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

The Behavioral Insights Team was and is a unit established to implement the ideas of (now Nobel-memorial-laureate) Richard Thaler on 'nudging'. The tiny problem with those ideas is that the evidence we have is that they don't work, at least no more than the RCT equivalent of a placebo (ie publication bias) [0]. Mentioning it in the context of psyops seems like a bit of a red herring to me, the biggest conspiracy here might just be the one to create prestigious and lucrative jobs in government for social science graduates.

[0] https://www.pnas.org/doi/10.1073/pnas.2200300119



Regardless of its effectiveness (I largely agree with with your posted study), the negative externalities of attempting to "nudge" and control the narrative are enormous. You now have government departments with an existential motive to subtly alter news stories, guidelines, public messaging - even if the results are dubious, and could have unbounded negative impact on public trust and understanding, etc..


> could have unbounded negative impact on public trust and understanding

Very much this!

And to the parents comment, they are seldom "lucrative" or "prestigious" gigs afaics.

The bind is that to remain silent, to not engage in counter-influence in a world where the Internet is "weponised", seems like giving tacit assent to mischief makers.

The mistake is to think that lies can be countered with more lies. The best that the "truth" can do, whether as science, education or accurate news media, is to de-escalate and neutralise information warfare itself.


I think the study you posted is valid, but I'll bet that some individual nudges really are effective.


But how to prove your particular nudge really worked? I'm reminded of Malcolm Gladwell's The Tipping Point - ok, yes, maybe a bunch of kids in New York started a Hush Puppy revival, but a million other fads didn't take off, so if you try to study it you're just studying survivorship bias. Or maybe coincidence.


You affirm that the study is valid yet still decide to override it with your gut feeling.


Valid for an average of all nudges. We need to do the same publication bias test for individual nudges.


It seems Meta is one of their prized clients.

Color me surprised... not.




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

Search: