I found Daniel Kahneman’s Thinking Fast and Slow persuasive on this question. I don’t have a psych or neuro degree, but Crudely speaking and possibly butchering it: the mind reflects two types of thinking, which Kahneman terms System 1 (fast) and System 2 (slow). System 1 is closer to instinct and helps us respond quickly in, for example, fight or flight situations. System 2 requires significant mental effort to more thoroughly analyze things like complex math problems. It’s easier to coast on System 1 thinking. The book provides examples and much better and more in depth explanations. There was a replication controversy about some of it, but still very worth reading I think.
Fair enough. I’ve read Kahnemann’s book, and hadn’t considered how to integrate it with TFA.
The sheer number of cognitive biases presented in TFA, and the overlap of the categorisations had me bamboozled.
But yeah, if these are System 1 staples then I can imagine how they might have made do for regular people for a long time. Some biases and prejudices affirm survival in a less civilised setting eg prehistoric tribes that compete for land and food with neighbouring tribes.
So perhaps it is the compact of civilisation that opens up the vistas for System 2 thinking to yield benefits.
They don't have to have an individual benefit (or even be an intentional thing), evolution isn't great at perfecting things it deals very well in just good enough and doesn't really have a great mechanism to deal with small issues with small effects to reproductive strength. See the recurrent laryngeal nerve for a pretty simple example of it. [0] It has little/no downside to reproductive ability so there's no reason to 'fix' it evolutionarily speaking.
For the vast majority of our time evolving the biggest issue was spotting a stalking predator or hidden danger before it could hurt us or spotting food better than other chimps/homo */lizards (at various points on the evolutionary history). For that quick efficient pattern matching is what you need not solid reasoning about large groups so that's what we've got.
This is an active area of research in the cognitive science world.
First, 'bias' is defined with respect to an ideal rational actor with perfect information and infinite computing power. The reasons for this are mostly historical (rational animal, Homo economicus, etc), but it also serves as a useful baseline 'ideal observer' model to compare human behavior against.
In the 1950s, Herbert Simon coined the term 'bounded rationality' to describe rationality within a set of computational bounds. For example, if we have finite working memory and limited computing time, but were still trying to make optimal decisions within those bounds, what behavior would we see? In this case, decision making turns from unconstrained to constrained optimization. What may LOOK like a 'bias', with its connotation of sub-optimality, may actually be optimal behavior given constraints.
More recently, people like Gerd Gigerenzer suggested that human decision making is largely composed of heuristics and tricks that enable 'fast and frugal' responses to scenarios. They don't need to be perfect, just good enough - and 'cheap' enough with respect to time that they are worth developing. This is probably true to a certain extent, but to me it's scientifically unsatisfying, as there is no general principle (except for 'cognitive miserliness') to explain behavior - and specifically, there is no longer a way to specify 'normative' or expected behavior in any given situation.
More recently still, there is a trend to revive Simon's perspective under the name 'Resource Rationality'. Tom Griffiths is one of the active researchers in this field. The idea is the similar to Simon's - we have limited 'cognitive resources' and strive to be rational. Griffiths and others have attempted to show that many behaviors that are traditionally called 'cognitive biases' are actually predicted if we are behaving optimally but with constrained cognitive resources.
From the resource rationality perspective, a cognitive bias is a way that a solution to unconstrained optimization differs from a solution to a corresponding constrained optimization problem. Roughly speaking, any combination of limitation on (memory, computation, time, energy, information) will produce a 'bias', and different scenarios we encounter push up against these boundaries in different ways, leading to a plethora of 'biases'.
> a cognitive bias is a way that a solution to unconstrained optimization differs from a solution to a corresponding constrained optimization problem
That is a narrow and incorrect definition IMHO. Two resource constrained solutions can also differ wildly in their rationality (or reality approximation if you will) and there is no resource unconstrained general intelligence in the universe so we wouldn’t even really know what that normativity looks like.
Biases are more of a classification of errors specific to our cognitive machinery; framing errors, recall errors, precision errors, proportionality errors, inference errors etc wrt the best we could have done with it.
Yes it's narrow. I was trying to summarize the 'resource rationality' approach, and there are certainly other legitimate accounts of cognitive bias. My interpretation of that work is that it's a project to convert the list of cognitive biases from a classification of errors to a general generative principle.
I think it's mostly because they are quite ad-hoc and superficial. It's often not even clear what is the normative behavior that the bias is deviation of, or whether that normative behavior is somehow better than the "biased" one. It's a weird scene.
I think the most obvious answer to this is that understanding the world perfectly is impossible (for an individiual human in a finite lifespan), and so in reality we use a whole slew of hacks and simplifications. Things fall to the ground at about 9.8m/s. Is that true? No. Is it helpful to assume it's true, sure. And so it's extra-ordinarily easy to therefore find conditions where these simplifications don't hold. We optimize for survival and reproduction. Anchoring[1] is a great example - as long as no one knows about anchoring, anchoring is a great technique for negotiation, in fact there are studies that show some form of anchoring is optimal. It's far more optimal to use these cognitive biases than to invent new algorithms for life- because most of these are incredibly difficult problems- what is maybe more surprising is how effective our cognitive biases are and how they've propogated, isn't it more crazy that we're all fairly good at applying these intuitive rules of thumb?
[1]: Setting an initial price in order to later favourably negotiate a price.
My own speculation basically boils down to: Human minds are primarily meaning-seeking machines, to such a large extent that we even create meaning where there is none (apophenia).
Take pareidolia, for instance – it's likely been more evolutionary advantageous to see faces where there are none (and thus flee tot often), as opposed to not seeing faces where there are faces (and thus be eaten).
And in evolutionary terms, not everything that exists has some benefit. Some features just haven't been subject to evolutionary pressures; Not selected for, just not selected against. Vestigial features are prime examples of this.
They are only a problem if you trying to use your cognitive power to prove mathematical theorems or discover natural physical laws. Which are not things that our brains are designed for.
In all other situations, they are working as expected.
e.g. risk avoidance is very important because one single mistake in risk assessment (underestimating a real threat) can wipe you out.
