Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Comment author: fblogin 23 November 2017 10:19:32AM *  0 points [-]

Yes, the article I was looking for. Your article gives me another approach on the subject. I hope to read more articles from you. http://abcdeya.com

Comment author: arundelo 09 December 2017 05:58:09PM *  0 points [-]


This account has been posting spam since April 2017 (though all of their old comments have been deleted and are visible only on their overview and comments pages).

Comment author: Habryka 16 September 2017 11:35:15PM 5 points [-]

Being aware that this is probably the most bikesheddy thing in this whole discussion, I've actually thought about this a bit.

From skimming a lot of early Eliezer posts, I've seen all three uses "LessWrong", "Lesswrong" and "Less Wrong" and so there isn't a super clear precedent here, though I do agree that "Less Wrong" was used a bit more often.

I personally really like "Less Wrong", because it has two weirdly capitalized words, and I don't like brand names that are two words. It makes it sound too much like it wants to refer to the original meaning of the words, instead of being a pointer towards the brand/organization/online-community, and while one might think that is actually useful, it usually just results in a short state of confusion when I read a sentence that has "Less Wrong" in it, because I just didn't parse it as the correct reference.

I am currently going with "LessWrong" and "LESSWRONG", which is what I am planning to use in the site navigation, logos and other areas of the page. If enough people object I would probably change my mind.

Comment author: arundelo 17 September 2017 08:06:54PM *  6 points [-]

I just used Wei Dai's lesswrong_user script to download Eliezer's posts and comments (excluding, last I knew, those that don't show up on his "OVERVIEW" page e.g. for karma reasons). This went back to late December 2009 before the network connection got dropped.

I counted his uses of "LessWrong" versus "Less Wrong". (Of course I didn't count things such as the domain name "lesswrong.com", the English phrase "less wrong", or derived words like "LessWrongers".)

"LessWrong": 1 2 3* 4*

"Less Wrong": 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20* 21 22* 23 24 25 26

Entries with asterisks appear in both lists. Of his four uses of "LessWrong", three are modifying another word (e.g., "LessWrong hivemind").

(For what it's worth, "LessWrongers": 1 2; "Less Wrongians": 1.)

Comment author: arundelo 15 September 2017 06:51:45PM *  5 points [-]

Has the team explicitly decided to call it "LessWrong" (no space) instead of "Less Wrong" (with a space)?

The spaced version has more precedent behind it. It's used by Eliezer and by most of the static content on lesswrong.com, including the <title> element.

Comment author: DragonGod 20 August 2017 08:10:43AM *  2 points [-]

Which bets did he make against academia?

When I eventually become an AI researcher, I do plan to try another approach apart from Neural networks though (I have an idea I think might work, and enough people are already trying their hands at neural nets).

I agree that I do find Eliezer's overconfidence endearing.

Comment author: arundelo 25 August 2017 01:39:29PM 0 points [-]


Comment formatting note -- Less Wrong's subset of Markdown does not let you use inline HTML.

Comment author: Dagon 24 August 2017 09:25:05PM 0 points [-]

Wait. If you're talking about surprise because you have said "update your model based on how surprised you are", you can't turn around and say "surprise is defined by how much you should update your model". "update your model based on how much you should update your model" isn't very helpful.

Comment author: arundelo 25 August 2017 06:31:36AM 0 points [-]

The intuitive sense of what surprise is corresponds well to the rules for updating your probability distribution over models, which we can therefore take as a formal definition of surprise.

Comment author: cousin_it 23 August 2017 01:34:22PM *  7 points [-]

Here's an old puzzle:

Alice: How can we formalize the idea of "surprise"?

Bob: I think surprise is seeing an event of low probability.

Alice: This morning I saw a car whose license plate said 3817, and that didn't surprise me at all!

Bob: Huh.

For everyone still wondering about that, here's the correct answer! The numerical measure of surprise is information gain (Kullback-Leibler divergence) from your prior to your posterior over models after updating on the data. That gives the intuitive answer to the above puzzle, as long as none of your models assigned high probability to 3817 in advance. It also works for the opposite case, if you expected an ordered string but got a random one, or ordered in a different way.

This is actually well known, I just wanted to put it on LW.

Comment author: arundelo 23 August 2017 05:04:49PM 1 point [-]

Just to make sure I understand prior and posterior over models, is the following about right?

  • Alice starts with a prior of 0.999 that non-vanity plates are generated basically randomly (according to some rule of "N letters followed by M digits" or whatever, and with rules e.g. preventing swear words).
  • Alice sees "3817" (having seen many other 4-digit plates previously).
  • Alice's posterior probability over models is still about 0.999 on the same model.
Comment author: MaryCh 22 July 2017 04:14:40PM 0 points [-]

Why 1/3?

Comment author: arundelo 25 July 2017 10:33:42AM *  1 point [-]

Families with exactly two children:

| oldest | youngest |
| boy... | boy..... | two boys
| boy... | girl.... | one boy
| girl.. | boy..... | one boy
| girl.. | girl.... | no boys
Comment author: fmgn 19 July 2017 01:43:00AM 0 points [-]

And then your opponent still steals, gets all the money, and nothing goes to medical research. woopty doo.

Comment author: arundelo 19 July 2017 04:35:57AM *  0 points [-]

No, Romeo chooses steal. If his opponent also chooses steal (in spite of Romeo's credible commitment to choosing steal himself), the opponent does not get any money.

Comment author: arundelo 10 July 2017 02:07:33PM 1 point [-]

This is probably a known issue, and I know a rewritten version of the Less Wrong software is being worked on, but I just noticed that even if I'm using HTTPS, comment permalinks (the chain icon at the bottom of a comment) are HTTP URLs.

Comment author: arundelo 09 July 2017 03:01:31AM *  0 points [-]

archive.is has both things from Patri's LiveJournal:

(Unlike archive.org, archive.is does not, IIRC, respect robots.txt.)

Gwern Branwen has a page on link rot and URL archiving.

Comment author: arundelo 09 July 2017 06:24:27PM *  0 points [-]

Why does archive.is not obey robots.txt?

Because it is not a free-walking crawler, it saves only one page acting as a direct agent of the human user.

--archive.is faq

A few months ago we stopped referring to robots.txt files on U.S. government and military web sites [...] As we have moved towards broader access it has not caused problems, which we take as a good sign. We are now looking to do this more broadly.

--archive.org blog, 2017-04-17

View more: Next