Crossposted (with small tweaks) from my Substack.

Burdensome Details

• This post explains the conjunction fallacy, i.e., people sometimes assign a higher probability to A and B together than to either A or B alone. This is usually motivated by the fact that adding more details tends to make scenarios seem more plausible—and therefore more ‘believable’—but it is a mathematical truth that the probability of A and B cannot exceed the probability of either A or B, since their conjunction is a subset of both.
• A way to guard against this fallacy is to be alert to conjunctions in propositions. You should automatically assume that every additional ‘and’ can only lower—or leave unchanged—the probability of the proposition. 

Planning Fallacy

• Another fallacy to add to our list! The gist of the planning fallacy is that we are overoptimistic in our estimates of how long our projects will take to complete; in fact, our ‘realistic’ estimates often lie remarkably close to the most optimistic scenarios we can imagine.
• It is relatively easy to debug though: you need to apply an ‘outside view’ that compares the current project with similar past projects and considers how long they took to complete. The extra details of the current situation only obscure the comparison.

Illusion of Transparency: Why No One Understands You

• Hindsight bias is briefly mentioned (once you’ve seen the outcome of an event, you can’t help thinking it was inevitable and easy to foresee before it happened), but the gist of this post deals with a related, but different bias: the illusion of transparency.
• The latter consists in believing that what you mean by what you say will be transparent to people who lack your context and background knowledge. A closely related form of the curse of knowledge appears when, because we know something, we expect others to be aware of it as well and overlook possible ambiguities because, to us, ‘it is obvious!’

Expecting Short Inferential Distances

• This bias doesn’t actually get a name apart from the title of the post. Yud himself says that it is difficult to explain in a few words, but here’s my best shot: we tend to expect that we should easily understand technical explanations—and that others should easily understand ours—after only a few short inferential steps.
• This might have made sense in small hunter-gatherer societies, where there was relatively little specialized technical knowledge and much of what people knew was shared. But once you have complex societies and specialist areas of knowledge, this breaks down, making communication between experts and lay audiences extremely difficult.

The Lens That Sees Its Own Flaws

• This one isn’t a bias, but rather, a useful capacity we possess and a nice way to round off this sequence.
• Humans, as opposed to animals, possess deliberate rationality: we can understand and reflect on the processes through which we form beliefs and construct meaning. This is an exceptional skill. 
• Because we can ‘see how the machinery works’ and reflect on its failings, we can correct the errors we make (be them of perception or of cognition). 

Some Thoughts of Mine

• Not much to say here. What Yud explains seems to be, prima facie, plausible. I’ve encountered these biases in other sources before.
• As a teacher, you get plenty of practice in combating the illusion of transparency because you receive constant, real-time feedback about whether your students have really understood what you have explained, even when it seems trivial to you.
• While reading about inferential distances, what came to mind was how incredibly baffled I get when trying to understand moral realists. Because a lot of people whose intelligence I respect fall into this reference class, I try to read and listen to their arguments, but I have yet to find one that does not strike me as outrageously silly and circular—almost self-evidently false—to the point that it makes me doubt my sanity and/or that of its proponents^[1]. As many of them are philosophers and metaethicists, it could be a case of inferential distance, although I feel it is more that, when we reach the bedrock of our intuitions, mine are fundamentally different from theirs.
• Similarly, The Lens… makes me reflect on one particularly strong belief I hold: i.e., full-blooded mathematical Platonism. While I dislike psychologizing people (including myself), I think this belief is so emotionally important to me because what attracts me to mathematics is its ability to provide the kind of dogmatic, inflexible, certain, and eternal truth that, as a matter of deep temperament, I seem to crave^[2]. That doesn’t mean the belief is false, but it should update me towards being somewhat more critical of it.

Note: I used ChatGPT for minimal proofreading/copyediting - fixing typos, grammar mistakes and suggestions of awkward or unidiomatic phrasing. The text, views, and responsibility for the post are mine.

1. ^{^}

   If you’re reading this and a moral realist, feel free to suggest good arguments or works for me to read. I’d be grateful if you included a short summary of them, because I am not enthusiastic about devoting many hours to reading a book whose argument continues to seem foolish from page 1 to page 300.

2. ^{^}

   You might guess from this—and you’d be right—that I am strongly averse to uncertainty. I accept it as a curse we have to live with, but one that I ultimately loathe. In another universe, I suspect I would have been happy as an Auditor of Reality. This also means I’ve never felt particularly drawn to probability theory or to other areas of mathematics concerned with uncertainty.