A Human's Guide to Words (2010) | Hacker News Reader

A Human's Guide to Words (2010)

wiki.lesswrong.com

109 pointsSymmetry15 comments

While there's a lot of value in the rationalist literature, there's also a lot of, well, words. The standard "sequences" take up nearly a million words, and the introductory fanfic ("prereq material") spans 2/3 of a million. The combined literature is longer than War and Peace plus the entire Bible plus my bloated summary of most of physics, combined.

People with technical backgrounds (scientists, engineers, etc.) have already internalized many of the key ideas as part of their education. It should be possible to convey 99% of the remaining value to us with only 1% of the wordcount. Has anybody worked on making such a resource?

Looking at the first line, "37 Ways That Words Can Be Wrong", one wonders to what degree these are not logical fallacies enjoying a tryst with Zen loans.

It only takes a few pages to describe how to be rational. Cataloging the many ways people tend to be irrational, so they can detect and defeat their own biases, creates most of the word count. "Less Wrong" is apt, because it's organized around how people tend to be wrong.

Once you have nearly a million words about what's wrong with people, I suppose it needs a little narrative sauce or it'd be massively discouraging.

This particular collection of blog posts is IMO the most valuable part of that canon. It does have some redundancy, since some points are worth belaboring -- but if you'd like to skip that and just read the parts that interest you, there's an introductory post with one-paragraph summaries of each one:

https://www.lesswrong.com/posts/FaJaCgqBKphrDzDSj/37-ways-th...

I agree that this is the high point of the Sequences.

In my view, the two most important ideas are EY's idea about what the two alternate neural architectures "look like from the inside" for concepts, and the combinatorio-spatial metaphor for alternative regimes of concepts.

To me, David Chapman's writing under Meaningness is somewhere in that direction: https://meaningness.com

All maps are wrong; the map is never the territory. All statements are hedged. All signs can do is point, never indicate.

People are not special. People are humans; humans are primates; primates are animals.

If genes are like blueprints for constructing life-supporting structures, then memes are like blueprints for thinking viral ideas.

All computation is symbolic.

...That should be enough to let people skip all of the philosophy, at a first pass. The main problem is that folks don't actually absorb these slogans on their own; the slogans require support from a library of lore and learnings. It is obvious, once one's mind has the right scaffolding, to see that everything and everybody is wrong about everything, but actually setting up that scaffolding is unique to every different reader.

I've found that I, too, need to write a fictional story set inside a fictional universe akin to ours, if I want to communicate certain non-fictional abstract mathematical ideas. People cannot grasp some concepts without some sort of path which leads to the concept, and humanity paves those paths with words.

People cannot grasp some concepts without some sort of path which leads to the concept, and humanity paves those paths with words. -- beautifully spoken.

Ah, the Parable of the Dagger. Most people wouldn't be surprised if, when looking into a shed with word "S * * T" written on it in large letters, they'd see a shovel, a garden rake, a couple of empty buckets, and a broken wheelbarrow but no feces. Yet, write "This shed is full of manure. The previous sentence and this sentence either are both true, or both false" on the shed, and they apparently will be surprised to find garden tools inside.

Isn't LessWrong a pseudo-intellectual personality cult with lots of big words but little substance? I wonder how one can produce this much output tho.

It is certainly not pseudo-intellectual: people with impeccable academic credentials like David Chalmers and Stuart Russell take it seriously.

You've responded to a loaded question [1] with an argument from authority. [2]

In this circumstance I would recommend not engaging at all.

[1] https://en.wikipedia.org/wiki/Loaded_question

[2] https://en.wikipedia.org/wiki/Argument_from_authority

Calling out fallacies too early should be a fallacy by itself.

The parent is not settling matters with a fallacy, he's providing branches we can investigate to judge if Less Wrong is worth reading. Should we censor facts because they might be classified as argument from authority?

Just a bit long-winded, I think. It probably boils down into some useful stuff, but I'm not entirely sure because the here-we-go-again prose pushes me away. It's not crankery, certainly.

Yeah I agree. I think I feel the same about their work actually.. Calling it "pseudo-intellectual" is too provocative on my part.

Reply on news.ycombinator.com