The argument that the H&B camp makes is that heuristics and biases get in the way of proper judgments and decisions because they are shortcuts that our brain makes to conserve resources. The argument that NDM makes is that heuristics are how our brain works, and we should spend more time utilising how it naturally works to make better decisions, instead of fighting it and trying to build intuitions for probability (which are unnatural, and not what our brains are built for).
The disagreement led both of Kahneman (H&B) and Klein (NDM) to write a joint paper: https://www.fs.fed.us/rmrs/sites/default/files/Kahneman2009_... the key takeaway of which is that the H&B approach is best suited for domains where expert performance is bad, and the NDM approach is better suited for domains where expert performance is reliable. The paper describes the conditions of the two domains.
I've written a longer post summarising the findings from NDM at https://commoncog.com/blog/putting-mental-models-to-practice..., which exists as part of a longer series on putting mental models to practice. (Earlier parts cover the foundation of the judgment & decision making field, which is mostly built around the H&B tradition at the moment).
I’d like to invert your key takeaway by looking at it a bit differently by framing it in terms of the environment being observed rather than in terms of the expert’s performance.
I’m reminded of this Murray Gell-Mann quote: “Imagine how hard physics would be if electrons could think" which I came across while reading an Economics piece. The writer used the quote to illustrate how much harder Economics as a field is, relative to a field like Physics. I agree with that characterization because, in general, electrons and by extension inanimate objects are not susceptible to behavior change from external stimuli, but humans are. Which is why expertise can be considered reliable when the environment consists of inanimate entities (e.g. astronomers, accountants) or animate but motiveless entities (e.g. livestock judges, grain inspectors).
How reliable an expert performs in an environment largely depends on the level of autonomy enjoyed by the entities that constitute the environment under observation. An expert’s ability to reliably understand an environment correlates inversely with the number of entities that enjoy autonomy in that environment, because it is infeasible to model a real-world environment in your head. What experts do is to mentally model a tiny slice of reality from which to draw inferences from.
Humans are entities that can think and act independently before, during and after an expert has been asked to intervene, which is why Doctors, Nurses and Auditors appear in both columns of reliable and unreliable performance for experts.
(PS: I encountered 2 occurrences of the typo: “Marlie Chunger“)
EDIT: reworded my comment significantly to make my point more clear.
(Also, Marlie Chunger is a fictitious character invented to illustrate some of the ideas in the series.)
I did read your summary in the section titled A Personal Take, unless there is another summary? Link?
(I’m wondering if your blogging tool allows a way to jump directly to sub-sections using anchor-links. This would be a time-saver for long-form essays.)
(I was going to look at the essay again and ask myself if I made a mistake, since everyone seems to stop at that heading. It also meant my structure was badly chosen. And then I saw your reply!)
Then, to address your idea more directly: I can see where you are coming from, but I think Klein and Kahneman’s definition of ‘has reliable causal cues’ + ‘provides opportunities to learn those cues’ are better filters. Klein studies some incredibly high performing individuals — experts who are able to perform well even in very complex environments. So it’s not always true that complexity of environment determines existence of expertise.
If you get the former without the latter, all that means is you're better at defending the positions you happened to reach: you're just as likely to reach wrong conclusions, but are now less likely to correct them. You shot yourself in the foot.
Unfortunately, it's a lot easier to spot biases in other people's arguments that you disagree with, than it is in any arguments you do disagree with.
If you want to get better at critical reasoning skills, I think it's a lot more productive to start with Bayes' rule. Learning 100 ways to not build a light bulb may help, but getting just 1 correct blue print would help a whole lot more.
[1] https://www.lesswrong.com/posts/AdYdLP2sRqPMoe8fb/knowing-ab...
Hang around old people who know what they are doing, and observe. Its possible to do it without getting frustrated and without adding to frustrations. Observe their use of language. Observe what they avoid doing and ask why. Hang in there and keep practicing. It pays off.
https://slatestarcodex.com/2018/05/08/varieties-of-argumenta...
If you're sick of it, that implies it's all over the place, but I for one have no idea what you're talking about. Who focuses on cognitive biases in arguments? The blog post says nothing about arguments.
In poorly written news articles and most social media arguments, there's a trend of identifying the person you disagree with by a catgory like rich / poor, white / black, radical left / radical right, etc. before arguing the point.
The insinuation is that your identity reveals your biases and your biases invalidate your argument.
Of course, trying to stop a debate by telling someone they're biased can backfire: https://youtu.be/VSHyAwErdFM
Thanks for the feedback and kind words. I'm absolutely considering doing "random x definition". I'll have to think of other subjects that I'm interested in but in the meantime, I'm happy for you to make some suggestions :)
On a less positive note, and tangentially related:
I'm kind of sick of the whole "bias" obsession. It's everyone's go-to counterargument these days, and it's a shallow, poorly developed one. It's like everyone's lost critical reasoning skills, which require delicate attention to the particular strategies and propositions deployed in a given argument, and found these set of stock biases to use instead. In fact it's impossible to purge an argument or line of thinking of all so-called biases (though these don't actually exist in arguments, they are deduced from arguments)--if it were, it wouldn't be an argument or thought.
The goal of catching our own mistakes is an admirable one, and I'm not advocating people stop doing that--I just think it too frequently bleeds into trying to find so-called biases in arguments (whether written or verbal). In fact, this is more or less a fool's errand. What people are actually trying to point out in arguments are logical fallacies which are traits of the argument. Biases contrarily occur at the individual level and are operational flaws, they only occur during the thought process, and it's only meaningful to talk about them in these terms (that is, as they manifest in the ongoing practices of a person)--they are not properties of a line of thought's encoding (the written or spoken argument). Fallacies or viewpoints expressed in an argument may hint at the biases of the author, but it's a non-sequitur to start talking about them (when critiquing an argument), as the only way one could actually confirm this is by observing the author at work in daily life. To say, such an such an author is biased, is useless. It doesn't contribute meaningfully to a critique of the argument, and it would need to be verified through observation of the author.
Demonstrating to someone that they have developed/fall prey to particular bias frequently and working to rectify that one-on-one is a totally different story, or trying to catch biases operating in yourself is a totally different story.
Edit: I suppose you could say I'm biased against biases. A joke that illustrates my point.