5 ms·
Why We’re Biased About Being Biased
- surement 10y agoWikipedia has a pretty fascinating list of cognitive biases, it's fun to go through it and mark each that applies until you start looking for the bias about thinking you have all of them: https://en.wikipedia.org/wiki/List_of_cognitive_biases https://en.wikipedia.org/wiki/List_of_cognitive_biases
- antisthenes 10y agoThat's the elusive meta-bias.
- marcosdumay 10y agoOf course you have all of them. Those are common patterns in human thinking, and I guess you are human (correct me if I'm wrong). You just don't have all of them at the same time, and have more bias in some situations than in others.
- blackRust 10y agoI liked the graphic poster drawn up by Business Insider: http://uk.businessinsider.com/cognitive-biases-that-affect-decisions-2015-8 http://uk.businessinsider.com/cognitive-biases-that-affect-d... Here's the image directly to avoid potential paywall: https://static4.uk.businessinsider.com/image/56a111c8dd0895ac0c8b46ca-1200-2247/bi_graphics_20-cognitive-biases-that-screw-up-your-decisions.png https://static4.uk.businessinsider.com/image/56a111c8dd0895a...
- golemotron 10y ago> The “moral credential effect” describes this compensation in the context of moral reasoning. When study subjects were given an opportunity to disagree with sexist statements, for example, they were then more likely to favor giving a stereotypically male job to a man instead of a woman (compared to people who weren’t exposed to the statements). Likewise, people who believed they were morally good were more likely to cheat on a math test. That's interesting. It means that if we want social justice, the last thing we should do is make people feel that they need to take public stands for it.
- antisthenes 10y agoI took a look at the referenced paper the nautilus article cites, and it is extremely poorly designed, to the point of being near useless to make any sorts of conclusions and definitely not good enough to create any sort of social justice policies. Not surprising, considering it's a psychology paper, but still.
- golemotron 10y agoThere are 19 research references on the moral credential effect [1] and self licensing [2] on their respective wikipedia pages. It might be a bit of work to go through all of them looking for disqualifying flaws. Maybe we can assume they are ok tentatively? [1] https://en.wikipedia.org/wiki/Moral_credential_effect https://en.wikipedia.org/wiki/Moral_credential_effect [2] https://en.wikipedia.org/wiki/Self-licensing https://en.wikipedia.org/wiki/Self-licensing
- meowface 10y agoI think the parent post was referring to the specific study regarding disagreeing with sexist statements -> favoring gender stereotypes in hiring. I don't think he was questioning the validity of the moral credential effect.
- jkraker 10y agoSocial media gives a lot of people the platform to do this and some use it quite heavily. It would be interesting to see the correlation between something like "ferocity" on Facebook and real life actions.
- golemotron 10y agoI don't know if there are any studies about this but I don't see anything that would make it different. Social media gives us 24/7 opportunity to feel good being on the "right" side of an issue. I wonder whether there's an inverse correlation between social media use and direct action on civic and social causes: writing a Congress person, volunteering at a homeless shelter, or giving to charities.
- daveguy 10y agoThis is why AI practitioners will never make artificial human intelligence. It will be found to be way too bug ridden before the hard work (expense wrt power, computation, bandwidth) of getting it trained to a human level is completed.
- daveguy 10y agoTo clarify, the emphasis there is on human. I don't doubt 50+ years from now we will have general purpose artificial intelligences as sharp as humans. They just won't be very human like. Typical heuristic errors like confirmation bias will be quantified and that's about as inhuman as it gets.
- solipsism 10y agoYou've failed to account for the inevitability that general AI itself will be extremely interested in modeling human cognition.
- runeks 10y agoAll life forms are biased towards their own survival. That's why they're alive in the first place.