In a recent thread, SarahC said:

I'd prefer more posts that aim to teach something the author knows a lot about, as opposed to an insight somebody just thought of. Even something less immediately related to rationality -- I'd love, say, posts on science, or how-to posts, at the epistemic standard of LessWrong.

... so here's the place to float ideas around: is there an area you know a lot about? A topic you've been considering writing about? Here's the place to mention it!

From a poll on what people want to see more of, the most votes went to:

  • Statistics
  • Game Theory
  • Direct advice for young people
  • General cognitive enhancing tools (such as Adderall and N-Back)
  • Information Theory
  • Economics

Some that got less votes:

  • Data visualization
  • (Defence against the) Dark Arts
  • Moral Philosophy (looks like that's being done already)
  • Postmodernism
  • Getting along in an irrational world
  • Existential risks
  • Medicine, Applied Human Biology

... but there are certainly many more things that would be interesting and useful to the community. So what can you teach us?

New to LessWrong?

New Comment
28 comments, sorted by Click to highlight new comments since: Today at 8:06 AM

I always wanted to write a book about "snake oil". I.E. Profitable practices based on misinformation, or in general profitable practices which only benefit one party without crossing a certain threshold of parasitism through force (i.e slavery)

Although the dangerous thought now occurs to me that perhaps people selling weight loss cures have their own form of enslavement in play...

10 karma points if you post about it.

I cannot post yet, not enough karma. Although I suppose in principle the gist goes that a donkey led by the carrot instead of the stick is still a beast of burden.

A lot of consumer goods, especially the day time TV variety add no significant marginal benefit but still are marketed in a way that is exceptionally compelling - I believe that certain people barely have a rational choice to buy the product, its just that attractive. Especially toys. The systems are so well established to make people want things that it seems as if those systems are sometimes the entire value proposition.

Or in a more abstract sense, the way in which car companies have successfully convinced us that "badges" mean anything. I know people who are making big decisions based on something which seems to occupy the same head space as religion in terms of the way it compels through belief, and reinforces through society.

It sounds a bit too stereotypically anti-corporate, but I'm an enthusiastic business student and these things still seem worrisome. The first case provides no value and is straight up snake oil sometimes, and the second case Is the bad kind of value, the one created by pure inescapable and unspoken agreement.

I'll admit this is nothing like being compelled by arms into hard labour, and is not an amazingly well thought out opinion. But it's something.

Lent you 10 points (will take back when you post, as this messes up the karma system). I've gotten suckered by that rather often, so I'm interested in the mechanisms. (Though I'm atypical - I suspect wanting to look normal until the fast-talking salesjerk shuts up and goes away forever isn't a normal drive.) I'm also interested in the limits of consent.

Mm. I'm afraid I still don't have the karma and not able to write at the moment (sleep deprived). However I'm all for expanding the idea, and will calenderise a date within the week to sit down for a write up. Take the points back, I suppose I'll PM you if this goes further.

I'll remeber that limits of consent phrase...

Wait, you can "lend" karma?

You currently have 22, last time I checked the threshold for posts in Main was 20.

By "lent" I mean I upvoted all your comments, including those I normally wouldn't. I'll leave it as such for a week then remove those extra upvotes. Godspeed re: writing.

Edit: 'sbeen over a week, so took points back. You still have enough karma for a post in Main.

[-][anonymous]13y00

Upvoted, seems interesting enough. I'm currently trying to figure out how much of the entertainment I consume is exploitative and/or addictive with little benefit, especially with regards to TV and 'funny' websites, so a discussion on that might be useful.

I'm trying to work up a heuristic for dealing with this. For instance: if 2 minutes after consuming this product there is no benefit ,I mean ZERO benefit, avoid it. Like buying a soft drink over water - second I finish the can, I'll probably forget I ever had the the thing and my body is worse off. Same for most sitcoms in my opinion.

Probably worth noting it gets messy easily. I know I should avoid video games for long term goals, but frankly I savour some victories for a while If the challenge was decent and it makes me happier for some time. But the problem here is I generally won't know if the game was worth it until well after i've committed time and effort.

[-][anonymous]13y80

I can help out with things in the math/statistics vein. I can go over stuff commonly taught in schools but in a more fun way (maybe probability theory but I take requests), stuff that's particularly relevant to LessWrong (an intuitive explanation behind Occam's Razor), or ideas I'm interested in that are more current and rarely found in textbooks (sparse approximation, dictionary learning, diffusion maps).

I've thought about compiling a post or a few about "The Math They Didn't Teach You," as I've found that linear algebra courses don't teach the most useful linear algebra techniques & algorithms, probability classes don't teach the most useful theorems, etc. for very standard computational purposes.

ETA: I can take care of compression.

  • statistical machine learning
  • robotics
  • convex optimization
  • computer vision

I consider myself to have a decent understand of art that might be helpful.

I actually asked myself this question a while ago and started writing a post about it but I don't have the writing skill or focus to actually write something myself.

In descending order of expertise:

  • Cryptography (from a theoretical perspective)
  • Computational Complexity theory (particularly, proof systems)
  • Algorithms / Data Structures
  • Quantum computing (particularly quantum cryptography and quantum complexity)
  • Mathematical logic (including recursion theory)
  • Mechanism design (and the required notions of game theory / economics)
  • Compressed sensing / sketching

What do you hope to get out of learning about cryptography?

I have already learned about cryptography. You mean why did I learn about cryptography? Or why would I offer to teach it?

I don't think its too valuable, but (1) its an aesthetically pleasing field from the theoretical perspective (2) general knowledge of cryptography is likely to lead to better use of cryptography, which is a little valuable (3) general understanding of what it actually means to be paranoid is valuable.

I misread your post and thought it was a request, not an offer - sorry! What sort of theory would you want to cover?

[-][anonymous]13y10

Psychology!

More details would be useful.

I've been considering writing an article about game design and learning; possible topics:

  • Relationship between fun (in games) and learning

  • The psychology of learning (reward, punishment, boredom, challenge)

  • Analysis of these aspect in existing games (in genre evolution, "casual" vs. "hardcore" games, learnable game mechanics vs. story)

  • Game development (importance of iteration, playtesting, willingness to throw bad ideas away)

... though I feel the last two may be a bit off-topic for LessWrong, except maybe in the context of developing a game aimed at teaching specific rationality skills -- I'm not convinced myself that a game is the best approach to teach skills, it's just the one I know the best. Note that the lines are blurry between games and exercises such as Anki and Dual-n-back.

[-][anonymous]13y00

Could people say more about what they hope to get out of learning more about cryptography?

[-][anonymous]13y-30

I would like to see introductory posts covering the following fields:

  • Decision theory
  • Computation
  • Encryption
  • Compression
  • Complexity theory

Additionally I would like to see a post discussing possible conclusions that can be drawn from the Fermi paradox (e.g. how to update on the probability of a Paperclip maximizer given that we are unable to observe any paper-clippers burning the cosmic commons).

Can you do that? Thanks :-)

It may be more convenient if this thread focused on the supply side of the equation, the two threads I linked to already have a fair amount of discussion of what people would want more / less of, i.e. the demand.

I won't promise to write a full post on it, but you might be interested in my pet theory saying that we do not observe any paper-clippers burning the cosmic commons because they are approaching us with the speed of light. I never wrote it up, but there are some short scattered comments about it here and here.

It's pretty unlikely that there are space-faring aliens inside our own galaxy at all. If they were moving at near c they would take around 50,000 [edited] years to get here - next to no time. In which case it would be quite a coincidence that they evolved to the "space-travel" stage almost exactly when we did. So: it is more likely that we are locally first.

Our galaxy's radius is ~50,000 light years, not 50,000,000, which of course strengthens your point.

You are right. In the second thread I linked to it was Nick Tarleton who came up with this obvious counterpoint. But my reply convinced him that the idea is compatible with the existence of many civilizations. Check it out, it is an elegant argument, even if you don't accept the premises.

If they were traveling here directly or expanding in uniform radius at near light speed, yes. But if they're hopping from star to star in a grid network, the distance to travel is much greater. Plus, traveling at near c may not be a practical use of energy at any level of technological sophistication.

I wouldn't be at all surprised if intelligent life, let alone spacefaring life, is rare enough that you would expect to find less than one such species per average galaxy, but I don't think the fact that we haven't been encountered yet is very strong evidence for this.

I wouldn't give up hope on space-faring aliens inside our own galaxy - but the argument still holds pretty well down to c/1,000. Those are some pretty slow-moving aliens. One begins to wonder why they would think they can dawdle around.

Additionally I would like to see a post discussing possible conclusions that can be drawn from the Fermi paradox (e.g. how to update on the probability of a Paperclip maximizer given that we are unable to observe any paper-clippers burning the cosmic commons).

Outside of our own galaxy, I don't think we currently know enough / can see enough to be able to distinguish living galaxies from dead ones. So, not very much to update on there just now.