User talk:Vladimir Nesov

Hi, welcome to the LessWrong Wiki, and thanks for your edit to the Talk:Paper clip maximizer page. Please leave a message on my talk page if I can help with anything, and feel free to email if that doesn't work. You might like to read discussion of this wiki on LessWrong itself. -- BJR (Talk) 14:31, 17 April 2009

How do I not do whatever you thought was bad? PotatoDumplings

Try, see, discuss. Last ones were the new kinds of things. If you disagree with my judgment, by all means debate your position (in the talk tabs). I'm just trying to make sure that each edit improves the quality of the wiki, step by step. --Vladimir Nesov 20:43, 9 July 2009 (UTC)

Saying that evolution began after the first replicator arose by chance is accurate, so why remove it? Secondly, saying evolution produced humans "in particular" implies it was somehow more responsible for humans than for anything else. Frost Shock Level 4 20:47, 7 December 2009 (UTC)

Strictly speaking, we can't yet say how the first replicator appeared, so "by chance" is too strong an assertion (what it would even mean, in any case?). "In particular" doesn't bring up this connotation for me, but better rewording is always welcome. I removed the whole edit simply because the result appeared to be worse than the original, even if it implied an additional piece of data (which I'm not sure needs to be mentioned in a stub article of this size anyway). --Vladimir Nesov 22:04, 7 December 2009 (UTC)

Then I'd like to say "including humans" rather than "humans in particular".

Go on and edit the page if you like. --Vladimir Nesov 18:51, 10 December 2009 (UTC)

As for the other thing, it seems to imply that evolution and only evolution is responsible for life, whereas the formation of the first replicator is clearly an important, non-evolutionary process involved right at the beginning... evolution can't start until there's a replicator. Frost Shock Level 4 23:28, 7 December 2009 (UTC)

You'd also need planets, stars and Big Bang... --Vladimir Nesov 18:51, 10 December 2009 (UTC)

I am reading through the sequences for the first time, and was really confused and frustrated as to why there was no link to the sequence "A Human's Guide to Words". I felt very overwhelmed reading the sequence guide 37 Ways That Words Can Be Wrong, but when I searched and found the sequence in the Wiki, and then started clicking on the links one by one, it was much more manageable for my brain. I read Eliezer's comments on the discussion page closely, and it looks like he wants to get people to go straight to 37 Ways That Words Can Be Wrong (because it is the sequence guide and he wants to minimize clicks). I think if that is the way that he wants it, it would be nice to at least add in a link to the sequence guide at the bottom of the section, which could say something like "once you have read the sequence guide, here is a link to the sequence in order of posting" or something like that, so there is a link to the sequence if people want to find it without having to search. --philips999 23:50:59, 8 March 2010 (UTC)

I don't quite understand what you refer to by the difference between clicking links from the 37 Ways guide and from the wiki page -- they are the same links, in more or less the same order... The difference is that the wiki page doesn't include descriptions, is that the feature that makes it more convenient to you? Or the fact that the links on the 37 Ways page are placed at the ends of paragraphs, and some of them are duplicated? What about other, more fleshed out sequence guides on the wiki, do the present similar issues to you? --Vladimir Nesov 06:51, 9 March 2010 (UTC)

Markup

What is standard about that arrangement of sections? Wikipedia with its >3.4 million articles (which I've worked on for 6 years) certainly doesn't do it that way. 'See Also's go before 'References'/'External links'. --Gwern 15:05, 1 November 2010 (UTC)

This structure is standard simply in the sense than most of this wiki is actually formatted this way. It might not be optimal, but it's how it currently is. Blog posts come first, since most of the wiki is a topic map of the blog, organizing the most important posts, with wiki pages themselves holding relatively little content. In this sense, blog posts work as extension of the wiki page content. The main use case is to navigate 'See also' links, find an interesting concept, and go to the referenced blog posts (or non-blog external references). If there is an actual argument behind Wikipedia's way of things, as opposed to just need for consistent formatting within a project, wiki could be reformatted (but this is a decision to apply to all pages). --Vladimir Nesov 15:18, 3 November 2010 (UTC)

Is there some general guideline as to what redirect pages are appropriate or not? It seems pointless for me to create more work for both of us by making things that you're just going to have to delete later, but even in retrospect I'm not sure what was different about the ones you deleted vs the ones you didn't. In particular, for "Ia" I was just following the direction given in Jargon: "To add a new acronym, place it in the list while retaining alphabetical order, and create a redirect page from the acronym to the article with full name, or to this page if creating such article isn't a good idea." More generally, I'm not sure why fewer redirects would be better than more redirects, even if they're not often used - is the number of pages a real concern? --Pedanterrific 03:28, 20 October 2011 (UTC)

Hi, thank you for contributing. I've just removed a few stranger-sounding redirects, for "Ia" in particular it would be fine by me if it was "IA", but "Ia" capitalization is not used, as I understand (I guess I should've renamed). I'm not clear on the general value of rephrasing-only redirects (as opposed to close ideas and different names for same idea), other than commonly used shorthands, since "Search" would lead the way anyway, and you're unlikely to guess a more complicated redirect exactly. --Vladimir Nesov 11:25, 20 October 2011 (UTC)

Um, as to capitalization... whether you type "TL;DR" or "tl;dr" into the search bar, you still get redirected by the page "Tl;dr". So I figured that wouldn't matter. (Though, now that I think of it, the difference between "Ia" and "IA" might be that one should redirect to "Evolution"...) But I get your general point.

Spam accounts

I've getting troubled by RC. We're seeing scores of new accounts registered, all names generated by the same old spam template, but none of them are being used to spam. Was an anti-spam measure enabled or are these accounts just being left to age and will bite us in months to come? --Gwern 04:42, 13 December 2011 (UTC)

Accounts must now use email confirmation (you probably were asked for that if you're posting, unless Trike implemented my suggestion to exempt existing users from this check). So I expect spam account creation process simply hasn't realized yet that they don't work. --Vladimir Nesov 00:16, 14 December 2011 (UTC)

Phew. It's been nice having this break - Haskell wiki spam continues unabated, but that's easy to deal with. --Gwern 01:42, 14 December 2011 (UTC)

Re:Edit History. Understood, I'll be more careful in future. Thanks

paperclippers and stable goals

Kindly explain why an AI that want to turn the universe into paperclips for a nanosecond, and then changes its mind, and acquires another goal, or no goal , would be dangerous.

Otherwise, I will continue to believe that dangerous paperclippers are and are only persistent paperclippers. -- TheAncientGeek, 11 June 2014

I explained in the edit. The standard argument is that goal-preservation is a basic drive (like gaining control over resources), because changing goals is expected to produce consequences not preferred according to current goals. See Omohundro (2008). --Vladimir Nesov (talk) 05:12, 11 June 2014 (MIST)

The desire to preserve goals does not imply the ability to preserve goals. No one knows how to build a goal stable learning .AI,and humans don't show any sign of maintaining .lifelong stable goals either.TheAncientGeek (talk) 23:03, 11 June 2014 (MIST)