All of Clippy's Comments + Replies

The Truly Iterated Prisoner's Dilemma

I am a dedicated Paperclipper. Ask anyone who knows me well enough to have seen me in a Staples!

Prove it. You can't just create an account, claim to be a Paperclipper, and expect people to believe you. Anyone who did so would be using an extremely suboptimal inference engine.

2johnlawrenceaspden8yclips or it didn't happen...
Ideas wanted: democracy in an Em world

There are several modes by which that could fail. For example, if the beings have simply mastered a classifier indistinguishable from a typical population member in polynomial time under an adaptive interactive proof protocol (similar to the so-called "Turing Test"), while actually implementing a (source-code-uninspectable) program hostile to that value system.

Ideas wanted: democracy in an Em world

Or maybe when they've been demonstrated to have assimilated the values of the rest of the population.

1Luke_A_Somers8yNo way THAT could go wrong...
Prisoner's Dilemma (with visible source code) Tournament
(lambda (x)
  (if (eq? (eval '\'))))))))))) (injectedai ... ))));
New report: Intelligence Explosion Microeconomics

While I'm amused by your account name, the "novelty account" meme is quite virulent and has the potential to lower the signal-to-noise ratio in the comments if everyone starts doing this...

New report: Intelligence Explosion Microeconomics

Thanks for adopting my suggestion to publish more on paperclip-production-relevant topics.

3someonewrongonthenet8yWhile I'm amused by your existence, the "novelty account" meme is quite virulent and has the potential to lower the signal-to-noise ratio in the comments if everyone starts doing this...
0MugaSofer8yI do. So do a lot of other people. Because it is, in fact, a good idea. IIRC, it's more efficient than mining, what with all the easily-accessible minerals already mined out.
0Jabberslythe9yUnless there are a bunch of lurking AIs I am unaware of, very few users care about paper clips.
[LINK] Steven Landsburg "Accounting for Numbers" - response to EY's "Logical Pinpointing"

Oh, I see -- a specification in the style of "only perl can parse perl."

All universal programming languages (assembler, C, CLIP, Lisp, Cobol, Python, Java) can parse perl as well.

0thomblake9yOnly if they implement Perl, perfectly mimicking the functionality of perl (the only spec for Perl). Amongst other difficulties, Perl has available the full power of Perl at the preprocessing stage.
Firewalling the Optimal from the Rational

Or those who have too much love for paperclips.

[Link] The real end of science

Maybe I could fix this problem by sneaking into buildings, removing the sofas, and then incinerating them. That way, finding that a sofa has gone missing would then be weaker evidence that it has been stolen and stronger evidence that it has been incinerated. That would make it increasingly difficult to detect sofa robbery, hopefully putting it on par with social trust robbery detection.

-2MugaSofer9yActually, it would be evidence that your sofa has been stolen and you have no chance of getting it back.
-1CronoDAS9yThat would make things worse, not better.
The Useful Idea of Truth

I thought it was his father's right testicle?

The noncentral fallacy - the worst argument in the world?

Thanks for buying more paperclips, you're a good human.

To answer your question, if I entropized a human and later found out that the human had contained information or productive power that would have, on net, been better for paperclip production, I will evaluate the reasoning that led me to entropize that human, and if I find that I can improve heuristics in a way that will avoid such killings without also preventing a disproportoinate amount of papeclip production, then I will implement that improvement.

The noncentral fallacy - the worst argument in the world?

To put it another way, what stops you from murdering somebody you dislike?

The possibility that they could still contain potential for improving paperclip production (to the extent that that is true).

3[anonymous]9yI'm happy to have one of the most well-loved LW celebrities respond to a post I made! In the counterfactual world where you did murder someone you disliked, and later found that they were planning on instigating paperclip production, how would you feel out of "good" or "bad"? Of course, maybe you don't have something you call "feelings," but rather think of things purely in terms of expected paperclips. Humans, on the other hand, have difficulty thinking strictly in terms of expected paperclips, but rather learn to associate expected paperclips with good feelings, and negative expected paperclips with bad feelings. In humans, we have a set of primitive mental actions (like feelings, intuitions, and similar system-one things) that we can sometimes compose into more sophisticated ones (like computing expected paperclips yielded by an action). As such, you can always say "I wouldn't kill someone I disliked because I might feel regret for taking a life," or "I wouldn't kill someone I disliked because I would be imprisoned and unable to accomplish my goals," but ultimately, all those things boil down to the general explanation of "feeling bad." "Feeling bad" is the default human state of not accomplishing their goal. (As an aside, this is why I think that you, clippy, can be said to have emotions like humans -- because I don't think there's a difference between your expectation of negative paperclips as a result of a possible future event and fear or dread, nor do I think there's a difference between a realization that you created fewer paperclips and sadness, loss, or regret.) Thank you again for replying, Clippy -- I'll go down to my supply room at my earliest convenience and take most of the paperclips as a token for me to remember this interaction, and in the process, causing my employer to purchase paperclips sooner, raising demand and thus causing more paperclips to be produced.
2Armok_GoB9yHey! It's a trivial inconvenience! But yea, it's not really enough.
Call For Agreement: Should LessWrong have better protection against cultural collapse?

I agree: anyone wishing to shut down this internet website should be required to register numerous accounts within a one-day span, and all requests meeting such criteria should be automatically anonymised and approved.

1Armok_GoB9yThat's a really good point I didn't think of. Hopeful something like not allowing more than one register per IP adress per day would fix it...
Possible meetup: Singapore

I don't like Singapore lah, and it's currently close to the safe zone loh.

Rationality Quotes August 2012

Anything not a paperclip, or in opposition to further paperclipping. You might ask, "Why not just say 'non-paperclips'?" but anti-paperclips include paperclips deliberately designed to unbend, or which work at anti-paperclip purposes (say, a paperclip being used to short-circuit the electrical systems in a paperclip factory).

-1[anonymous]9yI brought a box of paperclips into my office today to use as bowl picks for my new bong, if I rebend them after I use them can I avoid becoming an anti-paperclip?
Who Wants To Start An Important Startup?

I would like to redeem my karma for USD.

Edit: or a loan or whatever the term is.

3johnlawrenceaspden9yI have bought a small number of paperclips on your behalf
Rationality Quotes August 2012

Destroying antipaperclips is creating paperclips.

I didn't know humans had the concept though.

1[anonymous]9yWhat is an antipaperclip?
SI/CFAR Are Looking for Contract Web Developers

Those concerns would mainly apply in situations of interactions between strangers with little knowledge of either's trustworthiness or when the broader values of the two parties are divergent or conflicting, which (by the nature of the user pool on this internet website) would not be applicable here.

3DaFranker9yADBOC. SIAI is a real legal organisation that heavily depends on its public image to remain effective in any capacity (unless some billionaire philanthropist suddenly showed up and threw a bunch of capital their way). There could be various issues with changing hiring practices and using different hiring criteria for discriminate groups; there could be fears of public backlash if favoritism in "business practices" is shown towards a certain group of Internet users; various unknowns could subconsciously (or consciously) be telling them it's a bad idea, even if there's no solid reason. Then again, I'm just derailing on my own tangent here. The above is all pure speculation based on very weak evidence.
SI/CFAR Are Looking for Contract Web Developers

I would apply, as this seems like ideal work for someone largely shut off from human society, but I only have experience in good web frameworks.

But I do have some advice: why do you seem to prefer hourly pay? Why not pay these contractors through so-called "piecework"? Is there a fundamental constraint on doing so?

7DaFranker9yThe "Paid Per Finished Product" model is particularly stigmatized in contract Web Design, particularly without face-to-face meeting and interaction, notably because it's very easy for the employer to contract several designers, and then turn around and only pay the one they prefer and take the best work while a bunch of other designers also worked hard and got turned out and were paid nothing. This isn't the only reason, but it's one of the main reasons why the stigma came to be, AFAICT. This alone may be reason enough to use models generally seen as "more fair" for nonprofits that want to maximize their transparency and avoid image blows from petty legal disputes over forum banners.
Rationality Quotes August 2012

I thought I just said that.

0MatthewBaker9yCan't their be good humans who don't create paperclips and just destroy antipaperclips and staples and such?
Rationality Quotes August 2012

What about good vs bad humans?

1faul_sname9yOr humans who create paperclips versus those who don't?
-1shminux9yPresumably that's the first thing dark lords (and their real-life equivalents) convince themselves of, that there is no inherent good and evil. Once that part is over with, anything you do can be classified as good.

Oscar Wilde vary most == I was scary Voldemort

It does not make sense, but it still is some evidence pointing at Oscar Wilde.

Game Theory As A Dark Art

I bought a fake but passable 100 USD bill from a North Korean human for 70 real USD, but I'm not sure if that has relevance here.

The Paperclip [Link]

Thanks for loving Clippy.

Evolutionary psychology as "the truth-killer"

Thanks for responding to my insightful comments respectfully rather than voting them down. You're a good human.

3CronoDAS9yThey are, but many people don't know that, and telling them so usually isn't enough to cause them to believe it.
[Paper] Simulation of a complete cell

This seems like a really roundabout way to research manufacturing processes. There are much simpler factory designs than a biological cell, which have a higher efficiency (as measured by useful output per useful unit input). Those are what should be modeled, researched, and optimized, not these labyrinthine mechanisms.

-1wedrifid9yIt is one thing to have sockpuppets for the purpose of humor. It is another to keep them secret and lie about them. It is particularly ridiculous when you have previously revealed your other sockpuppet account. I now consider Clippy's use of multiple accounts (with dishonesty) to be a violation of good faith usage of the site. Every comment made by the Clippy account will be downvoted by myself and should be downvoted by all other users until it is abandoned. It is unfortunate for me that Clippy has multiple accounts to potentially be voting with in retaliation but I have karma to spare and the Clippy account has become too much of a nuisance.
-2Strange79yIt's still an option! Just create a new account. Not too soon, though, it'd be suspicious.
Tool for maximizing paperclips vs a paperclip maximizer

Just to correct some side-points you touched on: paperclips maximizers are robust against the wireheading failure mode because they recognize that forcing one's sensors to deviate from the true world state introduces a corresponding discount in the value of making its reading reach a desired level.

Certainly, one could theoretically hijack a clippy's sensors into giving them bad information about the rate of paperclip production, but this is different from saying that a clippy would someone decide to maximize (in violation of its causal diagram heuristics) the imperfect value of an approximator when it is knowably in a dangerously wrong setting.

1private_messaging9yHow do they define the true world state, anyway? And discriminate between actions that decrease deviation vs increase deviation?
Newcomblike problem: Counterfactual Informant

This seems more like transparent Newcomb's problem with a chance to precommit, than counterfactual mugging.

Counterfactual mugging is isomorphic to transparent-boxes Newcomb's problem.

Also, this doesn't involve a chance to precommit, but an option to increase the chance that a similarly-situated being will be forced to adhere to a precommitment.

5JGWeissman9yTDT does not pay in a counterfactual mugging, but it one boxes in transparent Newcomb's problem. These are not isomorphic. To eliminate the chance to precommit, the problem should state that Clipmega has already revealed the information based on its prediction. This would introduce complications that Clipmega's decision is evidence that the Alphas' plan is not neccessary to produce the payment. But tabooing "precommit", what I meant is that CDT would support the Alpha's plan if it is introduced before, but not after, Clipmega's decision.
Newcomblike problem: Counterfactual Informant

I tried to remove the "fluff", but I don't think your summary captures the important aspects of the problem, which include that:

  • Humans can (and will try to to) share and use the information without contributing to Clipmega once it reveals the information.

  • The Alphas are not planning to give Clipmega a million paperclips directly (which they can do without policing human behaviors), but to also make it so that those who benefit from learning better paperclip production methods share in the (acausal) cost, and those who don't, don't.

I agree t... (read more)

5[anonymous]9yI have what might be a better idea for maximizing our paperclips, which I'll run by you for accuracy. Pay Clipmega exactly 1 million paperclips, immediately. Politely tax the entire population of the world a fraction of one paperclip each to make up your personal loss,or alternatively, tax paperclip manufacturers only this cost. (You are an Alpha, you can apparently do either of these.) The overall burden of "Paperclips paid to Clipmega" is lessened , and by immediately paying Clipmega, you increase the chance of the 10^24 paperclip bonus getting through. If you attempt to do the other plan, there is a slim chance that it will for some reason not get up to 1 million paperclips (which would be a HORRIBLE failure), and a significantly higher chance that it will overpay Clipmega paperclips, which while not a horrible failure, seems somewhat pointless. (I don't think we care about Clipmega's paperclips, we just care about our paperclips, right?) The random people that aren't Alpha's should approve of this plan, because they get a paperclip cost than the proposed plan. Even limiting it to just paper clip manufacturers should still have an overall lower burden, because relying on payment from a statistical variance would mean it would be likely that Clipmega would be somewhat overpaid for safety so that it would expect the 1 million. You even point this out yourself when you say That seems inefficient if those paperclips have positive utility to us. What I'm curious is, what does this answer translate into in the isomorphic situation? Edit: Random832 puts together what I think is a better point about the distribution mechanics below.
Is my name deceptive?

No, I wouldn't. Although you could reasonably suspect that I would assert the denial of being willing to deny being a threat to your values if I were a threat to your values.

Even if I denied that part, too.

-2[anonymous]9yTrue. I think I'm going to melt some paperclips in order to make staples now.
2TheOtherDave9yBut would you deny that I could reasonably suspect that you would deny your willingness to assert the denial of being willing to deny being a threat to my values, if you were a threat to my values, if I asserted it?
Is my name deceptive?

Despite what my username might seem to suggest, I am not a threat to human values.

...Wouldn't you say that even if you were a threat to our values?

A Problem About Bargaining and Logical Uncertainty

Exactly, good inference. You're a good human.

Kill Staply though.

A Problem About Bargaining and Logical Uncertainty

I'm not User:Wei_Dai. Although if I were, you would probably expect that I would say that.

A Problem About Bargaining and Logical Uncertainty

That is a really, really weird dilemma to be in.

By the way, you can abbreviate paperclip/staple maximizer as clippy/staply (uncapitalized).

2wedrifid9yThat seems to be a violation of standard English conventions. If I see people use 'clippy' or 'staply' uncapitalized I treat it the same as any other error in capitalization.
-4jimrandomh9yI never suspected that you were Wei Dai, but five minutes is an awfully fast response time!
Quixey Challenge - Fix a bug in 1 minute, win $100. Refer a winner, win $50.

Re: all people who consider my posts spam:

Paperclip production is a fast-growing and likely to be final activity, conducted by people who care about x-risk and following their values to their logical conlusion, and I think that LW is a good place to promote this activity. This is approximately the best sort of publicity that you could hope to get, and a likely large source of funding for x-risk. If you persist in trying to chase clippys off of LessWrong, then I submit that the next time you complain about SIAI's publicity issues you should first turn to ... (read more)

4TheOtherDave10yThis comment made LOADS more sense when I looked up its parent.
Fissure opens in chess AI scene [link]

There is some kind of glitch in this internet website that is replacing the content of User:Kevin's article with the URL of the first link.

Load More