Will Eliezer Yudkowsky write a tweet containing the word "rationalussy" by the end of 2023?
45%
chance

Il miglior fabbro, FerrisHueller.

Sort by:
ButtocksCocktoasten avatar
Buttocks Cocktoasten
is predicting YES at 47%

Related

ButtocksCocktoasten avatar
Buttocks Cocktoasten
is predicting YES at 63%

Help, I am out of manna and I need more rationalussy shares

EdwardKmett avatar
Edward Kmett
is predicting YES at 46%

@ButtocksCocktoasten clearly has the right of it. Rationalussy is going to happen, it is merely a question of when. I fear that only when they are on top of the trading world, financially dominating @jack, only then everyone will see the light.

kazoo avatar
Trent Yazzo
is predicting YES at 47%

I would like to just come say welcome to the Mr. Rationalussy. May the spirit guide your right!

EliezerYudkowsky avatar
Eliezer Yudkowsky
bought Ṁ10 of NO

I herewith warn all potential traders that I may manipulate this market or bet on it at odds that do not represent my true epistemic beliefs.

jbeshir avatar
John Beshir
is predicting NO at 42%

The tipping interface doesn't allow me to be as specific as a laugh react, but it's a laugh react in spirit.

Spindle avatar
Spindle
is predicting YES at 47%

@EliezerYudkowsky THATZ THE SPIRIT

EMBRACE THE TRAUNCH
FREE THE LUSSY

MartinRandall avatar
Martin Randall
is predicting NO at 43%

A gift to the YES buyers: https://manifold.markets/MartinRandall/what-is-the-best-definition-of-rati

Sinclair avatar
Sinclair Chen
is predicting NO at 54%

I've been spending too much time on twitter lately and my brain pattern matched this to those crypto impersonator scams that reply to everything

NiclasKupper avatar
Niclas Kupper
is predicting YES at 36%

Now that Eliezer definitely knows of this market the optimal percentage is now 50%, since it minimizes his potential gains. Although maybe its a bit lower because he values other things except for M$

MartinRandall avatar
Martin Randall
bought Ṁ55 of NO

@NiclasKupper Or a bit higher because one of those other things is trolling.

EliezerYudkowsky avatar

@NiclasKupper There's also the question of my discount rate.

jbeshir avatar
John Beshir
is predicting NO at 42%

The metric to look at to determine what you can get by resolving a market however suits you, is actually not the market probability but the total amount on the order book in each direction that you could bet against given your current liquidity, since that's the amount available to acquire by resolving the market against that direction.

("you could bet against" == if you can't actually buy through a large limit order at >95% or <5%, since you need >20 times the amount of the limit order, you only count the amount of bet you're actually able to take)

jbeshir avatar
John Beshir
is predicting NO at 42%

@jbeshir The 50% thing holds iff the order book is empty and you're only trading against the AMM, modulo discount rates and actual preferences, though.

Yev avatar
Yev
is predicting NO at 49%

@jbeshir That also assumes the AMM is symmetric around 50%. This one is centered around 45% because most liquidity injections happened below 50%.

MartinRandall avatar
Martin Randall
bought Ṁ100 of NO

@Yev Good point. For anyone who wants to validate: I moved the % to 46% and now the market details popup says:

Pool: 705 YES, 880 NO

BionicD0LPH1N avatar
Bionic
bought Ṁ30 of YES

@MartinRandall Because of the overwhelming amount of YES compared to NO limit orders, the pool is pretty much irrelevant right?

Relatedly, prosocial people will remove their YES limit orders because it incentivizes an absence of rationalussy tweet, and because they most likely will lose money (assuming Eliezer acts according to his incentives). (More specifically, ~M$7000 can be made in the NO direction and only ~M$1000 can be made in the YES direction)

Though in any case the discount rate is probably much stronger of a consideration. I’m guessing it’s worth waiting 1 extra year for 7x profits? Depends on the opportunity cost of money I suppose..

MartinRandall avatar
Martin Randall
is predicting NO at 45%

@BionicD0LPH1N you think a deleted rationalussy tweet is a positive externality?

EliezerYudkowsky avatar

Question for clarification: If I write a tweet like that, but don't post it, does the market resolve YES?

JulianH avatar
Julian
bought Ṁ60 of YES

@EliezerYudkowsky that would be reward hacking. But perhaps this question, in its current wording, assumes far too much yudlignment.

MartinRandall avatar
Martin Randall
is predicting NO at 40%

Is it actually a tweet if it is not posted? Hard to guess SG's ontology.

On the other hand, posted and immediately deleted seems clearly yes.

SG avatar

@EliezerYudkowsky No, you must post the tweet for the market to resolve YES.

EliezerYudkowsky avatar

@SG How about if I write it and Aella posts it for me?

SG avatar

@EliezerYudkowsky Wow, so many edge cases! I'd say yes if it is also attributed to you. If you write the post and have someone else post it as themselves without attribution, that would not count. (It's not in the spirit of "writing a tweet".)

IsaacKing avatar
Isaac King
is predicting NO at 39%

@SG Really? I don't think it should count if it's not from Eliezer's own account. The market wasn't about whether he would say it, it was about whether he would tweet it. If a friend of mine tweets a quote of something I said, that's not the same thing as me tweeting it.

SG avatar

@IsaacKing Ok, you have convinced me. @EliezerYudkowsky the tweet must be written by you and come from your own Twitter account (the one with your name on it, no alts).

EliezerYudkowsky avatar

@SG What if I don't tweet the word as text and instead tweet a picture containing the word?

IsaacKing avatar
Isaac King
is predicting NO at 40%

I'm waiting for someone to point out that "rationalussy" isn't actually a word, so even if Eliezer tweets that string of characters, this market should still resolve to NO unless that string gets added to some dictionaries before the close date.

IsaacKing avatar
Isaac King
is predicting NO at 40%

Also, a warning to @SG: I predict that if you answer "yes" to Eliezer's question about a picture containing the word, he will next ask just how distorted the characters in the image are allowed to be.

AndrewG avatar
Andrew G
is predicting NO at 37%

Well, I think the dictionary point doesn't work since words aren't a super well-defined concept to begin with, and "rationalussy" has a reasonably well-understood descriptivist definition. I think the string-of-characters argument is important in edge cases such as posting a URL along the lines of [something]?ref=rationalussy&s=5a7f98154f8cea60fc00e4b6d1e51158 where it shouldn't count.

IsaacKing avatar
Isaac King
is predicting NO at 32%

@AndrewG "Word" is a fuzzy category, for sure. Many strings fall in a grey area where it's unclear whether or not they are words. But that doesn't mean that all strings are words. "Sdmnwermnnsadcxiuvoixudsfase", for example, is not a word.

A google search for "rationalussy" returns 3 results, which are all about this market/tweet. I don't think the string has any meaningful definition; there's no possible way for Eliezer to use it as a word in a Tweet that wouldn't either be directly referring to this market or trying to give it a new definition. If I saw the string used as a word in any other context, I would have no idea what it's supposed to mean.

EliezerYudkowsky avatar

@IsaacKing The market is in fact not about what I would tweet, it is about what I would write. Like, fine for SG to say it's about tweeting too, but the market description sure doesn't use "tweet" the verb rather than "tweet" the noun. As for "word", who's to say that it means "word" in the dictionary-as-prescriptivist-legislator-of-wordness sense rather than in the regular-expressions sense, since the former sense makes the market impossible of satisfaction?

(You were totally right about my next question being how much the word needed to be easily readable / stand out within the image.)

EliezerYudkowsky avatar

(And Andrew is correct that I was going to ask about strings inside of URLs.)

MartinRandall avatar
Martin Randall
is predicting NO at 40%

@AndrewG What is the definition of rationalussy? This has been the biggest mystery of the market.

JamesBabcock avatar
James Babcock
is predicting YES at 48%

If you're going for maximum ambiguity, you should make sure to misspell it, perhaps with rare Unicode characters that look almost but not quite like English letters. Also you might write it in a format that isn't a tweet but allow someone to screenshot it as a tweet, so that it's ambiguous whether it was a tweet or not. And dictate it verbally so that it's ambiguous whether you wrote it. And do it on Dec 31, close to midnight, from an undisclosed timezone, so that it's ambiguous whether it was by the end of 2023.

JamesBabcock avatar
James Babcock
bought Ṁ500 of YES

Hey Eliezer, are you Qualy the lightbulb? Keeping in mind that Qualy wrote this tweet: https://twitter.com/QualyThe/status/1567967324925198345 and that you are committed to answer this question with "no comment": https://www.facebook.com/yudkowsky/posts/pfbid0tyaKqmc3pkCFbJf4GTMZu7v1RiauVyrkUW84dQf689b9ktuTVFMEeCSek99Qxi5Rl

MartinRandall avatar
Martin Randall
bought Ṁ10 of NO

@JamesBabcock "the one with your name on it, no alts"

JamesBabcock avatar
James Babcock
is predicting YES at 50%

Ah, well, that's slightly less fun. But maybe Eliezer could lend out his Twitter account, have someone else tweet it, and then glomarize about whether they were taking a dictation vs writing it themself? It's either that or new year's eve on an obscure island where determining its true timezone would require resolving an obscure territorial dispute between two countries.

MartinRandall avatar
Martin Randall
bought Ṁ0 of YES

@JamesBabcock True, but Eliezer is also a market creator, so he'd need to be worried that defecting on SG by confounding this market would cause SG to tit-for-tat by almost, but-not-quite, solving the alignment problem.

VictorLevoso avatar
Victor Levoso
is predicting YES at 43%

@MartinRandall Almost but not quite solving aligment is much better than what we have now.

I think Eliezer would consider SG working on aligment to make Eliezer’s markets resolve ambiguously an incentive to do it.

a avatar

@EliezerYudkowsky You can make a good profit on this market if you choose to. If you go all in you would make M$2,000.

jack avatar
Jack
bought Ṁ0 of YES

@a This can work for both yes and no. there's also a similar amount of profit on the table if he buys a ton of no and doesn't tweet

MartinRandall avatar
Martin Randall
bought Ṁ10 of NO

@jack Actually much more available from no, given current order book, so sticking with my no shares for now.

SG avatar

There's some rats in this (group) house.

There's some rats in this (group) house.

There's some rats in this (group) house.

I said, CFAR'd freak

Substack on fleek

Rationalussy

Make those epistemics weak

Yeah, yeah, yeah, yeah

You fooming with some rationalussy

Bring a calibration graph for this rationalussy

Give me all the Bayes you got for this rationalussy

SG avatar

@SG As penance for writing this, I may unlist this market...

Yev avatar
Yev
is predicting NO at 28%

@SG Can you do that? I don't see an unlist button on my markets.

SG avatar

@Yev I can do it. (We should probably make the option available to all market creators though.)

GavrielK avatar

@SG I think "CFAR'd freak" alone makes you a criminal mastermind. Like, this is brilliant, but also you should be jailed.

GeorgeVii avatar
GeorgeVii
is predicting NO at 28%

@SG Do not pass Go! Do not collect 200!
ngl I did buy more Steven after this lmao

Sinclair avatar
Sinclair Chen
is predicting NO at 36%

snort a line of amphetemines
with your wife's boyfriend's credit card

Sinclair avatar
Sinclair Chen
is predicting NO at 49%

@SG

Beat it up, my friend, make a bet
On Manifold Markets, not just yet
Put this market right in your face
(Mani)swipe your doubt with a thought experiment
Hop on the bandwagon, let's make some cash
I'll do the math while we invest
Spit in the screen, look at the prize
This market is hot, come take a bite

Tie me up with some logic chains
Let's debate, I'll take the pains
I want you to park that big brain right in this little skull
Make it think, make me shrink
At the thought of being wrong I don't judge, I don't sway
But let me tell you, I got this mind (ayy, ayy)

- gpt-mc-3

Spindle avatar
Spindle
bought Ṁ1,000 of YES

A TRAUNCHE for all seasons

Yev avatar
Yev
is predicting NO at 25%
Spindle avatar
Spindle
is predicting YES at 14%

Pampu my SHAM POO

GeorgeVii avatar

ffs, guess i really do need a notif bot if i wanna stay in leaderboard

jbeshir avatar
John Beshir
is predicting NO at 14%

@GeorgeVii I got lucky! I've been expecting to start getting scooped by automation, but it looks like no one has set it up yet. Or is watching it close enough.

GeorgeVii avatar

@jbeshir really! hmm.
btw how did you manage to make 4k this week even without this? Is it just that your second largest position is volatile. or was that somewhat realized profit?
Seemed like nothing could have really had the liquidity to get such gains (although perhaps you could snipe large limit orders if you knew where they were. I think it'd be very handy to have such a sort function to find high orderbook liquidity)

jbeshir avatar
John Beshir
is predicting NO at 14%

@GeorgeVii https://manifold.markets/Charlie/emotional-insurance-will-orlando-ci helped a lot, I've been throwing all my loans into the CEP markets which were overvalued to start with so I got some there, and I think a lot of the rest is volatility.

GeorgeVii avatar

@jbeshir Thanks. Ah yeah, looks like you were buying into massive limit orders there.
@ManifoldMarkets Would be nice to have a sort be able to find these (ofc other things are probs higher on priority list rn)

MattP avatar
Matt P
is predicting NO at 17%

@GeorgeVii Spindle's bounties are sporadic for sure

Spindle avatar
Spindle
bought Ṁ1,283 of YES

@MattP They're pretty regular if u know the pattern

juan avatar
Juan Gil
bought Ṁ10 of NO

While I'm buying NO shares, I think the move for a YES resolution is to popularize the term "rationalussy" such that Eliezer would use the term in a tweet even without this prediction market.

citrinitas avatar
Anton
bought Ṁ10 of NO

If Eliezer decides to troll the prediction markets by referencing this tweet, then I will post a series of markets asking whether he'll do things, in an attempt to control him by exploiting his desire to troll prediction markets.

This is not a threat. It's just properly following the incentives.

JamesBabcock avatar

The insider-trade bounty is currently about M$500, which I predict is not enough to compensate him for that particular indignity.

MartinRandall avatar
Martin Randall
is predicting NO at 25%

@JamesBabcock It's currently m10,000, does that change your odds?

Yev avatar
Yev
is predicting NO at 26%

@MartinRandall No, it's only M$723. Where did you get 10k?

Yev avatar
Yev
is predicting NO at 26%

@MartinRandall Wait, I can't do math. It's about M$7475.

MartinRandall avatar
Martin Randall
is predicting NO at 26%

@Yev I added up all the limit NO orders.

MartinRandall avatar
Martin Randall
is predicting NO at 26%

@Yev Badly, apparently.

Yev avatar
Yev
is predicting NO at 26%

@MartinRandall And I forgot to include the limit orders at first :D

JamesBabcock avatar
James Babcock
is predicting YES at 26%
SG avatar

"I am Eliezer Yudkowsky, sequencer of sequences, look on my rationalussy and despair."

JulianH avatar

Omfg