All of Bugmaster's Comments + Replies

Right, I was thinking in the context of our Western society. But in the third world, as you said, the opposite is true: an address like "123 Main St., Sometown Somecountry" simply does not work. So it is still not the case that you need to implement a fully general address database that covers all possible cases; you only need to cover the cases that you personally care about.

I realize this is supposed to be satire, but I am not convinced that, taken at face value, this poem actually constitutes bad advice. If you have the kind of personality that would allow you to implement these instructions, chances are good that -- in our modern society, at least -- you'll live a happier life; though the world may be poorer for it.

I think it's worth noting that, yes, if you want your database of names/addresses/times/etc. to be fully robust, you need to essentially represent these items as unconstrained strings of arbitrary length (including zero).

However, in practice, most likely you're not building a fully robust database. For example, you are not solving the problem of, "how can I fully represent all of the marvelous variety of human names and addresses ?", but rather, "how can I maximize the changes that the packages my company is shipping to customers will actual... (read more)

2Lumifer
As a nitpick -- yes, he can. In the third world it's not uncommon to NOT have a working system of usual addresses and locations are typically specified as town -- local landmark -- directions from that local landmark.

This article appears to encompass most of my objections:

http://thebulletin.org/artificial-intelligence-really-existential-threat-humanity8577

I do disagree with some of the things Geist says in there, but of course he's a professional AI researcher and I'm, well, me, so...

See my response to Caspar42, below. I'll write up my thoughts and post them, this way I have something to link to every time this issue comes up...

No, and that's a good point, I should really make one. I will try to post a discussion post about it, once I get more time.

My own set of objections to AI risk does not include any of these (except possibly #7); but it's possible that they are unusual and therefore do not qualify as "top 10". Still, FWIW, I remain unconvinced that AI risk is something we should be spending any amount of resources on.

1Stuart_Armstrong
That's a very strong statement, denoting very high certainty. Do you have a good basis for it?
1Caspar Oesterheld
Is there a write-up of your objections anywhere?

Bending is allowed; see above.

Folding is allowed, yes.

Parts can have different shapes (if you want), but must have the same area.

You cannot use compasses, but you can use an unmarked straightedge if you want to make precise creases, or to avoid ripping the paper in an untidy fashion. You are not allowed to mark the straightedge, of course.

If the procedure were carried out with infinite precision, then it would indeed produce exact fifths.

In keeping with the "puzzle" theme:

You are given a rectangular piece of paper (such as the placemat at a fast-food restaurant). Without using any measuring tools (such as a ruler, a tape measure, some clever length-measuring app on your smartphone, etc.), divide the paper into five equal parts.

2Kindly
I believe I have it. rot13: Sbyq naq hasbyq gur cncre ubevmbagnyyl, gura qb gur fnzr iregvpnyyl, gb znex gur zvqcbvag bs rnpu fvqr. Arkg, sbyq naq hasbyq gb znex sbhe yvarf: vs gur pbearef bs n cncre ner N, O, P, Q va beqre nebhaq gur crevzrgre, gura gur yvarf tb sebz N gb gur zvqcbvag bs O naq P, sebz O gb gur zvqcbvag bs P naq Q, sebz P gb gur zvqcbvag bs N naq Q, naq sebz Q gb gur zvqcbvag bs N naq O. Gurfr cnegvgvba gur erpgnatyr vagb avar cvrprf: sbhe gevnatyrf, sbhe gencrmbvqf, naq bar cnenyyrybtenz. Yrg gur cnenyyrybtenz or bar cneg, naq tebhc rnpu gencrmbvq jvgu vgf bja nqwnprag gevnatyr gb znxr gur sbhe bgure cnegf. Obahf: vs jr phg bhg nyy avar cvrprf, n gencrmbvq naq n gevnatyr pna or chg onpx gbtrgure va gur rknpg funcr bs gur cnenyyrybtenz.
7Manfred
First, recruit five paper-maximizing ideal economic agents, and give them a picking order one through five. Let the last-picking agent divide up the rectangular piece of paper into five parts. Downsides: will cost you the piece of paper.
gjm
120

Could you confirm or correct some guesses about exactly what problem is intended?

  • Folding is allowed.
  • The parts need to be of equal area but can have different shapes.
  • No other tools are allowed, not even "non-measuring" ones like an unmarked straightedge or a pair of compasses.
  • You want not an approximation but a procedure that, carried out with infinite precision, would produce exact fifths.
-1[anonymous]
No bending the paper?

...but note that astrophysicists admire the night sky at least as much as lay folks, despite being able to describe in some detail how the stars shine and galaxies form. So "reasonable" doesn't mean "plain".

Quite the opposite, since the astrophysicists can enjoy the night sky on many more levels than someone who believes that stars are just little holes in the celestial dome, or something. Some of these things we call "stars" are suns (much like our own Sol), but others are galaxies or globular clusters. What sounds more gr... (read more)

I like the taste of some alcoholic beverages; but, for some weird reason, being drunk makes me feel quite dizzy -- and that's it. There are no positive effects: no social disinhibition (as far as I can tell, anyway), no warm fuzzy feelings, just dizziness. For this reason, I tend to drink rarely, and little.

Edit: I love olives, FWIW.

0Elo
I can measure my intoxication by the rate of dizzyness. anywhere from slight unsteadyness to eventually everything spins enough that its difficult to stand up because the only thing not moving is the flat floor beneath me. I recently observed someone maintain sobreity on beers for several hours and then switch to rum&coke and become slurry and have difficulty walking within 30mins (cumulative effects aside). I suspect that different alcohol has different specific effects on people. Other alcohols might not make you so dizzy but might give you the "expected social buzz". If you feel like giving it a try and letting us know? (on the condition that you shout "FOR SCIENCE" as you drink...

I was in this exact situation, and I chose to buy some headphones. If I analyze my decision-making process, I can come up with two reasons:

1). My own personal cost of buying and wearing headphones was much lower than the cost of having pissed-off neighbours who hate me. Obviously, YMMV.

2). My neighbours were polite, and even somewhat deferential, in their request (for me to stop playing loud music). They did not threaten me with coercion, despite the fact that they had plenty of coercion at their disposal -- they could've complained to the building manager... (read more)

One way is to simulate a perfect computational agent, assume perfect information, and see what kind of models it would construct.

Right, but I meant, in practice.

that is, learning from observations.

Observations of what ? Since you do not have access to infinite computation or perfect observations in practice, you end up observing the outputs of models, as suggested in the original post.

For example, how about "relying on the accumulated knowledge of others"?

What is it that makes their accumulated knowledge worthy of being relied upon ?

0gedymin
I agree with pragmatist (the OP) that this is a problem for the correspondence theory of truth. Usefulness? Just don't say "experimental evidence". Don't oversimplify epistemic justification. There are many aspects - how well knowledge fits with existing models, with observations, what is it's predictive power, what is it's instrumental value (does it help to achieve one's goals) etc. For example, we don't have any experimental evidence that smoking causes cancer in humans, but we nevertheless believe that is does. The power of Bayesian approach is in the mechanism to fuse together all these different forms of evidence and to arrive at a single posterior probability.

We don't want to be confused with the uncritically thinking masses - the apologists of homoeopathy or astrology justifying their views by "yeah, I don't know how it works either, but it's useful!";

I think this statement underscores the problem with rejecting the correspondence theory of truth. Yes, one can say "homeopathy works", but what does that mean ? How do you evaluate whether any given model is useful of not ? If you reject the notion of an external reality that is accessible to us in at least some way, then you cannot really ... (read more)

0gedymin
One way is to simulate a perfect computational agent, assume perfect information, and see what kind of models it would construct. Solomonoff induction provides a universal standard for "perfect" inductive inference, that is, learning from observations. It is not entirely parameter-free, so it's "a standard", not "the standard". I doubt if there is the standard for the same reasons I doubt that Platonic Truth does exist. Umm, no, this is a false dichotomy. There is a large area in between "relying on one's intuition" and "relying on an objective external word". For example, how about "relying on the accumulated knowledge of others"? See also my comment in the other thread.

Not to mention, the leading cause of propeller-induced face laceration syndrome...

Bugmaster
140

This may be a stupid question, but is that mosquito laser drone thing really the best way to solve the problem of... what problem is it even solving ? "Too many mosquitoes" ? "Malaria" ?

Your confusion is a clever ruse, but your username gives away your true motives!

chaosmage
110

Wiki says the idea has been suggested in earnest as one of the forms a mosquito laser could take, and was rejected in favor of a better one.

I don't think there is any quadcopter that can fly for more than 30 minutes on one battery charge - and that's without mosquito recognition and zapping systems drawing on that same battery.

4skeptical_lurker
Wouldn't a stationary laser system be simpler? At least as an initial minimum viable product?
6Evan_Gaensbauer
Laser-mounted mosquito-exterminating drones seems an uncommon and difficult-to-engineer enough approach that if anyone was serious or successful enough in even launching such an endeavor that they have a public-facing interface to raise awareness or gauge interest, I figure they'd be both impressive and rare enough Eliezer would want to at least get in touch with them. Either that, or he is trolling us with all his projects near the end, or is partially trolling us by interjecting fake interest in ridiculously ambitious projects between his real interest in other ridiculously ambitious projects. Another source of perspective is the fact that Eliezer has made research into engineering safety mechanisms into advanced machine agents a seriously substantial movement, and he did this by blogging about rationality for two years, and then writing a Harry Potter fanfiction over the course of the succeeding five years. From Eliezer's perspective, the rationality of him and his network, combined with the gusto of the surrounding community, may be enough to achieve very ambitious projects which start from seemingly ridiculous premises.
[anonymous]160

There's a much cheaper and much older flying platform for mosquito elimination. It's called a bat.

EDIT: or perhaps the bred/genetically modified steriile mosquitos that can wipe out populations in large areas?

8knb
Some perspective: From here. Basically Eliezer thinks people should work outside but don't because of insect problems (among other things.)

I don't think this is a good data point, since the makeup they wear is explicitly designed to counteract visual artifacts (glare, unnatural-seeming skin tones, etc.) that are introduced by the camera. Thus, the makeup does not necessarily have a positive effect on people who see the movie stars in person.

3zedzed
This is a wonderful data point. It moves our model from "if you're a man, don't wear makeup" to "if you're a man, don't wear makeup unless you're going to appear on camera, in which case, wear just enough to counteract visual artifacts." I expect this to be a nontrivially better model for a significant amount of men here.

Note: will still be stuck to Harry's face, so no bombs or suchlike.

Why not ? I mean, yeah, obviously Harry would want to survive; but if there was some way to take out Voldemort while also taking out himself (and possibly Hermione), and there was nothing better that Harry could come up with in 60 seconds; then the logical course of action would be to go ahead and do that.

7MathMage
Because it is unlikely that Harry would have precommitted to blowing himself up as a general emergency countermeasure.
Bugmaster
-10

I'm not sure what "having the entire plot planned out from the beginning" really means, though. Eliezer ends up retconning things relatively frequently, so I wouldn't be surprised if he had a plot point like "Voldemort captures Harry" followed by "Harry escapes", but with too few details in between to make the logic ironclad.

If HPMoR was a conventional book, then Eliezer would have a lot of time to edit it and make all the retcons behind the scenes -- even fairly major ones -- but it isn't, so he can't.

No, I was serious, sorry. But, seeing as I believe that Eliezer is human like the rest of us, I think it's entirely plausible that he ran out of ideas (or, at least, out of good ideas) -- kind of analogous to writing a program so clever that the author cannot debug it...

4dxu
True, but seeing as he has stated that he had the entire plot planned out from the beginning, for him to have run out of ideas would require him to have been actively lying to us.

I only read 3WC after the fact, so I can't comment on that one. But I don't recall him saying "...solve this problem or you get the bad ending" in he previous HPMoR chapters...

5Eugene
Yes you can. Simply look at the time stamps for each post and do simple math. By making the assumption that only "people who were there" can answer correctly, you're giving up solving your own problem before even trying.
4dxu
Well, yes, because those chapters weren't anywhere close to the ending. (Though, if I recall correctly, he did say that whoever guessed the answer to Chapter 9 would be told the entire rest of the plot, which isn't quite the same as "earn your happy ending", but it's similar.) Still, that notwithstanding, do you honestly think Eliezer is doing this because, of all things, he ran out of ideas? That seems... somewhat unlikely to me, given his previous track record. (Of course, Poe's Law is in full effect on this one, and if you were in fact joking in your original comment, then I apologize for any misunderstanding.)
6Sheaman3773
Do you not remember the Courtroom, where something similar to this was done? People speculated at the time that it was a practice run for something later. It looks like they were right.
7skeptical_lurker
Except that the next chapter will be published 2 hours after the deadline
0[anonymous]
That would be so awesome!
dxu
130

This is an extremely uncharitable reading of Eliezer's intent. Remember, he did the same thing with "Three Worlds Collide", and in Chapter 9 of HPMoR, and in Chapter 81 of HPMoR. Would you say he wrote himself into a corner in all those places as well?

They couldn't easily check every possibility and see if it's consistent.

Why not ? It's not like the laws of our space-time apply to them or anything.

Jost
180

Atlantis-Human: “Where did you learn about computability, Harry?”

Harry: “… in the Matrix.”

A-H: “The Matrix tells elegant lies.”

I don't know if he put it into exactly those terms, but a). Harry points out a lot of things that aren't true, like "you can't turn into a cat", and b). if the laws of reality are simulated, then they don't have to make sense; they could just be a giant "switch" statement somewhere in the Atlantean VM code.

0DanielLC
If the laws of reality are simulated, then they must be computable. A giant switch statement isn't going to let you figure out how to make time travel consistent. They couldn't easily check every possibility and see if it's consistent. Even if they did, that would mean they're simulating all of them, including the inconsistent ones, and there'd be no reason for Harry to find himself in a consistent one.

Either that, or the world where HPMOR takes place is just one among many realms within the Mirror; i.e., the Simulation Argument is true, and the Atlanteans are the Matrix Lords. This explain the weird and inconsistent magic rules: they are just artificial constructs that the Atlanteans came up with on a lark.

1DanielLC
Didn't Harry already point out that the time travel was not computable, and as a result it couldn't be a simulation? Although he didn't go to great lengths to prove that. He assumed that there was no force subtly manipulating events to make sure time travel is consistent. In fact, he is in a simulation run on the computer that is Eliezer's brain.

Also, nice Duane shout-out.

As well as the Mass Effect shout-out. ASSUMING DIRECT CONTROL.

Yeah, and I am kind of surprised that neither Quirrellmort nor Harry thought of reversing the letters. I mean, we are dealing with a magical mirror here. How is this not the first thing they've tried ?

1The_Duck
The inscription is not in the Latin alphabet.
2Jost
* The letters are in reverse order, but not mirrored. No wonder they didn’t think of that! * It wouldn’t fit the narrative. (Just like it didn’t fit the narrative in canon, so nobody pointed it out there, either.) * Maybe the mirror has some kind of very narrow Confundus charm, that makes a person’s brain unable to process these letters? ;)
WalterL
130

I get the feeling from Voldemort's comments and Harry's thoughts that the Letters of False Comprehension have a mind affecting power that prevents you from understanding them. I think the inversion is a joke for the readers.

Killing him would be a grave mistake.

As far as I understand, Quirrell believes (or claims to believe) that killing Harry will put him one step closer to fulfilling his CEV. Thus, any amusement Harry could provide is to Quirrel kind of like as ice cream is to us mortals: a minor, fleeting, and ultimately inconsequential pleasure.

Maybe Quirrel and Harry are just individual vectors in a massively parallel multiverse-spanning genetic algorithm, designed to produce the ultimate Tom Riddle.

Oddly enough I really like LOTR as well as The Silmarillion... So maybe I should give this Gormenghast thing another shot, I don't know.

I think the difference between LOTR/Simlarillion and Gormenghast is that Tolkien's books contain well-crafted language and descriptions of scenery that are punctuated by moments of sheer epic overload; whereas Gormenghast contains the former but not the latter.

But again, I haven't made it that far into it, so I could be wrong.

I started reading the first book, but stopped about 20% of the way in (may have been less, it's been a while since then), because I found it stupefyingly boring. Does that trilogy get any better later on ?

0drethelin
If you find it boring it's probably not for you. I enjoy the language, the descriptions and so on. It's the same as my recommendation for LOTR. I liked it a lot but I don't think everyone should. I find both pleasant to read even when nothing is really happening in the story.

Also there are great number of possibilities that even the smartest persons could not even imagine, but powerful Superintelligence could.

I think that, if you want to discuss the notion of the Superintelligent in any kind of a rational way, it is useful to make a distinction between "the Superintelligent can do things we can't", and "the Superintelligence is literally omnipotent". If the latter is true, then any meaningful discussion of it is impossible -- for the same set of reasons that meaningful discussion of the omni-everything Christian god is impossible.

I would say it's a combination of being at the wrong company, and our education system being inadequate to the task.

There are many skills that are required in order to write complex software. You need to know how to organize your code in a maintainable and comprehensible way (Design Patterns, build/package systems, abstraction layers, even simple stuff like UML). You need to know how to find bugs in one's own code as well as in code written by other people (using debuggers, reading stack traces, writing logs, applying basic deductive reasoning). When you g... (read more)

If you join a cult, then even your physical survival will suddenly become a lot more perilous. You will likely have to conform, or die. Keep that in mind.

4dxu
...Not that I know much about cults or their relationship to the law, but that seems kind of illegal.

I was wondering that too; personally, I have no idea how to even begin answering the question. It would seem that at least some protests do work, as evidenced by the civil rights movement during the Martin Luther King era; but I don't know if this is true in general.

2knb
I think protests work if there is already a critical or near-critical mass of support in the relevant decision-making body (legislature, courts, civil service, etc.) Protests rarely change minds, but they can give already-sympathetic people a new impetus to take action in this area rather than another. ETA: It also helps if the protesters have specific, focused demands, like "end segregation," or "bring the troops home."

Ok, in the spirit of rationality and openness, I have to admit: I'd love to get most of the audiobook volumes, but $50 is too rich for my blood. Is there a way you guys could introduce a cheaper option ? Something like, "All of the Sequences except for the Quantum Physics ones" for something like $25..$30 ?

7lukeprog
$50 is normal or even cheap for an audiobook of this length. Also, I'm hoping Castify can get it into the Audible catalog, in which case Audible users should be able to get it for a single Audible credit.

That doesn't cover artillery, unlike the word "gun".

Bugmaster
-40

The word "utilitarianism" technically means something like, "an algorithm for determining whether any given action should or should not be undertaken, given some predetermined utility function". However, when most people think of utilitarianism, they usually have a very specific utility function in mind. Taken together, the algorithm and the function do indeed imply certain "ethical obligations", which are somewhat tautologically defined as "doing whatever maximizes this utility function".

In general, the word "u... (read more)

jefftk
110

an algorithm for determining whether any given action should or should not be undertaken, given some predetermined utility function

That's not how the term "utilitarianism" is used in philosophy. The utility function has to be agent neutral. So a utility function where your welfare counts 10x as much as everyone else's wouldn't be utilitarian.

Bugmaster
110

I downvoted this post not because I hate you, or because I love Eugine_Nier (o), but because I'd like to see fewer post like this one in the future. And I think that expressing my sentiment is what the "Downvote" button is for.

More specifically, I don't think that public shaming and witch hunts belong on Less Wrong, even when the person being hunted is actually a witch (oo). I think that the toxic culture such tactics create is likely to be more harmful than individual unruly posters, in the long term.

(o) I don't even remember who he is, though the name does sound familiar.
(oo) Metaphorically speaking.

I think that "прикол" is closer to "amusing" than to "fun". "клёво" is more like "cool". And I always thought that "пиздец" was universally bad, something akin to "game over, man ! game over !" -- but words do change over time...

0IlyaShpitser
I have seen the word пиздец used after surviving a near miss, or witnessing a particularly daring and successful stunt (?as an exclamation of relief?). As I said, none of them are exactly right.

I think the more proper translation of "it depends" would be "как сказать".

Also, while it is true that the Russian language has no word for "privacy", note that it also has no word for "gun" :-)

0IlyaShpitser
Стрелковое оружие (firearm) (?)

It didn't do much for me :-(

If there's no way for me to figure out whether there's a chocolate cake inside of the Sun or not, then I might as well assume there's no cake, because this makes the math easier. I see MWI vs. no MWI the same way, but apparently that's the wrong answer...

0ike
What about the question of whether something not in your light cone still exists? Also, assuming something to make the math easier does not mean that it is meaningless. It may have little utility to calculate it if your utility function only counts things that affect you physically, although you'd still need the general rules for priors for things that can only be tested long term but influence decisions made short term.

That's what I thought too, and apparently I was wrong...

Bugmaster
390

FWIW, I said I "strongly disagree" with Feminism and Social Justice, even though I find their Wikipedia descriptions generally agreeable. I think in the future, it would be good to split those questions into pairs: a) "Do you agree with the stated mission goals of X ?", and b). "Do you agree with the actions of people who identify as X ?"

If we're going to bother to ask (b) at all, it's probably best to frame it in a way that doesn't make "some but not all of them" the obvious answer.

For example, perhaps you could identify some groups you consider definitive of Feminism and Social Justice, and we could ask "How often do you agree with $group?" (IIRC, on their own blog Yvain often uses something called jezebel as a metric for what feminists believe.)

You are saying that shminux is "a worse person than you" and also "heartless", but I am not sure what these words mean. How do you measure which person is better as compared to another person ? If the answer is, "whoever cares about more people is better", then all you're saying is, "shminux cares about fewer people because he cares about fewer people". This is true, but tautologically so.

0roryokane
All morals are axioms, not theorems, and thus all moral claims are tautological. Whatever morals we choose, we are driven to choose them by the morals we already have – the ones we were born with and raised to have. We did not get our morals from an objective external source. So no matter what your morals, if you condemn someone else by them, your condemnation will be tautoligcal.

I could be wrong, but didn't Darwin actually formulate some hypotheses, and then go out there and count finches (and other things) to see if his predictions were true ? I think that's why his success rate was so much better (though, admittedly, not perfect): he conducted experiments in the real world, using real math.

Also, Freud did some kind of experiments. He was not merely a philosopher, he also cured people, and it seemed to him that his theories work. But he didn't have a control group, etc.

How did he know if his theories actually worked, then ? Was he even making his patients better in any way (as compared to other patients who saw other doctors, or perhaps no doctors at all) ?

2Viliam_Bur
He was convinced that "couch therapy" worked better than hypnosis, but I don't know whether he kept records to prove it. (Sorry, I have read all this decades ago, and then I was interested in his models of mind, not in technical details. Now I know that those details are critical, but I don't remember whether I read about them or not.)

Sorry, it gets difficult to keep all the commenting systems straight after a while.

Load More