3 ms·
There's 2 huge huge problems not mentioned in the article: 1. Take Taleb's piece on IQ[0]. Which journalists have the mathematical background to understand it?
by concinds 4y ago
There's 2 huge huge problems not mentioned in the article:
1. Take Taleb's piece on IQ[0]. Which journalists have the mathematical background to understand it? None. Most of the psychologists who reponded to Taleb didn't understand his math either (some did). So how do you know if Taleb's bullshitting? McClure tried to dumb it down in a great article[1], but even if you understand those arguments, practically no journalist would want to go on a limb and declare a "consensus" is dead wrong, when both sides have seemingly coherent arguments.
2. There's always what I'll call "pendulum effects". If science journalism took the "skeptic" POV to try to increase scientific literacy, by showing the many cases where the scientific consensus was wrong historically (with the hopes of improving people's critical thinking), you'd just end up giving fodder to people to arbitrarily distrust any consensus they don't like (for political reasons). So it becomes "socially useful" for people to believe that consensus means truth. If you taught people all the ways in which science can get it wrong, even highly cited studies in top journals, they'd burn Harvard down. The bias is in the opposite direction: Freakonomics-style 'science journalism' that takes one finding and turns it into an overarching narrative 'truth'.
Science in more complex fields (quantum, economics, sociology) is getting out of reach of journalists and average people. People apply simplistic and inadequate statistical tools to complex systems and reach BS conclusions. Complexity science isn't taught enough. Will journalists ever understand complexity theory? Power laws? If even AI researchers can misunderstand what they're doing[2], how can journalists keep up?
[0]: https://medium.com/incerto/iq-is-largely-a-pseudoscientific-swindle-f131c101ba39 https://medium.com/incerto/iq-is-largely-a-pseudoscientific-...
[1]: https://seanamcclure.medium.com/intelligence-complexity-and-the-failed-science-of-iq-4fb17ce3f12 https://seanamcclure.medium.com/intelligence-complexity-and-...
[2]: https://twitter.com/sean_a_mcclure/status/1534231723851907072 https://twitter.com/sean_a_mcclure/status/153423172385190707...
- cycomanic 4y agoTaleb's piece is great, thanks for posting!
- zeroonetwothree 4y agoI’m sympathetic to the argument but Taleb’s writing is horrendous. He doesn’t explain things well at all, seems to mainly criticize strawman datasets, and primarily engages in ad hominems against anyone that doesn’t agree with him.
- patrick451 4y ago> 2. There's always what I'll call "pendulum effects". If science journalism took the "skeptic" POV to try to increase scientific literacy, by showing the many cases where the scientific consensus was wrong historically (with the hopes of improving people's critical thinking), you'd just end up giving fodder to people to arbitrarily distrust any consensus they don't like (for political reasons). So it becomes "socially useful" for people to believe that consensus means truth. If you taught people all the ways in which science can get it wrong, even highly cited studies in top journals, they'd burn Harvard down. It's such a bizarre argument to simultaneously acknowledge that The Science is often wrong yet double down on manipulating the public with this idea that scientific consensus equals Truth. In my view, this practice adulterates science into a technical sophistry that serves as a fifth column for a political agenda. But the root cause here is that science itself is corrupted by the desire to effect change. That is the realm of politics and if scientists want to stop losing trust, they will get out of the business of "ought".