Chat Logs/2009-04-11

From Lesswrongwiki
Jump to: navigation, search

The following is a mostly-complete, slightly-edited log of the Less Wrong IRC chat from April 11, 2009.

See Chat logs for a list of other chat logs.

  • Session Start: Sat Apr 11 15:15:47 2009
  • Session Ident: #lesswrong
  • [11:32] Topic is 'Discussion on the topics of lesswrong.com and overcomingbias.com?'
  • [11:32] Set by ciphergoth on Thu Apr 02 17:08:44
  • [11:33] timewarp: Is that rational? :)
  • [11:34] Xuenay was irrational enough to go play online Civ4 FfH instead of waiting for the meetup
  • [11:40] MosaSaur: how long till the meetup?
  • [11:41] jimrandomh: 2h20m
  • [11:41] MosaSaur: thank jim, I forgot there was one but now I am counting the minutes ...
  • [11:50] timewarp: So what happens? Body shots? Strippers?
  • [11:52] ErikMesoy: Mind-altering substances. For a sufficiently broad definition of "substances". ;)
  • [11:54] Grey_Fox: religious level broadness
  • [11:54] Grey_Fox: :p
  • [11:57] pengvado has joined #lesswrong
  • [12:04] MosaSaur: vacuum pumps
  • [12:04] Grey_Fox: so timewarp
  • [12:04] Grey_Fox: I gotta teach you computer science
  • [12:04] Grey_Fox: or someone computer science
  • [12:04] Grey_Fox: if I don't, I'll die!
  • [12:05] AdeleneDawner has joined #lesswrong
  • [12:05] jimrandomh: Show of hands: how many computer scientists or programmers, and how many aren't?
  • [12:05] jimrandomh has a computer science degree
  • [12:05] ErikMesoy: Grey_Fox: Do you mean "I need to talk to someone about computer science"? Because I could pretend to listen and nod from time to time.
  • [12:06] ErikMesoy has taken programming courses but doesn't have a degree.
  • [12:06] Grey_Fox: lol
  • [12:06] Grey_Fox is undergoing a degree
  • [12:07] MosaSaur: python programmer here, started out as a social psychologist
  • [12:07] PeerInfinity: hi Adelene :)
  • [12:07] jimrandomh: Any non-programmers?
  • [12:07] PeerInfinity is a programmer :)
  • [12:07] AdeleneDawner: Hiay Peer. ^.^ *hug*
  • [12:07] PeerInfinity: *hugs* :D
  • [12:07] AdeleneDawner codes.
  • [12:08] PeerInfinity has a Computer Science degree :)
  • [12:09] jimrandomh: Ok, I feel a sudden need to one-up
  • [12:09] jimrandomh has *two* Computer Science degrees
  • [12:10] jimrandomh: Hmm
  • [12:10] Grey_Fox: lol
  • [12:10] PeerInfinity has a Joint Honours degree in Computer Science and Mathematics, actually :)
  • [12:10] jimrandomh: Maybe Less Wrong should have some articles about applying rationality to debugging and programming
  • [12:10] ErikMesoy: Possibly more interesting question: What are there more of hanging around LW, theists or non-(computer scientists/programmers)?
  • [12:11] jimrandomh: I suspect that active posters and lurkers are demographically different
  • [12:11] jimrandomh: that complicates things
  • [12:11] MosaSaur: to me, degrees signal slave mentality
  • [12:13] PeerInfinity is actually kinda embarassed to have wasted 4 years getting a degree, rather than just going ahead and learning and doing programming on my
  • own...
  • [12:14] AdeleneDawner chuckles.
  • [12:18] timewarp is not a programmer
  • [12:19] Xuenay: I'm minoring in CompSci
  • [12:19] Xuenay: And thinking about switching majors to CompSci once I have my Bachelor's
  • [12:20] Grey_Fox: any lisp fans here?
  • [12:20] Xuenay: (Currently cognitive science)
  • [12:20] jimrandomh: Tried it briefly, hated it
  • [12:21] jimrandomh: When you're skimming quickly, it all looks the same; you can't meaningfully syntax highlight it
  • [12:22] PeerInfinity wonders if I can count myself as a LISP fan, despite not actually having used LISP for anything yet...
  • [12:22] MosaSaur: In my days, computer science was still called informatics and thought to be a subdiscipline of linguistics or psychology, only when the physics
  • guys took over it became known as computer science
  • [12:23] PeerInfinity: hehe... "computer psychology" :)
  • [12:23] Grey_Fox: PeerInfinity: well, you have to know it
  • [12:24] PeerInfinity: would you believe that back in high school, I practically, um... reinvented LISP? this was before I had access to the internet. I got as
  • far as implementing a mostly-functional interpreter for a new language I invented, called StrandC, for lack of a better name...
  • [12:25] PeerInfinity: StrandC was basically a horribly inefficient implementation of LISP :P
  • [12:26] PeerInfinity: I created this language as a subproject of my misguided attempt to create an AI...
  • [12:27] PeerInfinity: I ended up abandoning the project after I finally found out that LISP had already been around for decades, and lots of people had used it to
  • try to make their own AIs :P
  • [12:28] Grey_Fox: lisp is pretty cool
  • [12:28] PeerInfinity: I spent a few hours reading about how LISP actually works, but I never got around to actually using it for anything...
  • [12:28] Grey_Fox: I'm checking out joy
  • [12:28] PeerInfinity: and yeah, LISP is awesome :D
  • [12:29] jimrandomh: ...until you need to link against a library
  • [12:29] PeerInfinity: :P
  • [12:33] stillflame has joined #lesswrong
  • [12:44] Grey_Fox: (defun eval (x) x)
  • [12:44] Grey_Fox: :p
  • [12:45] ErikMesoy: I'm not familiar with Lisp. That looks as though it returns the input. What does it actually do?
  • [12:45] Grey_Fox: returns the input
  • [12:50] MosaSaur: import gravity
  • [12:51] stillflame: from _future_ import mind.unbiased, reality.safe
  • [12:51] Grey_Fox: (defun primrec (test base acc step) (labels ((iter (x) (if test base (acc (iter (step x))))))))
  • [12:52] Grey_Fox: (primrec #'(lambda (x) (<= x 0)) 1 #'* #'-1) <- factorial
  • [12:53] Grey_Fox: of course, this particular primrec is inefficient as sin
  • [12:53] Grey_Fox: but that's the power of lisp
  • [12:53] Grey_Fox: abstracting away control flow
  • [12:54] Grey_Fox: of course most good languages can do this nowadays
  • [12:54] Grey_Fox: they can't do macros though
  • [12:55] Grey_Fox: oh, silly me
  • [12:56] Grey_Fox: (defun primrec (test base acc step) (labels ((iter (x) (if test base (acc (iter (step x)))))) #'iter))
  • [12:56] Grey_Fox: now the code actually works
  • [12:58] ErikMesoy hums the Final Countdown tune. 1 hour until the black-hole-powered hoover of time and life force is activated.
  • [13:00] chronophasiac has joined #lesswrong
  • [13:00] PeerInfinity: hi chrono :D
  • [13:01] chronophasiac: hullo
  • [13:01] PeerInfinity: thanks for setting up the Less Wrong wiki :)
  • [13:01] chronophasiac: oh, that's not me
  • [13:01] chronophasiac: i think you are thinking of ChronoDAS
  • [13:01] PeerInfinity: oops, that was Cyphergoth...
  • [13:02] PeerInfinity: silly me and my habit of only looking at the first letter of a person's name :P
  • [13:02] chronophasiac has left #lesswrong
  • [13:03] chronophasiac has joined #lesswrong
  • [13:03] chronophasiac: ...I haven't used IRC in a while
  • [13:04] PeerInfinity suggests "how we can use the wiki" as one of the main topics for the meeting...
  • [13:06] Grey_Fox: yeah
  • [13:07] MosaSaur suggests unnamable topics
  • [13:09] Grey_Fox: lol
  • [13:09] Grey_Fox: do you know who's coming to the meeting?
  • [13:09] ErikMesoy wonders which those are. (Seriously: the Dark Arts or the Institute Which Must Not Be Named?)
  • [13:10] MosaSaur: if it's xenu I just failed
  • [13:11] chronophasiac: related question: in this chat, shall we uphold the moratorium on transhumanist-related topics?
  • [13:11] Grey_Fox: might as well
  • [13:12] jimrandomh: I suspect more than half of the people coming are already here
  • [13:12] jimrandomh: the number of people in this channel has been steadily rising all day
  • [13:13] PeerInfinity: another obvious topic is "how we can use these weekly chat meetings"
  • [13:14] PeerInfinity: and, of course, "should these meetings be weekly? at what time?"
  • [13:14] jimrandomh: If weekly, I think it should be the same time as this one every week
  • [13:14] jimrandomh: so people can keep track of it
  • [13:15] jimrandomh: the alternative is monthly
  • [13:15] Grey_Fox: I don't
  • [13:15] jimrandomh: weekly can turn into a time sink
  • [13:16] MosaSaur suggests the topic to be "how to avoid meta-topics"
  • [13:17] PeerInfinity: but... it's traditional for the first chat meeting to consist mostly of meta-topics...
  • [13:17] jimrandomh: This isn't a first meeting
  • [13:17] jimrandomh: that would be the thread on the site
  • [13:18] jimrandomh: and that's full of meta-topics, rightly
  • [13:18] PeerInfinity: oh
  • [13:18] jimrandomh: (Actually, my only reason for choosing to see it that way is because that perspective suppresses discussion of meta-topics. But that was the
  • goal anyways.)
  • [13:19] MosaSaur: That topic would be a paradox
  • [13:20] Grey_Fox: anyway
  • [13:21] Grey_Fox: we've got 40 minutes
  • [13:21] Grey_Fox: right?
  • [13:21] stillflame: confirmed.
  • [13:21] jimrandomh: yeah
  • [13:21] jimrandomh: but that doesn't mean we can't discuss things before the official 'start of discussion' bell rings
  • [13:22] Grey_Fox: ok
  • [13:22] MosaSaur: in fact now would be the time to assign some people to their contrapositions
  • [13:24] Grey_Fox: ???
  • [13:25] stillflame: i expect we would need more clarification of the irc-specific rules for playing Paranoid Debating right off the bat.
  • [13:25] stillflame: s/ for / before /
  • [13:25] jimrandomh: Does anyone have Paranoid Debating questions handy?
  • [13:27] Grey_Fox: what about playing mafia?
  • [13:28] jimrandomh: That would definitely require a bot
  • [13:28] jimrandomh: otherwise it'd be logistically unmanageable
  • [13:28] stillflame looks around for a mafia bot
  • [13:29] jimrandomh: How about a vote tallying bot?
  • [13:29] ErikMesoy: I think getting a bot and what functions it should have is worth a topic in itself.
  • [13:29] jimrandomh: So we can call aye/nay votes for official stuff like future meeting times without cluttering the chat
  • [13:30] Grey_Fox: yeah
  • [13:31] Grey_Fox: jimrandomh: we could take turns adminning
  • [13:32] Grey_Fox: find a mafia bot
  • [13:32] Grey_Fox: http://code.google.com/p/erkbot/
  • [13:33] Grey_Fox: could someone get that up and running?
  • [13:37] Erkbot has joined #lesswrong
  • [13:38] jimrandomh: That's mine
  • [13:38] jimrandomh: Not sure how it works yet
  • [13:38] stillflame: Erkbot: help
  • [13:38] ErikMesoy: !help
  • [13:38] jimrandomh: >>help
  • [13:38] Erkbot: Commands: checkmode settings nicks create list channels destroy
  • [13:38] jimrandomh: >>help settings
  • [13:38] Erkbot: Debug. Shows IRC server settings, optional argument will list setting value.
  • [13:39] jimrandomh: >>checkmode
  • [13:39] Erkbot: No mode matches in #lesswrong
  • [13:39] jimrandomh: >>list
  • [13:39] Erkbot: There are no games.
  • [13:39] jimrandomh: >>create
  • [13:39] Erkbot: Game game1 created. Join #lesswrong_game1 to play.
  • [13:39] stillflame: bah
  • [13:40] jimrandomh: We need a few more for Werewolf
  • [13:40] jimrandomh: We have 3
  • [13:42] jimrandomh: Hmm
  • [13:42] jimrandomh: The Werewolf bot didn't do at all what I expected when I started the game
  • [13:42] jimrandomh: In fact, it publically announced the roles it assigned to everyone
  • [13:42] Grey_Fox: yeah
  • [13:42] timewarp has quit IRC ("Leaving"?)
  • [13:42] Grey_Fox: that's super crazy
  • [13:43] timewarp has joined #lesswrong
  • [13:43] Grey_Fox: hi timewarp
  • [13:43] Grey_Fox: so jimrandomh can you fix it?
  • [13:43] timewarp: hey
  • [13:44] jimrandomh: Reading its source code now
  • [13:44] jimrandomh: It doesn't really have any documentation
  • [13:44] jimrandomh: >>help modes
  • [13:44] Erkbot: Invalid help topic.
  • [13:45] Grey_Fox: well, I've found another bot for the game
  • [13:46] Erkbot has quit IRC (Remote closed the connection?)
  • [13:46] jimrandomh has quit IRC (Client Quit?)
  • [13:46] jimrandomh has joined #lesswrong
  • [13:47] Grey_Fox: jimrandomh: if the bot is too much trouble, there's another one
  • [13:47] Grey_Fox: http://code.google.com/p/wolfbot-pl/
  • [13:51] EnglishGent has joined #lesswrong
  • [13:51] EnglishGent: hello :)
  • [13:51] jimrandomh: Hello
  • [13:52] EnglishGent: hi jimrandomh
  • [13:52] LeopoldTal has joined #lesswrong
  • [13:52] LeopoldTal: Hello.
  • [13:52] stillflame: Welcome.
  • [13:52] EnglishGent: hi LeopoldTal, stillflame :)
  • [13:53] jimrandomh: Currently installing dependencies for wolfbot-pl
  • [13:53] Grey_Fox: cool
  • [13:53] Grey_Fox: hopefully this gives us less trouble
  • [13:53] jimrandomh: Does anyone have questions suitable for paranoid debating to try?
  • [13:54] Grey_Fox: ideally, questions that can't be solved with a quick google
  • [13:54] EliezerYudkowsky has joined #lesswrong
  • [13:54] stillflame: which state/country has the smallest difference between highest point and lowest point?
  • [13:54] EliezerYudkowsky is now known as EliezerIsWatchin
  • [13:54] stillflame: i don't know how googleable that is
  • [13:55] EliezerIsWatchin is now known as EliezerWatches
  • [13:55] PeerInfinity wonders what to do now that Eliezer has entered... do I bow?
  • [13:55] Grey_Fox waves to EliezerWatches
  • [13:55] vnesov double-checks the cord of the Soul-Sucking Black Hole-Powered Hoover Vacuum Cleaner in preparation to turning it on.
  • [13:56] Yvain has joined #lesswrong
  • [13:56] jimrandomh: The second werewolf bot doesn't work, either
  • [13:56] Grey_Fox: hi Yvain
  • [13:56] Grey_Fox: oh come on!
  • [13:56] jimrandomh: Compilation error. Might be a Perl version issue
  • [13:56] MattSimpson has joined #lesswrong
  • [13:56] Grey_Fox: EnglishGent: could you get this perl script working?
  • [13:56] EnglishGent: what perl script?
  • [13:56] Grey_Fox: http://code.google.com/p/wolfbot-pl/
  • [13:56] EnglishGent has a look
  • [13:57] rwallace has joined #lesswrong
  • [13:57] PeerInfinity decides to ressurect an old meme, and says: "Hail Ilpalazzo-sama" :)
  • [13:57] EnglishGent: I try to avoid perl you know Grey_Fox....
  • [13:57] Grey_Fox: you don't actually have to look at it
  • [13:57] EnglishGent: I dont believe in immortal souls... but somehow it still makes me fear for mine :P
  • [13:57] EnglishGent: :)
  • [13:57] rwallace: Hi all! ah yes, Excel Saga, classic :)
  • [13:57] Grey_Fox: just point it to irc
  • [13:58] vnesov: So, Omega Watch was an omen.
  • [13:58] vnesov waves EliezerWatches
  • [13:58] Erkbot has joined #lesswrong
  • [13:58] Yvain: Ah! Got it. Hi everyone.
  • [13:58] dfranke has joined #lesswrong
  • [13:58] jimrandomh: I think Erkbot (the first Werewolf bot) may be worth another try
  • [13:58] Grey_Fox: alright
  • [13:58] jimrandomh: >>create
  • [13:58] Erkbot: Game game1 created. Join #lesswrong_game1 to play.
  • [13:59] Grey_Fox: so, we're opening up the meeting with a game of mafia
  • [13:59] Infotropism has joined #lesswrong
  • [13:59] vnesov turns on the Soul-Sucking Black Hole-Powered Hoover Vacuum Cleaner.
  • [14:00] Wolfbot has joined #lesswrong
  • [14:01] stillflame: UTC1900!
  • [14:01] jimrandomh: Argh
  • [14:01] Grey_Fox: same bug as last time
  • [14:01] jimrandomh: I thought I'd fixed it wouldn't do that
  • [14:01] jimrandomh: Oh well. Forget about Werewolf, then
  • [14:01] EliezerWatches chants, "The hour is come, but not the man."
  • [14:01] stillflame: wolfbot is the perl version - reading the source now.
  • [14:01] Grey_Fox: aww
  • [14:01] Grey_Fox: I'm happy to preform the function of the bot
  • [14:02] Grey_Fox: a *working* bot
  • [14:02] ErikMesoy: stillflame: 1900?
  • [14:02] stillflame: 7pm
  • [14:02] Grey_Fox: if that's an issue
  • [14:03] stillflame: !help
  • [14:03] -Wolfbot:#lesswrong- !start --- Begins a game of Werewolf.
  • [14:03] -Wolfbot:#lesswrong- !daystart / !nightstart --- Force day/night start, as opposed to the default behavior of randomly choosing one (using again returns to
  • default behavior).
  • [14:03] -Wolfbot:#lesswrong- !randomroles --- Randomize which special roles are present in the game (though roles disabled via the next command stay disabled).
  • [14:03] -Wolfbot:#lesswrong- !toggle<role> --- Turn a role on or off (note: Village Idiot is simply "Idiot", some roles cannot be toggled).
  • [14:03] -Wolfbot:#lesswrong- !rolelist --- Lists all the special roles
  • [14:03] jimrandomh: Ah, you got the other one running
  • [14:03] Grey_Fox: awesome
  • [14:03] jimrandomh: !rolelist
  • [14:03] -Wolfbot:#lesswrong- Villager --- Default player. No special abilities. Non-toggleable.
  • [14:03] -Wolfbot:#lesswrong- Wolf --- Kills one other player each night. Non-toggleable.
  • [14:03] -Wolfbot:#lesswrong- Seer --- Can determine the role of one other player each night.
  • [14:03] -Wolfbot:#lesswrong- Angel --- Can protect one other player each night from attack by the Wolves.
  • [14:03] -Wolfbot:#lesswrong- Doppleganger --- Can duplicate another player's role, but only once.
  • [14:03] -Wolfbot:#lesswrong- Village Idiot --- Can only win by getting lynched.
  • [14:03] -Wolfbot:#lesswrong- Finder --- Finds the target of another player's abilities.
  • [14:04] jimrandomh: !start
  • [14:04] -Wolfbot:#lesswrong- {jimrandomh} has started a game of werewolf!
  • [14:04] -Wolfbot:#lesswrong- A game of Werewolf has begun! You have {60} seconds to register. You can register by typing "/msg Wolfbot join" and you will be added
  • to the game. You only need to send the message once, but don't worry if you accidentally send it twice, it won't cause the bot to crash or anything.
  • [14:04] -Wolfbot:#lesswrong- {dfranke} has joined the game!
  • [14:04] -Wolfbot:#lesswrong- {stillflame} has joined the game!
  • [14:04] steven0461 has joined #lesswrong
  • [14:04] -Wolfbot:#lesswrong- {EliezerWatches} has joined the game!
  • [14:04] -Wolfbot:#lesswrong- {Grey_Fox} has joined the game!
  • [14:04] -Wolfbot:#lesswrong- {rwallace} has joined the game!
  • [14:04] -Wolfbot:#lesswrong- {Evercat} has joined the game!
  • [14:04] -Wolfbot:#lesswrong- {chronophasiac} has joined the game!
  • [14:04] -Wolfbot:#lesswrong- {LeopoldTal} has joined the game!
  • [14:04] -Wolfbot:#lesswrong- {Infotropism} has joined the game!
  • [14:04] -Wolfbot:#lesswrong- {vnesov} has joined the game!
  • [14:04] stillflame: steven0461: type "/msg Wolfbot join" if you want in. quick.
  • [14:04] -Wolfbot:#lesswrong- {steven0461} has joined the game!
  • [14:04] saturnnn has joined #lesswrong
  • [14:05] -Wolfbot:#lesswrong- Registration has closed.
  • [14:05] Wolfbot has quit IRC (Excess Flood?)
  • [14:05] jimrandomh: !toggle Idiot
  • [14:05] steven0461: thanks
  • [14:05] stillflame: awww!
  • [14:05] jimrandomh: !daystart
  • [14:05] chronophasiac: he's dead, jim
  • [14:05] stillflame: well, not quite a winner there (it flooded itself out of the channel)
  • [14:05] jimrandomh: Hmm
  • [14:05] timewarp: lol
  • [14:05] Grey_Fox: lol
  • [14:05] EliezerWatches: So we're playing Werewolf... and the game itself commits *suicide*.
  • [14:06] jimrandomh: Well, we don't actually need to play games
  • [14:06] jimrandomh: We could have actual discussions
  • [14:06] Grey_Fox: that's no fun
  • [14:06] EnglishGent: a strange game...
  • [14:06] EnglishGent: the only winning move - is not to play
  • [14:06] stillflame will put in flood controls while we discuss things
  • [14:06] Grey_Fox: or to manipulate everyone around you
  • [14:06] Grey_Fox: ok, well items for discussion
  • [14:06] Grey_Fox: the wiki
  • [14:07] Grey_Fox: I'd like to get kind of a syllabus going
  • [14:07] jimrandomh: There are two types of articles: biases/techniques, and examples
  • [14:07] LeopoldTal: Paranoid debating doesn't require a bot, and there are fewer occasions to play it than Werewolf.
  • [14:07] PeerInfinity: and definitions
  • [14:07] dfranke: can someone point me to an explanation of the rules of Werewolf? I think I know the general idea but I've never played it.
  • [14:07] stillflame: i've heard topic suggestions so far: avoiding meta-topics, using the wiki, irc meeting times, Paranoid Debating bot creation
  • [14:07] jimrandomh: So it seems like the wiki should naturally be bipartite
  • [14:07] vnesov: Concepts.
  • [14:07] Grey_Fox: http://en.wikipedia.org/wiki/Mafia_(party_game)
  • [14:08] jimrandomh: Right, we'll hit all those topics, but one at a time
  • [14:08] Grey_Fox: ok
  • [14:08] jimrandomh: The wiki first, since it was mentioned first
  • [14:08] Grey_Fox: yeah
  • [14:08] MosaSaur: where is the current wiki?
  • [14:09] PeerInfinity: http://lesswrong.wikia.com/wiki/LessWrong_Wiki
  • [14:09] jimrandomh: I think that if we took the current OB and Less Wrong, and added even more links to them, they'd make a great wiki
  • [14:09] vnesov: Or rather http://lesswrong.wikia.com
  • [14:09] dfranke: Heh! Andrew Plotkin came up with the 'Werewolf' theme?
  • [14:09] EliezerWatches is now known as Eliezer
  • [14:09] jimrandomh: Eliezer's posts have lots of links, but only to older articles. We could add links from old articles to new ones, and fill in links in other
  • authors' articles
  • [14:10] dfranke: I've played all his text adventures, and I ran into him at the MIT mystery hunt.
  • [14:10] Eliezer: I've also been thinking about the wiki thing. It seems to me that we have an opportunity to do something new here, a blog+wiki format.
  • [14:10] Infotropism: integrated; side by side ?
  • [14:10] jimrandomh: In particular, the articles on biases are almost all older than the examples that refer to them, but they'd be much more useful if they had an
  • appendix full of links to examples
  • [14:11] Eliezer: Wiki articles are short. For any extended discussions they link to blog posts. It's a general rule that blog posts don't go in the Wiki. Wikis
  • name concepts, give quick summaries, possibly link to a few other Wiki articles, and mostly link to blog posts.
  • [14:11] PeerInfinity: (side-topic: acceleratingfuture.com is experimenting with a blog+forum+wiki format)
  • [14:11] vnesov: When someone wants to reference an old article, it is an opportunity to add a little bit to an existing wiki article, and link to it.
  • [14:11] Eliezer: Blog posts link ubiquitously to Wiki pages. Every instance of "epistemic rationality" for example would get wikified.
  • [14:11] vnesov: I suggest encouraging linking to wiki, as opposed to the older articles.
  • [14:12] Eliezer: This would be my shot at making LW as addictive as TVTropes.
  • [14:12] Grey_Fox: so, so addictive
  • [14:12] Roland_LW has joined #lesswrong
  • [14:12] Roland_LW: Hi
  • [14:12] Grey_Fox: what's the magic formula of TVTropes?
  • [14:13] vnesov: Wiki pages should give summaries, then extended summaries, and then they grow into the full-length articles.
  • [14:13] EnglishGent: hello Roland_LW
  • [14:13] Eliezer: the ubiquitous linkage and the interaction between jumping from trope to show to trope
  • [14:13] saturnnn: Grey_Fox, huge amounts of interlinked content I'd say
  • [14:13] jimrandomh: TVTropes has just enough content that whenever you read an article, you'll always find links to something you haven't seen before
  • [14:13] Grey_Fox: yeah
  • [14:13] Grey_Fox: plus the humour
  • [14:14] EnglishGent: hmmm... what about wikipedia - doesnt that have huge amounts of interlinked material?
  • [14:14] Grey_Fox: or at least some sort of sense that there are real people writing
  • [14:14] LeopoldTal: IAWYC but linking to everything in the wiki from so much material would create inertia as changes in the wiki would create discrepancies
  • between the new version and the blog posts/comments about the old version.
  • [14:14] rwallace: "the interaction between jumping from trope to show to trope" is critical
  • [14:14] Grey_Fox: EnglishGent: well, wikipedia isn't as addictive as tvtropes
  • [14:14] Grey_Fox: or for that matter, the c2 wiki
  • [14:14] Xuenay: http://xkcd.com/214/
  • [14:14] saturnnn: some people find wikipedia addictive, the dry style puts a damper on it though
  • [14:14] rwallace: in other words, examples matter
  • [14:14] EnglishGent: surely that is a matter of perspective Grey_Fox?
  • [14:15] jimrandomh: How about if instead of a separate wiki, we make the blog more wiki-like
  • [14:15] Grey_Fox: I agree wikipedia is an invaluable resource
  • [14:15] jimrandomh: in particular, article authors should have the ability to allow future authors to modify their posts
  • [14:15] LeopoldTal: Examples don't matter that much - I can't stop reading TVTropes but if I shouldn't waste much time I can skip the examples.
  • [14:15] Grey_Fox: but I've never spent hours at a time getting lost in its comforting folds
  • [14:15] EnglishGent feels that the xkcd Xuenay linked to illustrates the problem nicely
  • [14:15] Eliezer: Jim, due to limited development resources that would be a far-Future update.
  • [14:16] saturnnn: i would guess the number of links per word is more than an order of magnitude lower on wikipedia
  • [14:16] Eliezer: At least if we want Wikipedian change control.
  • [14:16] steven0461: I like the stream of consciousness style of the original C2 wiki
  • [14:16] rwallace: Hmm. Different tastes, then. For me, tvtropes would be dust-dry without the examples.
  • [14:16] jimrandomh: Otherwise, we have to duplicate just about everything between the blog and the wiki
  • [14:16] Eliezer: s /Wikipedian/Wikian/
  • [14:17] jimrandomh: the only reason it seems like there's content to put in the wiki that isn't also on the blog, is because of all the old OB blog content that
  • needs to be adapted
  • [14:17] vnesov: No, blog posts should be personal, but what stable concepts they introduce/develop should be gradually integrated in the wiki format.
  • [14:17] Grey_Fox: so, I suppose an isomorphic dynamic would be technique/fallacy to domain to technique
  • [14:17] Eliezer: Jim, the idea is that blogs are for long articles. Each time the article uses a concept, it links back to the Wiki. Someone going to the Wiki
  • finds a quick definition of the concept (possibly with some more internal links) and more importantly a sequence-ish list of blog posts with quick summaries.
  • [14:17] LeopoldTal: More tree-ish than sequence-ish.
  • [14:18] Grey_Fox: a linked tree maybe
  • [14:18] Eliezer: Even if we have editable blog posts this can still imply a different dynamic - because of the idea that when you have original content to
  • contribute, you write a blog post, and substantial user discussions are visible there.
  • [14:18] Grey_Fox: Eliezer: do you think we should have a reference of the more pedestrian fallacies?
  • [14:19] Eliezer: Linking ubiquitously from the blog to the Wiki is new-user-friendlier than linking directly to blog posts because the Wiki provides a quick
  • summary and the index.
  • [14:19] Xuenay: Eliezer's idea sounds good to me
  • [14:19] vnesov: Blog posts are more short-term and less stable than wiki articles.
  • [14:19] timewarp has quit IRC ("Leaving"?)
  • [14:19] rwallace: Yeah.
  • [14:19] Eliezer: Grey: On this theory, if you want to write up a fallacy, you would start by writing a blog post. If it got voted up enough, or if anyone else
  • wanted to reference the concept in the blog post, they would create a Wiki page that summarizes the concept and links to your blog post.
  • [14:19] jimrandomh: How about links in the other direction - from the wiki to blog posts
  • [14:19] vnesov: A wiki article is a stable concept, than can be gradually elaborated with whatever details get sorted out.
  • [14:19] PeerInfinity: random idea: a script to automatically update the blog pages, so that every time a concept is mentioned that has a corresponding page on the
  • wiki, a link is created to the wiki?
  • [14:19] timewarp has joined #lesswrong
  • [14:20] jimrandomh: For working through an example, blog posts work better than wikis
  • [14:20] rwallace: jim - I think that is a very good idea
  • [14:20] Yvain has quit IRC ("http://www.mibbit.com ajax IRC Client"?)
  • [14:20] Eliezer: Jim, in the other direction, this is a different solution than needing it to all be consensus voting everything; instead you write your own blog
  • posts.
  • [14:20] jimrandomh: so maybe, blog posts have lots of links into the wiki, wiki articles have a main section with lots of links that only go to wiki articles, and
  • wiki articles have an 'examples' section at the bottom which contains only links to blog posts
  • [14:20] LeopoldTal: jimrandomh: All/the main blog posts (comments?) mentioning the concept the article is about get linked, with their structure if they're a
  • sequence.
  • [14:20] Grey_Fox: Eliezer: so, new pages are first vetted through the blog
  • [14:20] Eliezer: No, I'm proposing that blogging is a more efficient way of creating detailed discussions than Wiki'ing.
  • [14:21] vnesov: A blog may present controversial assertions, opening a debate; a wiki article should state stable facts.
  • [14:21] Grey_Fox: and then a summary of the concepts is posted on the wiki for posterity
  • [14:21] Grey_Fox: oh I see
  • [14:21] saturnnn: perhaps each blog post should have a wikiable summary section where new links can be added even when the post is old
  • [14:21] vnesov: PeerInfinity: I've never seen this idea about automatic linking work out non-horribly.
  • [14:22] jimrandomh: You're right about wikis being more stable/better for stable things than blogs, and blogs being better for discussion. I think it's mainly
  • because people approach blogs with articles sorted by date.
  • [14:22] Eliezer: Saturn, I would propose that instead the Wiki entry for that concept (not the particular blog post) should be wikiable (of course) and have new
  • links added.
  • [14:22] Grey_Fox: just trying to get the dynamic here
  • [14:22] PeerInfinity: ok then, not entirely automatic, but a script to speed up the process of manually making the links?
  • [14:22] Eliezer: Wiki for concepts and summaries, posts for theses and discussions. Each format doing what they do best.
  • [14:22] PeerInfinity: by the way, here's a good example of tree structure for wiki content: http://www.acceleratingfuture.com/wiki/index.php?title=All_Topics
  • [14:22] Grey_Fox: alright
  • [14:22] SomeGuyM has joined #lesswrong
  • [14:23] jimrandomh: Ok, sounds like we agree on the distinction between what goes in the blog vs. what goes on the wiki
  • [14:23] PeerInfinity: (I was planning to integrate the Less Wrong wiki content into the Accelerating Future wiki, by the way)
  • [14:23] vnesov: Wiki articles about existing content are of course vetted through the talk pages of those articles, blog is for pushing new content.
  • [14:23] EnglishGent: hello SomeGuyM
  • [14:23] Yvain has joined #lesswrong
  • [14:23] Eliezer: Peer, by "integrate" do you mean copy or link?
  • [14:23] EnglishGent: hello Yvain
  • [14:23] jimrandomh: It would be nice to have a good way of adding links to new wiki articles on old blog posts, particularly if the wiki is going to have a lot of
  • links to blog posts, but that's not essential
  • [14:24] Yvain: Hi. Sorry, Mibbit was acting weird for me. Hope I didn't miss much.
  • [14:24] noko has joined #lesswrong
  • [14:24] Eliezer: Yvain: We solved everything.
  • [14:24] noko: ARE THERE ANY ROBOTS IN HERE
  • [14:24] Grey_Fox: the secrets of the universe were laid bare
  • [14:24] Eliezer: Jim: I agree, the thing to consider is that for now development resources are highly limited.
  • [14:24] saturnnn: Eliezer, i meant that once i finished reading the blog post, at the bottom there would be the relevant wiki content... rather than having to
  • click on it to see it
  • [14:24] Erkbot has quit IRC (Remote closed the connection?)
  • [14:25] Yvain: Neat. You'll all have to tell me about it sometime.
  • [14:25] saturnnn: this is assuming the wiki content is very short
  • [14:25] Eliezer: Saturn: But a post may belong to more than one concept.
  • [14:25] PeerInfinity: by integrate, I meant either transclude or link
  • [14:25] stillflame: process note: we've spent 15 minutes on this topic.
  • [14:26] jimrandomh: Are we all happy with the wiki's software/hosting/etc?
  • [14:26] Grey_Fox: so, long stuff goes in the blogs, the wiki is for summaries and concepts
  • [14:26] komponisto has joined #lesswrong
  • [14:26] Grey_Fox: links to concepts should link to the wiki
  • [14:26] Roland_LW: NOKO: I'M A ROBOT.
  • [14:27] jimrandomh: Minor note: the front page needs a prominent link to the wiki
  • [14:27] Grey_Fox: the concepts page should link to relevant posts on the concept
  • [14:27] jimrandomh: Please don't shout
  • [14:27] Eliezer: Jim, we're scheduled to get a MediaWiki (same format) integrated into lesswrong.com, don't know when that's scheduled to go live.
  • [14:27] vnesov: It would be neat to get the wiki installed somewhere without advertising.
  • [14:27] Eliezer: vnesov, see above
  • [14:27] jimrandomh: Will the software change preserve content added before then?
  • [14:27] vnesov: rihgt
  • [14:27] PeerInfinity <3 wikis :)
  • [14:28] Eliezer: Jim, that is my understanding.
  • [14:28] mib_1dg94ml0 has joined #lesswrong
  • [14:28] mib_cwvwwh has joined #lesswrong
  • [14:28] mib_cwvwwh has quit IRC (Client Quit?)
  • [14:28] michaelblume has joined #lesswrong
  • [14:28] jimrandomh: Anything else to say about the wiki before moving on to the next topic?
  • [14:28] MichaelGR: what about licensing? will Less Wrong wiki's be creative commons or GPL? If it is compatible with wikipedia's license, we could use some of thei
  • content to build on (when appropriate)
  • [14:28] vnesov: From the point of view of the wiki, blog posts should have the same status as the rest of the Internet.
  • [14:28] komponisto has quit IRC (Client Quit?)
  • [14:28] dfranke: GPL makes no sense for !(code).
  • [14:29] PeerInfinity: should I ask Michael Anissimov if he would agree to hosting the wiki on acceleratingfuture.com?
  • [14:29] saturnnn: maybe you meand GFDL
  • [14:29] vnesov: Wiki should prefer to link to itself, and blog posters should prefer to link to the wiki.
  • [14:29] Eliezer: No. We are hosting it at lesswrong.com.
  • [14:29] PeerInfinity: ok
  • [14:29] MichaelGR: dfranke: I meant GFDL, or whatever wikipedia use(d)
  • [14:29] komponisto has joined #lesswrong
  • [14:29] dfranke: MichaelGR: wikipedia uses CC.
  • [14:29] Eliezer: Both blogs and wikis should prefer to link to the wiki for *concepts* and to the blog for *discussions*.
  • [14:29] MichaelGR: dfranke: Glad they made the switch. thx
  • [14:30] Xuenay is now known as KajSotala
  • [14:30] KajSotala: (Might as well go by a name most people will recognize...)
  • [14:30] Evercat: I don't believe Wikipedia has switched to CC yet
  • [14:30] KajSotala: (Even if this feels weird)
  • [14:30] vnesov: How do you link to discussion? Old discussions are generally dead.
  • [14:30] jimrandomh: I don't think Wikipedia is capable of changing its licensing at this point
  • [14:30] LeopoldTal: Much-linked-to discussions won't necessarily stay dead.
  • [14:30] michaelblume: vnesov: We have recent comments linked from the front page
  • [14:31] jimrandomh: Better distinction: link to the wiki for rationality terms, the blog for examples and outside applications
  • [14:31] vnesov: You *create* blog posts to start discussions, but also to introduce concepts new to the community.
  • [14:31] Roland_LW: It would be interesting to have discussions like the wikipedia "talk" pages
  • [14:31] Eliezer: Vnesov, by "discussion" I mean a previously existing extended discourse upon a particular topic, including the presentation of a long thesis with
  • support and its discussion in the comments. This should be a blog post, and if you want to link to it you should link to the blog.
  • [14:31] dfranke: michaelblume: ah nm, looks like GFDL is still the default.
  • [14:31] MichaelGR: Roland_LW: If mediawiki is used, I think we'll get Talk pages by default
  • [14:31] dfranke: s/michaelblume/MichaelGR/
  • [14:32] Traveler2 has joined #lesswrong
  • [14:32] Grey_Fox: so, about licensing
  • [14:32] vnesov: Eliezer: but the conclusions from any discussion should be integrated into wiki articles, so that there would be no point in actually *reading* the
  • old discussion. Any new part of the discussion can start anew, in the global namespace.
  • [14:32] Grey_Fox: is CC fine?
  • [14:32] Eliezer: For so long as Wikis don't support threaded comments with *voting*, they can't do what the current LW blog does. And conversely, so long as the
  • LW reddit-software doesn't support multiple editors with change control and comments linked to the version that they originally commented on, it can't do what the
  • Wiki does.
  • [14:33] PeerInfinity: yes, MediaWiki has Talk pages by default, but we don't necessarily need to use them
  • [14:33] michaelblume: Grey_Fox: I'm pretty sure CC and GPL are interoperable these days
  • [14:33] saturnnn: it would be cool if the talk pages could be replaced by the reddit-style comment threads... i assume that's way too much work to be worth it
  • [14:33] stillflame: would we point to a irc for "active discussion"? does that usually amount to newbies reasking questions that eventually get put into a faq?
  • [14:33] Grey_Fox: k
  • [14:33] michaelblume: I usually use CC for my own writing
  • [14:33] vnesov: Yes, doing the wiki article talk bages LessWrong-style would be an improvement.
  • [14:33] vnesov: But it should be a separate sevice from the blog.
  • [14:34] Roland_LW: Eliezer, you mentioned voting, I'm starting quesitoning the usefulness of this feature
  • [14:34] Eliezer: Vnesov, to the extent that conclusions are short and can be summarized and don't sound stupid without their attached support, they can appear in
  • the wiki's brief concept descriptions. But this does not obsolete the page which contains the support and the criticisms and the replies.
  • [14:34] PeerInfinity makes a note to check if there is a MediaWiki extension to do threaded comments with voting...
  • [14:34] michaelblume: Vnesov, Eliezer it seems it would be useful to still include a link to the original discussion in which the conclusions were linked
  • [14:34] Grey_Fox: we could just put the lesswrong irc channel link somewhere on the site
  • [14:35] Eliezer: saturnn, if you've got enough money you can always *pay* for work to get done... but in the absence of OSS volunteers, that's the issue basically.
  • [14:35] jimrandomh: Wiki talk pages should only be for discussing presention issues. For discussion about content, it makes more sense to post on Less Wrong where
  • people will see it
  • [14:35] jimrandomh: On the blog, that is
  • [14:35] Grey_Fox: yeah
  • [14:35] vnesov: Of course, everything is archived, whenever it's necessary blog can be linked to. I think it would be rarely used (how often do we link to
  • *discussions* now?)
  • [14:35] michaelblume: jimrandom: agreed
  • [14:35] Eliezer: Vnesov, you appear to believe that we should wiki everything possible. Why? Wikis have a different tone than blog discussions. I see that this
  • is good for some things, but not all things.
  • [14:36] LeopoldTal: If we have interesting findings, they will be nonobvious (i.e. sound stupid), we must be OK with crazy-sounding summaries or find a way to do
  • without summaries.
  • [14:36] vnesov: michaelblume: yes, that may be a good idea, to link in the footnotes.
  • [14:37] Eliezer: Also, you seem to assume an authoritative knowledge structure in which a definite right answer is known (a la most of Wikipedia). Even when LW
  • agrees on something, the primary usefulness to newbies is not being *told* what we think but of having the issue *explained*.
  • [14:37] vnesov: jimrandomh: sounds good, any substantial discussion should turn into full-fledged blog discussion, so there is no need for fancy talk pages.
  • [14:37] saturnnn: LeopoldTal, we can write the summaries in a neutral style rather than argumentative style
  • [14:37] dfranke: I think the wiki should be used for 1. conclusions we almost all agree on, 2. defining jargon, 3. explaining standard positions in long-running
  • debates.
  • [14:37] MichaelGR: talk pages might still be useful for more housekeeping discussions about wiki pages that don't necessarily deserve blog posts
  • [14:38] Eliezer: "The following was argued in such-and-such a blog post using this primary argument" (3 line summary)
  • [14:38] Roland_LW: Vnesov: talk pages have more structure(at least those in wikipedia)
  • [14:38] vnesov: Eliezer: no, everything should be *potentially* wikified, but the stub articles with a few-words summary and link to a blog post should be OK for
  • the most stuff.
  • [14:38] Eliezer: MGR, agreed
  • [14:38] jimrandomh: How about if wiki articles start with a list of links to detailed explanations, followed by a link-laden but useless to newcomers summary,
  • followed by a list of links to examples?
  • [14:38] Roland_LW: What I wanted to say is: the blog discussions tend to big and unstructured so it becomes hard to find the relevant information
  • [14:38] Eliezer: Vnesov, what do you mean "potentially wikified"?
  • [14:38] saturnnn: I think the wiki shoud be used for making it easy to explore old blog posts that are still really interesting but currently hard to wander across
  • [14:39] vnesov: Eliezer: wiki may list multiple points of view, and link to different posts and discussions where the controversy is a topic.
  • [14:39] dfranke: Talk pages should be *exclusively* housekeeping discussions. They break down if you try to use them for anything else.
  • [14:39] stillflame: (...didn't i read something recently about talking not being about info...)
  • [14:39] vnesov: dfranke: right.
  • [14:39] LeopoldTal: saturnnn: That's the problem! "A diet that tries to make you lose weight by drinking oil" is neutral and sounds crazy - explaining Roberts'
  • theory is argumentative, sounds true and is way too long.
  • [14:40] dfranke: ongoing debate needs to be on the blog and nowhere else, until it's gone on long enough for "standard positions" to be established, at which point
  • they can be documented on the wiki.
  • [14:40] Eliezer: Documented, but not necessarily argued, unless the argument is short.
  • [14:40] saturnnn: LeopoldTal, how about "A diet that tries to make you lose weight by exploiting flavor-calorie association"
  • [14:41] vnesov: Eliezer: I mean that an article starts as a few-words summary, and may grow to include most of the content in some blog post, but not word-by-word.
  • [14:41] dfranke: I think the Shangri-La posts are a perfect example of what doesn't belong on the wiki.
  • [14:41] Eliezer: "Used to illustrate issues of _akrasia, _willpower, and _individual_variation in what works for someone."
  • [14:41] taw has joined #lesswrong
  • [14:41] Eliezer: where the underlined are other wiki links
  • [14:41] taw: hi guys
  • [14:41] michaelblume: hi taw
  • [14:42] dfranke: pages about jargon can link to them as examples, but they shouldn't have their own article.
  • [14:42] Eliezer: dfranke, but maybe I want to refer to Shangri-La in passing in the future. So I link to Shangri-La the Wiki article and then it gives a quick
  • summary and refers back to the blog posts (with one-sentence summaries of their relevance)
  • [14:42] Yvain: I think the best place for the Shangri-La posts on the wiki would be in the article on akrasia, unless lots of other posters start using Shangri-La
  • as an example to the point where new readers of LW would probably wonder what this Shangri-La thing was.
  • [14:42] vnesov: For example, if a standard position on a portion of an old blog post explaining a concept gets changed, the corresponding wiki article should
  • introduce that part explicitly and describe the changes (or updated position).
  • [14:43] Eliezer: dfranke, I don't understand the reason for that policy - quickly defining unfamiliar concepts would seem to be a *primary* use of a Wiki
  • [14:43] MichaelGR: wiki is not paper, so there could be a stub about shangri la on the wiki, but it would probably have a lot less inbound links than other more
  • reference pages
  • [14:43] stillflame: process note: we're at 35 minutes for the topic of the wiki.
  • [14:43] Eliezer: stillflame, are there other topics pending? (also who says we can only discuss one topic at once?)
  • [14:43] taw: i'm actually quite curious why shangri la diet has any appeal to you guys
  • [14:43] saturnnn: I'm pretty sure I remember a lot of discussion of Shangri-La on OB a while back, too
  • [14:43] taw: it has absolutely zero research behind it
  • [14:44] taw: and they seem to be research-hostile
  • [14:44] jimrandomh: Fake Wiki page that shows how I think they should be structured: http://pastebin.com/m5a6acbe9
  • [14:44] PeerInfinity: create a new IRC channel for discussing another topic?
  • [14:44] taw: even atkins etc. are more research friendly than that
  • [14:44] stillflame: other topics suggested: meeting schedule, Paranoid Debating bot, avoiding meta-topics
  • [14:44] vnesov: Hey, if we discuss more then one topic my head will explode. It's already to fast as it is.
  • [14:44] jimrandomh: All three of the diet-related posts used diets as examples to get at other, rationality related topics
  • [14:45] jimrandomh: It's only the comments that talked about dieting directly
  • [14:45] KajSotala: vnesov: You have it easy, I'm playing an RP-intensive game of multiplayer Civ4 at the same time :)
  • [14:45] LeopoldTal: Eliezer: In my experience, when >1 topics are discussed, one eats the whole conversation.
  • [14:45] saturnnn: taw, I think the interest is more in the person who created it rather than it being such a wonderful thing per se.
  • [14:45] Eliezer: Jim, I'd suggest putting the summary above the links - that's user-friendlier.
  • [14:45] Roland_LW: I have a new TOPIC: does anyone know what is the current scientific position on the question if all physical computation is turing computable?
  • [14:45] PeerInfinity makes a wiki link for the new topic: Turing Computability
  • [14:45] dfranke: Eliezer: Shangri-La is not a working concept that we're likely to reuse. The concepts are akrasia or individual variation.
  • [14:45] Eliezer: Roland: "YES"
  • [14:46] taw: Roland_LW: what would it mean if it wasn't?
  • [14:46] dfranke: Eliezer: those things should have wiki articles, but if you want to link about Shangri-La, just link to the blog posts.
  • [14:46] jimrandomh: The idea behind putting the Explanations section before the summary section was that if the topic is unfamiliar, you should read an
  • article-length explanation first, before going click-happy
  • [14:46] Eliezer: Dfranke, that's a testable hypothesis, and if I want to refer to Shangri-La in a future post, I could just create the Wiki article at that time.
  • [14:46] Eliezer: Df: why?
  • [14:47] dfranke: Eliezer: sure, if the idea gets reused regularly, then it gets a wiki page.
  • [14:47] Eliezer: Jim: It seems to me that what the user *wants* is a fast explanation. Throwing links at them is just some valueless, inexplicable resource until
  • they know what the concept is *about*.
  • [14:47] vnesov: So, who writes an article discussing the suggested wiki-blog roles?
  • [14:47] dfranke: Eliezer: but right now it's just something that's been brought up in passing.
  • [14:47] Eliezer: df: why not if it gets reused *once*?
  • [14:47] Roland_LW: Eliezer: so all PC is Turing computable? Can you give me a link?
  • [14:48] michaelblume: Roland_LW what is "physical computation"?
  • [14:48] PeerInfinity hopes that dfranke won't be given permission to delete pages from the wiki...
  • [14:48] EnglishGent: Roland_LW - http://en.wikipedia.org/wiki/Church–Turing_thesis
  • [14:48] EnglishGent: it is *the* basic idea underlying computer science
  • [14:48] vnesov: computer science?
  • [14:48] ErikMesoy: Bad link for me. Try http://en.wikipedia.org/wiki/Church-Turing_thesis ?
  • [14:49] EnglishGent: same link ErikMesoy :)
  • [14:49] Roland_LW: michaelblume: physical computation is anything a physical system can do, while turing computation is what computers do.(in simple words)
  • [14:49] ErikMesoy: Might be my charset then. Your hyphen turned into a black block.
  • [14:49] jimrandomh: I yield to Eliezer on putting the summary block before the Explanations links
  • [14:49] Eliezer: Yes, we shall definitely be avoiding deletinionism except for the most egregious cases of irrelevance or original research.  :)
  • [14:49] KajSotala: EnglishGent: There was some special character in yours
  • [14:49] jimrandomh: Anything more to say about wikis? I think we've pretty much covered it
  • [14:49] LeopoldTal: What's wrong with original research?
  • [14:49] Eliezer: it goes in a blog post
  • [14:49] LeopoldTal: Right.
  • [14:49] EnglishGent: they are _probably_ equivalent - as far as we know this universe doesnt allow harnessable hypercomputation
  • [14:49] vnesov: Right, linking to posts such as http://www.overcomingbias.com/2008/07/the-meaning-of.html is just useless, too much info.
  • [14:50] dfranke: Eliezer: actually... I could see using Wiki pages for sequences, that are just a list of links with one-line summaries.
  • [14:50] Eliezer: dfranke: yes - although I'd wanted to see sequences with Prev/Next clickthroughs the Wiki sounds like a more important feature
  • [14:50] Roland_LW: Church-Turing thesis doesn't talk about PC, only about effectively calculable functions
  • [14:50] PeerInfinity: another topic suggestion: IRC chatlogs - do we want to post them online?
  • [14:50] Eliezer: (to be implemented first)
  • [14:51] Eliezer: Peer: sounds fine to me, anyone object?
  • [14:51] jimrandomh: I agree that they should be posted
  • [14:51] Roland_LW: Where I'm trying to go with my question is: how do we know that a human brain is turing computable in the first place, this is a fairly strong
  • assumption IMHO
  • [14:51] komponisto: (Probably should ask after we're done!)
  • [14:51] michaelblume: Roland_LW: The universe, as far as we can tell, uses real numbers
  • [14:51] mib_1dg94ml0 has left #lesswrong
  • [14:51] dfranke: PeerInfinity: yes, we do.
  • [14:52] jimrandomh: michaelblume: Disagree; as far as we can tell, the universe uses only discrete math, but our module is incomplete so we have to use real
  • approximations
  • [14:52] EnglishGent: actually Roland_LW it's a very weak assumption - for the human brain *not* to be Turing computable it would have to be exploting unknown
  • physics
  • [14:52] Eliezer: Roland, Scott Aaronson has written a good deal about physics not being able to solve NP-complete problems even, so hypercomputation presumably
  • falls under the same rubric
  • [14:52] Roland_LW: michaelblume: I don't think this is the case.
  • [14:52] EnglishGent: and it's very hard to believe that any unknown physics would be available at the mass/energy levels
  • [14:52] EnglishGent: that the brain is restricted to
  • [14:52] Grey_Fox: so, what was the next topic after the wiki?
  • [14:52] jimrandomh: IRC meeting schedule
  • [14:53] Grey_Fox: ah right
  • [14:53] vnesov has quit IRC ("http://www.mibbit.com ajax IRC Client"?)
  • [14:53] Eliezer: I'll recommend once a month at most, and "official" gatherings once a quarter maybe
  • [14:53] PeerInfinity: also, where should the chat logs be posted? to the wiki?
  • [14:53] Eliezer: because IRC eats the soul
  • [14:53] EnglishGent: even if the brain turned out to be a quantum computer (which is also hard to imagine - there's far too much thermal noise) it would *still* be
  • Turing computable
  • [14:53] Roland_LW: EnglishGent: I don't think that your conclusion follows.
  • [14:53] Grey_Fox: I do agree that IRC eats the soul
  • [14:54] vnesov has joined #lesswrong
  • [14:54] Grey_Fox: but it's such a delicious feeling!
  • [14:54] EnglishGent: Roland_LW - why not?
  • [14:54] vnesov wonders if he missed something
  • [14:54] taw: the idea that brain might be a quantum computer is nonsense
  • [14:54] Infotropism: why would the brain need to be a quantum computer ?
  • [14:54] Eliezer isn't sure whether the current LW blog model allows for attachments, so maybe post it as an attachment to the Wiki under "Less Wrong IRC chats"
  • (note that this is a continually edited page...)
  • [14:54] taw: it's just way too many orders of magnitude wrong for it
  • [14:54] EnglishGent: taw - I agree.. I think it's highly unlikely
  • [14:54] taw: not unlikely - pure nonsense
  • [14:54] PeerInfinity volunteers to post the chat log
  • [14:54] Eliezer: it is not *nonsense*, it is *implausible*
  • [14:55] jimrandomh: Ok, not much to say on the schedule topic
  • [14:55] EnglishGent: but the idea it's a Turing oracle is surely even more unlikely
  • [14:55] PeerInfinity: with proper wiki formatting
  • [14:55] jimrandomh: I support monthly. Does anyone disagree?
  • [14:55] Grey_Fox: could we have weekly paranoid debating meetups?
  • [14:55] ErikMesoy: How would monthly work with regard to weekdays?
  • [14:55] Roland_LW: EnglishGent you assumption was: not turing computable -> exploit unknown physics
  • [14:55] jimrandomh: The channel is always here, whether there's an "official" meetup scheduled or not
  • [14:55] Eliezer: would suggest varying time vs weekday/weekend
  • [14:55] Roland_LW: you start with the assumption that all knowable physics is turing computable, this is the very question I'm posing
  • [14:55] vnesov: If it's just implausible, then nonsense must be meaningless.
  • [14:56] Eliezer: I'd say we should have one Paranoid Debating meetup before debating how often to schedule them
  • [14:56] EnglishGent: Roland_LW - why dont you think that follows? *all* known physics is computable
  • [14:56] EnglishGent: see for example: http://en.wikipedia.org/wiki/Hypercomputation
  • [14:56] michaelblume: Roland_LW: we understand physics at the level that on which the brain operates
  • [14:56] jimrandomh: Ok, monthly, varying times
  • [14:56] Grey_Fox: monthly for meetings like this
  • [14:56] taw: Eliezer: what's Paranoid Debating?
  • [14:57] EnglishGent: to be a hypercomputer one needs a machine capable of doing rather implausible things - like measuring real-numbered physical constants with
  • infinite precision - which Quantum mechanics explicitly dissalows
  • [14:57] Grey_Fox: mafia + aumann
  • [14:57] PeerInfinity: Paranoid Debating
  • [14:57] taw: Grey_Fox: more?
  • [14:57] PeerInfinity: (that will be a clickable link in the chatlog)
  • [14:57] vnesov: Scheduling paranoid debates? How many people are actually enthused by the idea? I'm not.
  • [14:57] michaelblume: taw: http://lesswrong.com/lw/77/selecting_rationalist_groups/
  • [14:57] Boxo has joined #lesswrong
  • [14:57] jimrandomh: We can schedule *a* paranoid debate
  • [14:57] Grey_Fox: vnesov: I am
  • [14:57] Roland_LW: Folks: lets make it simple: take a glass full of water and shake it: is this computable? And why do you think it is?
  • [14:57] taw: is it the thing we played in london?
  • [14:58] jimrandomh: we have to see how well it works on IRC, thoguh
  • [14:58] dfranke: heh... I have an idea...
  • [14:58] Wolfbot has joined #lesswrong
  • [14:58] taw: estimation game with deceiver?
  • [14:58] Grey_Fox: yes
  • [14:58] jimrandomh: yes, that one
  • [14:58] stillflame: (this Wolfbot may have better flood protection)
  • [14:58] michaelblume: Roland_LW: the question is, does it matter?
  • [14:58] Grey_Fox: awesome
  • [14:58] taw: ok then
  • [14:58] dfranke: if we have a chatroom open all the time, then we could have paranoid debates without knowing we're having them.
  • [14:58] Roland_LW: michaelblume: yes it matters
  • [14:58] jimrandomh: stillflame: Try registering it with NickServ
  • [14:58] taw: i don't think deceiver changes that much
  • [14:58] vnesov: jimrandomh: right, with the most enthused responsible for getting the bot running.
  • [14:58] EnglishGent: yes it matters - it would have all sorts of practical implications
  • [14:58] Roland_LW: michaelblume: if the glass of water is not turing computable why do you suppose a brain is?
  • [14:59] dfranke: the bot just PMs someone telling them to be the deceiver, then ten minutes later puts someone on the spot as spokesman :-)
  • [14:59] Eliezer: roland, for relevance we need a noncomputable *mind* not just a noncomputable brain
  • [14:59] EnglishGent: Roland_LW - yes - that's computable - none of the physical laws required to describe what happens to the water require solving the halting
  • problem
  • [14:59] Grey_Fox glares at the halting problem
  • [15:00] Grey_Fox: it's a real annoyance
  • [15:00] Wolfbot has quit IRC (Remote closed the connection?)
  • [15:00] stillflame is now known as Wolfbot
  • [15:00] EnglishGent: and in practice one wouldnt even *need* to worry about quantum mechanical / relativistic effects - simple classical thermodynamics / fluid
  • dynamics would describe the outcome in as much detail as is interesting
  • [15:00] steven0461: why ten minutes later?
  • [15:00] Roland_LW: Eliezer: do you know a mind design that is Turing computable?
  • [15:00] Wolfbot is now known as neverwrong
  • [15:00] Roland_LW: I mean an intelligent mind
  • [15:00] jimrandomh: I've noticed that Less Wrong seems to make great strides when dealing with topics where bad explanations are prevalent, but spin in circles
  • elsewhere
  • [15:00] Grey_Fox: ours
  • [15:00] vnesov: Grey_Fox: not as much as it's purported to be.
  • [15:01] michaelblume: EnglishGent: exactly, we don't *care* about the paths of individual molecules of neurotransmitters
  • [15:01] michaelblume: that's just noise
  • [15:01] jimrandomh: One topic came to mind, as having many bad explanations presented but no good ones: "flow state" (the state of mind where you're really well
  • focused)
  • [15:01] EnglishGent: point - there *might* be a prehistoric monster the size of a double decker-bus living in Loch Ness -- but it's most certainly not the default
  • assumption
  • [15:01] Grey_Fox: jimrandomh: your mission is to write more bad explanations :p
  • [15:01] Eliezer: Roland, in the absence of a TOE or a working AI, what kind of evidence would you accept to update on, given that no known abilities of
  • intelligence or operations of physics are non-Turing-computable?
  • [15:01] EnglishGent: if your suggesting the mind is non-computable you need to explain why - as given what we currently know about the brain shows no hint of
  • noncomputable processes at work
  • [15:02] SomeGuyM has quit IRC ("http://www.mibbit.com ajax IRC Client"?)
  • [15:02] EnglishGent: *and* what we currently know about physics says there cant be any non-computable processes
  • [15:02] jimrandomh: Does anyone know if there's any research which has tried distinguishing flow state in an fMRI?
  • [15:02] Eliezer: English: not true, electrons could have oracular mass
  • [15:02] Roland_LW: Eliezer: the problem is all our physics knowledge are the result of deliberative reasoning which is itself subject to TC-limitations. So how do
  • we know that we didn't miss important aspects of reality in our abstractions?
  • [15:02] Eliezer: (mass equal to Chaitin's halting omega, say)
  • [15:03] jimrandomh: Caveat: We can have computable processes which can't be computed by anything with less mass than the process has
  • [15:03] Eliezer: Roland: Don't worry, I'd have caught it.
  • [15:03] EnglishGent: yes - they could have Eliezer - but that still wouldnt be good enough - you'd have to be able to measure that mass with arbitrary precision
  • [15:03] EnglishGent: which Heisneberg & Planck forbid
  • [15:03] Eliezer: use lots of electrons
  • [15:03] Eliezer: take lots of measurements, central limit theorem
  • [15:03] EnglishGent: you still wouldnt have access to unbounded precision - unless you could measure an infinite number of them
  • [15:03] Infotropism: Even if hypercomputing is possible, what reason is there to expect it in the human brain ? Because the latter seems complicated, mysterious,
  • doing things we can't reproduce yet artificially ?
  • [15:03] vnesov: jimrandomh: I vaguely recall I might've seen something.
  • [15:04] Eliezer: Well, let the universe grow unboundedly over time and then let me perform unboundedly large finite measurements.
  • [15:04] EnglishGent: bear in mind that to recover omega & not just some finite approximation to it you'd need to measure aleph-null electron masses
  • [15:04] Grey_Fox: is that going to happen though Eliezer?
  • [15:04] Eliezer: Nope.
  • [15:04] EnglishGent: not to mention there's no reason to believe that any of the physical constants are oracular numbers
  • [15:05] Eliezer: well, if *I* were designing a universe *I* would make them oracular numbers.
  • [15:05] michaelblume: EnglishGent: it would be awfully convenient
  • [15:05] Roland_LW: Eliezer:"Roland: Don't worry, I'd have caught it." you are joking or what?
  • [15:05] Eliezer: therefore it is true QED
  • [15:05] EnglishGent: perhaps it would be - but there's no reason to believe it at present
  • [15:05] Grey_Fox: lol
  • [15:05] neverwrong is now known as stillflame
  • [15:05] EnglishGent: perhaps Omega is *itself* a physical constant
  • [15:05] Grey_Fox would like to know the trick of making universes
  • [15:05] smoofra: Eliezer: you know mvassar's email? the one on singinst.org doesn't seem to work.
  • [15:06] EnglishGent: one might be able to define the physics of a universe as the rules specifying what is computable within it
  • [15:06] EnglishGent speculates wildly
  • [15:06] saturnnn: Roland, I bet Eliezer can't even prove that reality is really real!
  • [15:06] Eliezer: michael no space aruna at yahoo dot com
  • [15:06] neverwrong has joined #lesswrong
  • [15:06] EnglishGent: but - the point stands - as far as we know there's no harnessable hypercomputation in this universe... and even if there is - there's no
  • evidence to suggest that the brain uses it
  • [15:06] Stormoog_ has joined #lesswrong
  • [15:06] stillflame: (neverwrong is the Wolfbot with a registered nick)
  • [15:07] Grey_Fox: k
  • [15:07] EnglishGent: that doenst mean that *necessairly* there isnt & it doenst -
  • [15:07] LeopoldTal: As far as we know, the Flying Spaghetti Monster threw up on the candle next to me.
  • [15:07] EnglishGent: but well... there *might* be a prehistoric monster the size of a double decker bus living in Loch Ness
  • [15:07] smoofra: Eliezer: thanks.
  • [15:07] EnglishGent: I still wouldnt bet on it :)
  • [15:07] Grey_Fox: occams razor strikes back!
  • [15:07] Roland_LW: One of the points I'm trying to make is that we only see Turing computability everywhere because that's all what we can capture in our formal
  • systems. Every mathematical formula is TC, so as soon as we see a physical law it is TC. What about the things we didn't manage to put into "laws"
  • [15:08] jimrandomh: This conversation is one of the sort where it's very easy to get lost and talk past eachother
  • [15:08] Grey_Fox: or as it has been redesignated "Solomonoff's lightsaber"
  • [15:08] michaelblume: EnglishGent: when we want to probe the physics we don't understand yet, we build machines the size of cities
  • [15:08] jimrandomh: And soI don't think it's well suited for IRC
  • [15:08] vnesov: People, a chat channel is definitely not a medium for many-tentacled discussions like this one. I'm not sure it's even fun.
  • [15:08] MosaSaur: What timezones do the times on the lesswrong blog refer to?
  • [15:08] michaelblume: vnesov: agreed
  • [15:08] EnglishGent: oh - CERN is rather bigger than many cities michaelblume... :)
  • [15:08] Grey_Fox: it's 6 am here
  • [15:08] dfranke: It seems to me that the suggestion that the human brain is uncomputable is just a rationalization to justify the tendency that we've seen
  • throughout history to believe that mysterious things have to stay mysterious.
  • [15:08] Tiiba has joined #lesswrong
  • [15:08] Eliezer: Roland, but there are no known phenomena *not* covered by the laws
  • [15:08] LeopoldTal: vnesov: Problem, every conversation wants to fork.
  • [15:08] jimrandomh: We need a better topic to displace this one
  • [15:09] Grey_Fox: well, the iternary has sort of run out
  • [15:09] Grey_Fox: unless we want to play mafia now?
  • [15:09] Stormoog_: or a forum with threaded chats :)
  • [15:09] jimrandomh: Hmm, pjeby isn't here
  • [15:09] Grey_Fox: I think stillflame got the bot working
  • [15:09] EnglishGent: I agree dfranke - it seems to me just a continuation of - the earth is at the center of the universe... okay .... solar system... galaxy...
  • well - at least we're differnt from animals... etc
  • [15:09] Eliezer: Grey, who invented "Solomonoff's lightsaber" is that original?
  • [15:09] jimrandomh: I wanted to Other-Optimize his writing style
  • [15:09] vnesov: Leopold: and it also wants to be a cult, as we all know. ;-)
  • [15:09] LeopoldTal: vnesov: No, those are causes
  • [15:09] Grey_Fox: Eliezer: well, I invented that specific phrase
  • [15:09] Roland_LW: Eliezer: what about chaotic flow of water for example
  • [15:09] Grey_Fox: I think
  • [15:09] EnglishGent: the evidence says we're computers ... if we are to believe otherwise we need a good reason to believe otherwise
  • [15:09] dfranke: EnglishGent: bingo.
  • [15:09] saturnnn: Roland, why don't you write a post about this on LW
  • [15:10] Eliezer: Grey, do I attribute it to Grey_Fox or what?
  • [15:10] LeopoldTal: I have a question for people who realized religion was wrong from an early age
  • [15:10] PeerInfinity makes a wiki link for the new topic: reasons for rejecting religion at a young age
  • [15:10] EnglishGent: that we *might* not be, or that it somehow offends my sense of dignity that we are - are not sufficient reasons to doubt it
  • [15:10] Grey_Fox: Eliezer: attribute it to Patrick
  • [15:10] Eliezer: Just "Patrick"?
  • [15:10] Grey_Fox: well, that's my lesswrong account
  • [15:10] vnesov: A perpetually forking causal cult.
  • [15:10] Grey_Fox: so it's the thing that'll let me jockey for status :p
  • [15:10] Roland_LW: saturnnn: that's what I intended, I just wanted to make sure there was no simple answer to what I was proposing.
  • [15:10] Eliezer: English: reversed stupidity != intelligence
  • [15:10] taw has left #lesswrong
  • [15:11] Grey_Fox: either Patrick or Patrick Robotham
  • [15:11] Eliezer: okeydoke
  • [15:11] saturnnn: Roland, there is a simple answer it's just not quite simple enough for IRC
  • [15:11] LeopoldTal: That whole argument about mind-engines is complicated, and you didn't invent it on your own before puberty - so why the hell were you so sure
  • you really needed to observe stuff to form accurate beliefs, and not just your socks as opposed to dragons?
  • [15:11] Roland_LW: saturnnn point to it please? link?
  • [15:11] Eliezer: Wei Dai and the everything list have probably debated this extensively
  • [15:11] EnglishGent: I'm not sure I get what your saying Eliezer (serious)
  • [15:12] Infotropism: leopold : you don't necessarily need that to reject religion
  • [15:12] EnglishGent: I freely admit that we *might* be hypercomputers btw... it just seems to me that we _are_ not
  • [15:12] saturnnn: Roland, search OB for Eliezer's posts about quantum physics, p-zombies etc.
  • [15:12] Roland_LW: Eliezer: where do I find the everything list?
  • [15:12] Eliezer: English: in a lot of mathematical literature - I mean that in the literary sense - the computable is a tiny speck within the vast realms of the
  • uncomputable
  • [15:12] steven0461: Roland, http://groups.google.com/group/everything-list
  • [15:12] EnglishGent: if we're going to adopt a rationalist stance we cant throw out the default scientific hypothesis just becuase we dont like it
  • [15:12] Roland_LW: saturnn: I read most of Eliezer writings
  • [15:12] Grey_Fox: Eliezer: by the way, I'm sorry for that cryonics example
  • [15:12] Roland_LW: Ok
  • [15:12] Eliezer: now it appears that all of this is fantasy, but if it weren't, for all things to be computable would be quite a coincidence
  • [15:12] Eliezer: i.e. there *is* a perspective from which it is a sensible question
  • [15:13] steven0461: but I don't recommend digging in the list archives, the explanations you're looking for are probably spread out a lot and hard to find
  • [15:13] LeopoldTal: Infotropism: Can't see how - saying "belief in god isn't supported by evidence" sounds really silly if you can't explain why it requires
  • evidence.
  • [15:13] Roland_LW: eliezer: you got my point
  • [15:14] Grey_Fox: so, it might be, but it doesn't appear that way
  • [15:14] Eliezer: Leopold, it is sufficient to have read it in a book by Richard Feynman
  • [15:14] ciphergoth has joined #lesswrong
  • [15:14] Grey_Fox: good old feynman
  • [15:15] PeerInfinity: hi ciphergoth :D
  • [15:15] Roland_LW: I guess I'll write a post soon, to expose my thinking more clearly
  • [15:15] Eliezer: (if you're a kid)
  • [15:15] smoofra: i'm coming in the middle of this, but "all things computable"??? as in non-turing-computable things being computable? in the physical univerise?
  • is that what's being hypothesized?
  • [15:15] PeerInfinity: thanks for setting up the Less Wrong wiki :)
  • [15:15] EnglishGent: smoofra - it started as 'are we computers'
  • [15:15] Roland_LW: smoofra: yes
  • [15:15] EnglishGent: and went on to 'are the laws of physics computable?'
  • [15:15] EnglishGent: as the latter implies the former
  • [15:15] Infotropism: Because belief in God, being unsupported by evidence, is supported by something else. Something that feels and looks suspiciously like wishful thinking. That and the fact religion had been proven wrong often in the past, was what made it difficult in my case, to believe in it.
  • [15:15] ciphergoth: PeerInfinity: thank Badger for starting us off
  • [15:15] Grey_Fox: I'm interested in the rationality thing
  • [15:16] dfranke: Eliezer: I sense there may be a connection between the locality of physical laws and the idea that what computers can calculate are pretty much the same as what the universe can calculate.
  • [15:16] Roland_LW: EnglishGent: No! It started as the latter and then became the former
  • [15:16] Grey_Fox: not rationality, Solomonoff's lightsaber
  • [15:16] LeopoldTal: Eliezer: So if you had read a book by Aumann instead, you'd be religious?
  • [15:16] EnglishGent: did it? I cant claim to have read the whole scroll - but that's where I came in
  • [15:16] Grey_Fox: from a certain vantage point, it's a solved problem
  • [15:16] Tiiba has quit IRC ("http://www.mibbit.com ajax IRC Client"?)
  • [15:16] jimrandomh: This, here, is what Eliezer was talking about when he said IRC is a great big black hole of time wasting
  • [15:16] ciphergoth: I have time to waste right now
  • [15:16] jimrandomh: Not only won't this discussion go anywhere, it won't be preserved in a way that lets it be built upon
  • [15:16] mib_983n4c has joined #lesswrong
  • [15:16] Grey_Fox: yeah, I'd like to get a mafia game underway
  • [15:16] ciphergoth: Jess is home in a couple of hours
  • [15:17] Roland_LW: EnglishGent, I started this topic, I think I know. ;)
  • [15:17] dfranke: Eliezer: (I don't have much follow-on for this... I'm just pulling on a thread and hoping somebody sees what I mean and provides some insights)
  • [15:17] EnglishGent: anyway - I've stated my position - yes we are, and yes they are - I'm happy to consider alternatives but 'mights' arent one of them ... the bible *might* be literally true & all evidence to the contrary has been faked by an omnioptent being
  • [15:17] EnglishGent: I cant *disprove* that - but I certainly wouldnt accept it as a good argument
  • [15:17] Infotropism: I have a question, a very simple one
  • [15:17] Grey_Fox: just because something isn't a contradiction doesn't mean it's plausible
  • [15:18] EnglishGent: the moment one starts allowing completely unsubstantiated 'mights' - one is from a rationalist viewpoint in very murky water
  • [15:18] Eliezer: English, again, the problem is that there does exist a perspective from which "noncomputability" is not a specific burdensome thing like Jehovah, but actually constitutes the infinite majority of the possible
  • [15:18] Eliezer: I just think that the explanation for this is, "Those infinite realms are literary inventions, like Oz" but the perspective exists
  • [15:18] dfranke: i.e., if physical laws are local in nature, then you don't get any extra new magic by invoking the power of the whole universe.
  • [15:18] Infotropism: As a child, did any one of you just, notice that childish beliefs like santa or monsters under the bed, felt exactly like the belief in god or an afterlife ?
  • [15:18] Eliezer: Not I!
  • [15:18] ciphergoth: Eliezer: is there something that's like Solomonoff induction but allows for a super-Turing universe?
  • [15:19] EnglishGent: if there's some evidence that isnt completely convincing - well there's *some* evidence - but if there's no evidence beyond 'I dont like that conclusion' then really there's no evidence whatsoever
  • [15:19] Grey_Fox: Infotropism: I was raised an atheist, so I couldn't comment
  • [15:19] vnesov turns off the Soul-Sucking Black Hole-Powered Hoover Vacuum Cleaner, changes the filter, scrubs off the soulshatter, and turns it on again.
  • [15:19] dfranke: you ought to be able to do the same thing with a more modest device, just for smaller input.
  • [15:19] saturnnn: maybe we shoud have an informal opprobrium against arguments that go on for more than a couple minutes with no obvious progress
  • [15:19] Eliezer: Ciphergoth: Solomonoff induction is already incomputable but it would have analogues for any hypercomputers. In an oracle-1 universe you just use oracle-2 Solomonoff induction
  • [15:19] EnglishGent: that goes for *any* hypothesis btw - you cant simply deny global warming by saying 'well it might not be true' or evolution by saying 'well there might be an intelligent designer'
  • [15:19] EnglishGent: it becomes a universal cop-out otherwise
  • [15:19] tarbo2 has joined #lesswrong
  • [15:20] saturnnn: Infotropism, yes that's exactly how I thought of God as a kid
  • [15:20] ciphergoth: saturnnn: feel free to change the subject...
  • [15:20] Eliezer: yes, if anyone has a subject change just use it
  • [15:20] Grey_Fox: the mafia game?
  • [15:21] ciphergoth: Eliezer: there's a hierarchy mirroring the ordinals, I guess? So for any source of priors there's an orderly universe it doesn't allow?
  • [15:21] Grey_Fox: I think the bot is up now
  • [15:21] stillflame: ya.
  • [15:21] jimrandomh: >>help
  • [15:21] Eliezer: cipher: in that sense of "orderly", yes
  • [15:21] jimrandomh: !help
  • [15:21] -neverwrong:#lesswrong- !start --- Begins a game of Werewolf.
  • [15:21] -neverwrong:#lesswrong- !daystart / !nightstart --- Force day/night start, as opposed to the default behavior of randomly choosing one (using again returns to default behavior).
  • [15:21] -neverwrong:#lesswrong- !randomroles --- Randomize which special roles are present in the game (though roles disabled via the next command stay disabled).
  • [15:21] stillflame: !help
  • [15:21] -neverwrong:#lesswrong- !toggle<role> --- Turn a role on or off (note: Village Idiot is simply "Idiot", some roles cannot be toggled).
  • [15:21] jimrandomh: !start
  • [15:21] -neverwrong:#lesswrong- !rolelist --- Lists all the special roles
  • [15:21] -neverwrong:#lesswrong- !start --- Begins a game of Werewolf.
  • [15:21] -neverwrong:#lesswrong- !daystart / !nightstart --- Force day/night start, as opposed to the default behavior of randomly choosing one (using again returns to default behavior).
  • [15:21] ciphergoth: Grey_Fox: you programmed a bot for Mafia?
  • [15:21] -neverwrong:#lesswrong- !randomroles --- Randomize which special roles are present in the game (though roles disabled via the next command stay disabled).
  • [15:21] -neverwrong:#lesswrong- !toggle<role> --- Turn a role on or off (note: Village Idiot is simply "Idiot", some roles cannot be toggled).
  • [15:21] Grey_Fox: ciphergoth: no
  • [15:21] -neverwrong:#lesswrong- !rolelist --- Lists all the special roles
  • [15:21] jimrandomh: No, downloaded one
  • [15:21] Grey_Fox: stillflame: has one up and running
  • [15:21] -neverwrong:#lesswrong- {jimrandomh} has started a game of werewolf!
  • [15:21] LeopoldTal: Infotropism: They don't. Belief in god goes with religious devotion, which makes you feel a very special kind of good (I'm trying to harness it, half-fake religions like Discordianism work but are not organized and ritualized enough); belief in Santa doesn't.
  • [15:21] -neverwrong:#lesswrong- A game of Werewolf has begun! You have {60} seconds to register. You can register by typing "/msg Wolfbot join" and you will be added to the game. You only need to send the message once, but don't worry if you accidentally send it twice, it won't cause the bot to crash or anything.
  • [15:22] EnglishGent invests in silver
  • [15:22] EnglishGent: :)
  • [15:22] -neverwrong:#lesswrong- {ciphergoth} has joined the game!
  • [15:22] jimrandomh: The flood protection may be a bit of an issue
  • [15:22] stillflame: make that /msg neverwrong
  • [15:22] -neverwrong:#lesswrong- {Grey_Fox} has joined the game!
  • [15:22] -neverwrong:#lesswrong- {Eliezer} has joined the game!
  • [15:22] -neverwrong:#lesswrong- {LeopoldTal} has joined the game!
  • [15:22] ciphergoth: I was thinking I should write such a thing. I should have anticipated that it would exist.
  • [15:22] -neverwrong:#lesswrong- {dfranke} has joined the game!
  • [15:22] neverwrong has quit IRC (Excess Flood?)
  • [15:22] stillflame: aw...
  • [15:22] stillflame: nope.
  • [15:22] Infotropism: Even if we don't have conclusive evidence one way or the other for the existence of a god, we have evidence for the fact that people are making this stuff up and then rationalize it away. Why bother with speculation about god, if that means forgetting the whole story was playind pretending from the beginning on anyway
  • [15:22] stillflame: well, we need one that implements the Paranoid rules, anyway.
  • [15:23] ciphergoth: someone explain how we map paranoid debating onto a standard Werewolf but?
  • [15:23] Roland_LW is away, be back later
  • [15:23] jimrandomh: We don't, we would've just played Werewolf
  • [15:23] ciphergoth: tell me more about this one?
  • [15:23] stillflame: who wants to facilitate the debate?
  • [15:23] Infotropism: I see what you mean leopold, but there's still something at the core of both religion, and santa, which feels the same
  • [15:23] mib_983n4c: I personally tended to link god with my mother's mental illness which made it easy to reject. I have no memory of my beliefs regarding santa as a child. The idea of an afterlife felt a little different to me for whatever reason.
  • [15:24] mib_veqt8v has joined #lesswrong
  • [15:24] LeopoldTal: Infotropism: We do have evidence. Problems 1) most people who say "evidence supports no-god" don't understand why evidence matters 2) religious devotion has a bunch of benefits and doesn't need weird beliefs to work.
  • [15:24] Eliezer: I think the main thing I linked religion with was "adult stupidity"
  • [15:24] ciphergoth: wow. how on Earth do you decide who to lynch over IRC?
  • [15:24] jimrandomh: You private-message a bot
  • [15:24] ciphergoth: I was brought up by two ex-Irish-catholic atheists
  • [15:24] mib_983n4c: The Jehovah's Witnesses that my mom invited inside *MY HOME* really irked me.
  • [15:24] Eliezer: there are two of those?
  • [15:25] saturnnn: By age 2 I was quite aware that my parents were liars so I never really believed in santa or god
  • [15:25] mib_veqt8v has quit IRC (Client Quit?)
  • [15:25] ciphergoth: jimrandomh: oh, no public voting?
  • [15:25] MichaelGR: you have memories from age 2?
  • [15:25] saturnnn: yes
  • [15:25] jimrandomh: Having not gotten it working, we don't know the details
  • [15:25] ciphergoth: we don't have enough players yet anyway
  • [15:26] GuySrinivasan has joined #lesswrong
  • [15:26] LeopoldTal: I'm pretty sure people like, say, Eliezer never experienced religious devotion - or we'd have a bunch of atheist churches around.
  • [15:26] jimrandomh: No, we have a ton of players
  • [15:26] vnesov: I linked religion with "stuff that doesn't matter".
  • [15:27] dfranke: try registering the bot's nick. IIRC that increases the flood limit.
  • [15:27] stillflame: it was registered.
  • [15:27] timewarp: Who knew the chaos of Efnet has benefits
  • [15:27] ciphergoth: can channel ops remove flood limits
  • [15:27] Stormoog_: religion feels like an easy solution to a hard problem
  • [15:27] Stormoog_: santa is not
  • [15:27] rwallace: or have it do all communication by individual messaging?
  • [15:28] LeopoldTal: Stormoog_: That problem would be?
  • [15:28] Stormoog_: the meaning of life
  • [15:28] Stormoog_: the origin of the universe
  • [15:28] Stormoog_: what is Good?
  • [15:28] Stormoog_: to name a few
  • [15:28] komponisto: infotropism: I agree about Santa; when I was about 10 I was the equivalent of a "learned theologian"
  • [15:28] abd_uz_zahir has joined #lesswrong
  • [15:28] saturnnn: can the bot be adjusted to send its messages more slowly?
  • [15:28] komponisto: I believed in belief in Santa
  • [15:28] Eliezer is now known as EliezerWatches
  • [15:29] Stormoog_ is now known as Stormoog
  • [15:29] abd_uz_zahir: what up, wrongsters?
  • [15:29] Grey_Fox: so what's up with the bot?
  • [15:29] stillflame: saturnnn: i tried ajusting the delays by 20%
  • [15:29] Grey_Fox: try making it 50%
  • [15:29] ciphergoth hears the sound of the bot's engine turning over and failing to start
  • [15:30] vnesov: Stormoog: right, that's different. The question was about what's common.
  • [15:30] Grey_Fox: I'll just admin
  • [15:30] Grey_Fox: anyone who wants to play, pm me
  • [15:30] LeopoldTal: People mostly use religion to get a kick out of it - religious devotion is pleasurable, healthy, and increases willpower - not sure how the weird beliefs got into the mix (the weird rituals are helps to focus).
  • [15:30] EnglishGent: mib_983n4c - why? I've invited them inside on several occasions... if they want to discuss theology I'm happy to
  • [15:30] jimrandomh: You're going to have a lot of work on your hands, Grey_Fox
  • [15:30] Grey_Fox: ok, ciphergoth jimrandomh and dfranke so far
  • [15:30] EnglishGent: I just dont agree with them
  • [15:30] ciphergoth: Grey_Fox: if you're admining, does that mean we can play paranoid debating instead of werewolf?
  • [15:30] Infotropism: I actually remember having a conversation with a school peer about santa when I was about 8, about whether or not it existed, and doubting my answer. I stopped believing in God as soon as I learned about it, at 6.
  • [15:30] Grey_Fox: ciphergoth: sure, just explain the rules to me
  • [15:31] EnglishGent: :)
  • [15:31] Grey_Fox: actually, I think this would go better in a seperate room
  • [15:31] Grey_Fox: try #paranoiddebating
  • [15:31] smoofra: i know paranoid debating, what's werewolf?
  • [15:31] AdeleneDawner has quit IRC ("http://www.mibbit.com ajax IRC Client"?)
  • [15:31] vnesov: LeopoldTal: those are standard justifications, I'n not sure they are actually used as reasons to join a religion.
  • [15:31] Infotropism: the reason was that believing in god had consequences, on my life choices, while santa belief was unimportant. But settling the god question had very practical consequence on the life choices I'd have to make
  • [15:31] mib_983n4c is now known as taiyo
  • [15:32] LeopoldTal: vnesov: Well, they're the reasons why I'm going to join a religion. Now to find an organized religion that accepts atheists.
  • [15:32] Stormoog: FSM :)
  • [15:32] EnglishGent: there's more evidence to support the santa hypothesis ... I mean santa (supposedly) leaves tangible presents... and does so *today*
  • [15:33] komponisto: Leopold: unitarians
  • [15:33] abd_uz_zahir: Robert M. Price is an atheist Anglican
  • [15:33] EnglishGent: where as all of god's miracles are conviniently thousands of years ago & have never been repeated since
  • [15:33] Stormoog: That depends.
  • [15:33] LeopoldTal: Stormoog: Not organized/ritualized enough.
  • [15:33] Grey_Fox: ok, so the game is being conducted in #paranoiddebating
  • [15:33] Stormoog: I saw Maria in a dream a few weeks ago.
  • [15:34] Stormoog: It/she felt very tangible.
  • [15:34] Stormoog: I have stored it in my memory as a dream.
  • [15:34] Stormoog: Probably would have been very different had I not had an atheist upbringing.
  • [15:35] Stormoog: More tangible than santa's presents; I have never actually seen santa.
  • [15:35] Yvain: You know, my religious friends used to think atheists were people who sat around all day talking about how stupid religion was and how they were so smart for rejecting it. I told them that was an offensive mischaracterization.
  • [15:35] Yvain: ...and then I started hanging out with other atheists :-/
  • [15:35] taiyo: It isnt plausible to you that the benefits you're pointed out happen to people who join religion organizations who have all the trappings of irrationality and inanity that come with religious belief?
  • [15:36] LeopoldTal: Yvain: I did meet many atheists like that. Turned me off atheism initially, until I found the post about minds as engines.
  • [15:36] smoofra: Yvain: hear. hear.
  • [15:36] EnglishGent: Yvain - sadly some atheists do seem to radiate that attitude... I'm honestly not sure if Dawkins is doing atheism any favours at all
  • [15:36] vnesov is now known as vnesov_watches
  • [15:36] EnglishGent: but then ... many theists are just as bad
  • [15:36] EnglishGent shrugs
  • [15:36] EnglishGent: :|
  • [15:36] GuySrinivasan: Has anyone here played the Battlestar Galactica board game?
  • [15:36] KajSotala: I have
  • [15:36] KajSotala: It was fun
  • [15:36] GuySrinivasan: It bears a lot of similarities to paranoid debating IMO... :)
  • [15:37] Yvain: How do you play?
  • [15:37] GuySrinivasan: In essence decide how to allocate private resources to solving problems, with some unknown number of people trying to sabotage effectiveness.
  • [15:37] MattSimpson has quit IRC ("http://www.mibbit.com ajax IRC Client"?)
  • [15:38] Infotropism has quit IRC (Read error: 104 (Connection reset by peer)?)
  • [15:38] vnesov_watches is now known as vnesov
  • [15:38] GuySrinivasan: Everyone can take actions on their turn, everyone can discuss which is "best".
  • [15:40] LeopoldTal: komponisto: You sure? Looks like unitarians actually believe (that they believe) there is a bearded guy sitting in the sky.
  • [15:40] EnglishGent: <deadpan>GuySrinivasan - I thought that game was called 'goverment'</deadpan>
  • [15:40] GuySrinivasan: 0-25% players are secretly (or not) on the other side - half way through that turns into 50%
  • [15:40] wpearson has joined #lesswrong
  • [15:40] smoofra: LeopoldTal some of them do, some don't.
  • [15:40] stillflame: LeopoldTal: never was taught to me that way in UU sunday school.
  • [15:40] Yvain: EnglishGent: :)
  • [15:40] saturnnn: LeopoldTal, what about buddhism?
  • [15:41] Infotropism has joined #lesswrong
  • [15:42] Yvain: Speaking of games: I keep hearing references to Civilization on LW. How many other people play?
  • [15:42] LeopoldTal: saturnnn: The popular version is a personality cult around Buddha based on Hinduism, the sophisticated version doesn't get you the kick of religion.
  • [15:42] Infotropism: played non multiplayer civ since the first one
  • [15:42] wpearson: I played Civ2 the most.
  • [15:43] EnglishGent: <deadpan> Yvain- allgedly .. most of the planet ... though I'm *really* not sure about the reality.... </deadpan>
  • [15:43] ErikMesoy: Me, but I mostly lurk LW proper.
  • [15:43] EnglishGent sighs
  • [15:43] EnglishGent tries to combat his inner cynic
  • [15:43] EnglishGent fails :|
  • [15:43] saturnnn: LeopoldTal, they have no problem with atheism though, right?
  • [15:44] LeopoldTal: saturnnn: The popular version does (it's not popular in the West, anyway)
  • [15:44] EnglishGent: "where there is conflict between science & buddhism - buddhism will have to change" -- the Dali Lama (current)
  • [15:44] steven0461: it always annoyed me in civ2 that "Future Technology N" didn't actually do anything
  • [15:44] KajSotala: Yvain: I'm playing Civ4 as we speak.
  • [15:44] Stormoog: except for the points
  • [15:44] Yvain: KajSotala: I know *you* play, I see you at the Fall From Heaven forums all the time.
  • [15:44] saturnnn: I could never quite get the hang of Civ... the computer would always completely crush me. :(
  • [15:45] Stormoog: @steven, did they change that in later incarnations?
  • [15:45] steven0461: no idea
  • [15:45] KajSotala: Yvain: *blinks, looks at your nick* Oh, right, I hadn't made the connection :)
  • [15:45] ErikMesoy: Not in Civ3, but in Civ4 Future Tech is +1 happy and +1 sanitation, iirc.
  • [15:45] Yvain: Kaj: Actually, there is no connection. Total coincidence.
  • [15:45] KajSotala: :)
  • [15:46] EnglishGent is playing Civilisation 2009 -- it's a massively multi-player version with approximately 6 billion players....
  • [15:46] EnglishGent: :)
  • [15:46] Yvain: I'm not much of a Leaves fan anyway. Empyrean forever! Rational debate and wise counsel FTW!
  • [15:46] wpearson: I didn't realise that Shane that I met at the LW meetup in London was the Shane Legg that worked with Marcus Hutter on AGI.
  • [15:46] wpearson: Until he said.
  • [15:47] KajSotala: The Empyrean feels a bit, I dunno... they're a bit too good for a dark fantasy setting, I think.
  • [15:48] wpearson: So what have I missed discussion wise? Anything interesting?
  • [15:48] Yvain: Description of Battlestar Galactica and Civilization. Also, apparently people here don't like religion.
  • [15:49] Yvain: And there's paranoid debating possibly going on in a #paranoiddebating channel
  • [15:49] wpearson: The BSG board game is an interesting cooperative multi-player game, a bit like werewolf but more complex.
  • [15:50] Stormoog: Almost midnight here: time to go to bed. It has been a pleasure.
  • [15:51] Stormoog has quit IRC ("Ex-Chat"?)
  • [15:51] wpearson lurks and tries to write an essay
  • [15:51] wpearson is now known as wpearson_lurk
  • [15:52] Grey_Fox: sign up for the next paranoid debating session at #paranoidnext
  • [15:53] paulp_ has joined #lesswrong
  • [15:54] MosaSaur has left #lesswrong
  • [15:56] EliezerWatches: Shall we declare this chat at an end?
  • [15:56] Yvain: Seems like the humane thing to do at this point.
  • [15:56] taiyo has quit IRC ("http://www.mibbit.com ajax IRC Client"?)
  • [15:57] EliezerWatches has quit IRC
  • [15:57] Grey_Fox: aww, we had some paranoid debating happening
  • [15:57] Grey_Fox: oh well
  • [15:57] Grey_Fox: next month I guess
  • [15:57] Yvain: See you.
  • [15:57] Yvain has quit IRC ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]"?)
  • [15:57] PeerInfinity: hmm, I should have asked where I should start the chat log: at the beginning of the official chat, or including the stuff from before?
  • [15:58] GuySrinivasan has quit IRC ("http://www.mibbit.com ajax IRC Client"?)
  • [15:58] PeerInfinity: for now I'll include just the official chat
  • [15:58] saturnnn: why not include everything?
  • [15:58] PeerInfinity: no reason not to, I guess...
  • [15:59] saturnnn: i'm interested to see what i missed
  • [15:59] wpearson_lurk has left #lesswrong ("Leaving"?)
  • [15:59] ErikMesoy: On a related note, should be a more visible note somewhere that chatlogs will be published at so and so location?
  • [15:59] PeerInfinity: yes, there should be a more visible note
  • [15:59] ErikMesoy: *should there
  • [15:59] LeopoldTal: Yes.
  • [15:59] PeerInfinity: unless our policy is that we will only post chat logs if we have everyone's explicit permission
  • [16:00] PeerInfinity: this time we got everyone's implicit permission, and a few people's explicit permission
  • [16:00] saturnnn: better to ask forgiveness than permission, i say :)
  • [16:00] PeerInfinity: :)
  • [16:00] ErikMesoy: Besides that, there's also findability for those wanting to get at the logs.
  • [16:00] Roland_LW: Ok folks, bye
  • [16:00] LeopoldTal: saturnnn: No. Released information is public forever.
  • [16:00] Roland_LW has left #lesswrong