - because it rides on credibility it acquired in the domains it is effective in, it creates false confidence in results it produces in domains it doesn't fit.