Hacker News new | past | comments | ask | show | jobs | submit login

Are humans capable of detecting and understanding what that amplification means? Even though our pattern-matching is a wonderful skill, it doesn't always serve us well. We exaggerate local effects. We are known to be terrible at understanding risk. And I've even heard that nine out of ten dentists are bad at statistics.

I'm not saying we shouldn't ask the question. I'm saying we need a way to answer it that factors out human perceptual error.




Feedback loops are shorter, news reports are increasingly more useless[0], and our current age is somewhat unique - reporting of the past didn't have a strong economic system attached whose sole purpose is making it less truthful, less accurate, and more disagreeable. I'm, of course, talking about funding media through advertising impressions at article granularity.

--

[0] - Gwern makes a really good argument that with the amount of people we have on the planet and how information moves near-instantly around the world, you can plausibly assume that all news reports are flukes, one-in-a-million events, "rare datapoints driven by unusual processes such as the mentally ill or hoaxers are increasingly unreliable as evidence of anything at all and must be ignored." https://www.gwern.net/Littlewood




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

Search: