Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Comment author: Viliam_Bur 26 February 2015 03:13:27PM *  53 points [-]

This is how I understand it:

  • At some sufficiently high meta level, all people believe "I am trying to do the right thing".
  • But people have widely different models of world; therefore their "trying to do the right thing" may result in almost anything, including many things that we consider wrong.
  • Therefore, we are likely to conclude that our enemy, at his high meta level, believes "I am trying to do the wrong thing." Because this makes our model of him simpler for us. But that is not a realistic model. And having a wrong model of reality is potentially dangerous to our goals.

Generally, I think I agree with this. The question is, how specifically high is the level where people are trying to do "the right thing". Here I can imagine that people who didn't have experience with (conventionally called) evil people, can underestimate the necessary level of abstraction.

For some people, in their model of the world, "the right thing" includes e.g. torturing the nonbelievers. Not only because of some reasons that we could consider palatable -- for example, when someone burns a witch, we could say "well, in their model of the world, the witch is able to bring a lot of human suffering, and probably already did, so... while I think their model of the world is stupid, I can on the abstract level empathise with the concept of 'doing the right thing by killing someone who causes a lot of suffering'." No, this is actually very shallow thinking; something that naive people imagine, and that expert Dark Arts religious apologists like Chesterton make them believe. This is not a credible model of a religious fanatic.

For a credible model one has to go yet a few levels deeper. One has to imagine someone who would torture the nonbeliever because he believes that torturing nonbelievers per se is a good thing. Not because of some consequentialist reasoning based on wrong models... this is trying to push our thinking into someone else's head. No, there are people who believe that torturing nonbelievers is intrinsically the right thing. To explain why, it would be like to explain why having an orgasm is pleasant. It obviously is, that's the whole answer. This is how some true believers think, whether they are Nazis, Communists, Muslims, Christians, etc. They may also have some rationalization for why the stuff they consider right will have good (from their point of view) consequences, for example burning the heretics may appease the angry God. But those are just rationalizations. Some people burn heretics because burning the heretics is the right thing to do. Tell them they are wrong, and they will consider you insane.

And yes, one level higher, both these people and the happy hippies of San Francisco are trying to do the right thing.

Maybe the values are not all in the mind. Some values are on level of feelings, associations, reflexes. Orgasm is good because because it feels good. Eating chocolate feels good. Hearing a lullaby feels good. With the right kind of upbringing, the idea of burning the heretic will feel good. Not because we have a specific "heretic burning" sense receptor, but because the parts of the brain containing the idea of burning the heretics were connected by neural pathways to the pleasure centers, just like all associations are created. Given some upbringing, values like this can be hardwired. (The human hardware is not read-only. Associations in human brain develop as we live.) People from different cultures and subcultures can be wired differently, so they may perceive as inherently pleasant the things we abhor, and vice versa. They may feel genuine discomfort from the idea of not burning the heretic.

So what does this mean? Did "doing the right thing" become completely meaningless? Can it predict anything; explain any observation? I don't think it is completely meaningless. Specifically, I believe that some cultures require more brainwashing, some less. (Although I have no specific methodology for measuring the amount of brainwashing. Even the environment can be a component of brainwashing; what exactly would a "neutral" environment look like?) Some cultures are more reflectively coherent than others. Some cultures promote better models of reality. Giving better information and more intelligence to humans would destroy many cultures. It's just... the predictions of this hypothesis are less straightforward than a naive reader would imagine.

Comment author: Fhyve 02 March 2015 09:06:18AM *  3 points [-]

Burning cats is another good example. Can you feel how much fun it is to burn cats? Some people used to have all sorts of fun by burning cats. And this one is harder to do the wrong sort of justification based on bad models than either burning witches or torturing heretics.

Edit: Well, just scrolled down to where you talk about torturing animals. Beat me to it I guess...

Comment author: jimrandomh 28 February 2015 09:56:22PM *  29 points [-]

(EDIT: I thought about this some more, produced a better solution, submitted it as a review, and also posted it here)

  1. Voldemort knows that Harry possesses an altered version of the Patronus spell, and something else that affects dementors, but not the spell's nature. Harry can buy time by offering to explain how these work, by doing so, and negotiating for more names. Harry can truthfully say that learning certain things about the Patronus and about Dementors will have side-effects that Voldemort may not want, and that this means he needs to think about what to say. Voldemort will surely exert time pressure, but can only speed this up by so much if he wants to fulfill his goal of gaining all of Harry's secret powers. He can also try to convince Voldemort to let him cast his modified patronus; this is very unlikely to work, but should be done anyways because seeming to not try any tricks would itself be suspicious.

  2. A winning strategy should simultaneously disable all of the death eaters present and Voldemort himself. Voldemort will be disabled if their magics touch, especially if that touch can be sustained.

  3. The main weapon at Harry's disposal is partial transfiguration. I spoke first of buying time, because we also know that transfiguration takes time, although we don't know all the details of how the time requirement is determined or where in the process there is first external evidence of a transfiguration in progress. Harry has already practiced transfiguring thin cross-sections through things, and used this power in anger twice (in Azkaban and against the troll), so he should know what he can do there, how long it will take, and when it will become visible.

  4. Harry can only transfigure something that touches his wand. So he transfigures a narrow segment starting at his wand, passing through himself into the ground, and proceeding from there.

  5. Since one turn of Harry's time turner remains, if a plan ends with Harry recovering his items, then there may be an extra invisibility-cloaked, fully equipped copy of him nearby. While there is a charm that Voldemort could have cast which would detect if that were so, we don't actually see him cast it. However, the resonance effect would reveal his presence; so instead, if Harry manages to time turn and escape, he should immediately contact and lend his cloak to Mad-Eye Moody.

  6. If Harry can get a partial transfiguration off, there are a ton of possibilities. Harry has recently learned how, in a complex transfiguration, to control the order in which parts appear. So these options are not mutually exclusive; he can perform several or all of them at once.

    • There is a time turner nearby, in a pile of objects on the floor which no one is paying much attention to. He can modify the floor under it, flipping it over.
    • There are objects transfigured by Voldemort nearby (false teeth), in the same pile; if Harry's transfiguration touches one of them, it could cause resonance. Many of the other objects in that pile will also be enchanted by Voldemort's magic; transfiguring the entire surface underneath the heap would strengthen the effect. (This is important because Harry's transfiguration can't reach Voldemort himself, since Voldemort isn't touching the ground).
    • There is a Hermione horcrux nearby, which activates by touch. What "touch" means in the context of partially-transfigured cross sections has not been nailed down, but if that does count, then Harry can cause the hermione horcrux to simultaneously touch all of the death eaters. Furthermore, he can cause it to not simultaneously touch himself, by transfiguring a bit of floor into a delay switch, and lifting up his foot after the transfiguration is complete.
    • There are weapons Harry can make with low volume and low mass. If the configuration is not too complex for the time available, he could try transfiguring a cross-section that passes through the brains of each of the death eaters.
    • Hermione is nearby, asleep. She will wake up if a little bit of her skin turns into acid.

Ideally, he would do all of these things as part of the same transfiguration; but that might be too much.

Comment author: Fhyve 02 March 2015 08:55:54AM 3 points [-]

He doesn't need to stall for time to transfigure. He could have already been doing it over the last two chapters.

Comment author: jsalvatier 07 January 2015 08:24:35PM 4 points [-]

The standard advice for the best quality/price tradeoff seems to be Victorinox knives with the fibrox handle.

Comment author: Fhyve 08 January 2015 07:59:52PM 0 points [-]

I have one of these. Can confirm, pretty good relative to other similarly priced knives I've tried, and even better than a high quality knife of the same age, when both hadn't been properly maintained.

Comment author: DanielLC 06 January 2015 08:23:23PM 13 points [-]

If you type hunt-and-peck, learn to type ten-fingered, then get an ergonomic keyboard.

Comment author: Fhyve 08 January 2015 07:56:16PM 0 points [-]

In the spirit of this thread, take a typing class. I find that taking classes are an effective way to get over motivation blocks, if that's what is preventing you from learning touch typing.

Comment author: ruelian 27 October 2014 05:13:24PM 8 points [-]

I have a question for anyone who spends a fair amount of their time thinking about math: how exactly do you do it, and why?

To specify, I've tried thinking about math in two rather distinct ways. One is verbal and involves stating terms, definitions, and the logical steps of inference I'm making in my head or out loud, as I frequently talk to myself during this process. This type of thinking is slow, but it tends to work better for actually writing proofs and when I don't yet have an intuitive understanding of the concepts involved.

The other is nonverbal and based on understanding terms, definitions, theorems, and the ways they connect to each other on an intuitive level (note: this takes a while to achieve, and I haven't always managed it) and letting my mind think it out, making logical steps of inference in my head, somewhat less consciously. This type of thinking is much faster, though it has a tendency to get derailed or stuck and produces good results less reliably.

Which of those, if any, sounds closer to the way you think about math? (Note: most of the people I've talked to about this don't polarize it quite so much and tend to do a bit of both, i.e. thinking through a proof consciously but solving potential problems that come up while writing it more intuitively. Do you also divide different types of thinking into separate processes, or use them together?)

The reason I'm asking is that I'm trying to transition to spending more of my time thinking about math not in a classroom setting and I need to figure out how I should go about it. The fast kind of thinking would be much more convenient, but it appears to have downsides that I haven't been able to study properly due to insufficient data.

Comment author: Fhyve 28 October 2014 07:18:47AM 1 point [-]

I'm a math undergrad, and I definitely spend more time in the second sort of style. I find that my intuition is rather reliable, so maybe that's why I'm so successful at math. This might be hitting into the "two cultures of mathematics", where I am definitely on the theory builder/algebraist side. I study category theory and other abstract nonsense, and I am rather bad (relative to my peers) at Putnam style problems.

Comment author: ChristianKl 08 August 2014 09:57:39AM 0 points [-]

I'll also give my pragmatism a slogan: "It's just a model."

In what aspect is your idea of pragmatism supposed to differ from general semantics with the slogan "The map is not the territory"? How about reading Science and Sanity and seeing whether that's the philosophy that you are looking for?

Instead I mean that I don't even think that the truth is a useful or coherent concept when stretched to accommodate what philosophers have tried to make it accommodate.

That's a claim about ontology not a claim about epistemology. When it comes to modern source I consider Barry Smith worth reading. He's doing practical ontology for bioinformatical problems.

Comment author: Fhyve 08 August 2014 11:49:33PM 0 points [-]

The difference is that saying there is a territory is also a model. The way I would rephrase map/territory into this language is "the model is not the data."

Comment author: Protagoras 05 August 2014 07:08:26PM 0 points [-]

As I said, I'm sympathetic to pragmatism. But I guess I'd turn the question around, and ask what you think pragmatism will improve. Serious researchers are pretty good at rationalizing how procedures that work fit into their paradigm (or just not thinking about it and using the procedures that work regardless of any conflicting absolutist principles they might have). I'm sure removing the hypocrisy would be of some benefit, but given the history it would also likely be extremely difficult; in what cases do you think it is clear that this would be the best place to apply effort, and why?

Oh, and on reductionism (and to some extent truth absolutism generally), trying to give a unified account of everything requires thoroughly exploring the connections between different realms, and there are definitely tendencies to view realms as much more isolated than they are for purposes of simplification. To take what is admittedly a small scale reductionist project rather than a global reductionist project, there seems to be a strong tendency to sharply separate the physiological from the psychological when looking at behavior, in ways that seem to hinder understanding, not to mention the ability to deal with serious problems. For example, the pointless disputes about drugs for psychological therapy that focus on the bogus question of whether the psychological disorders have a biological base (how could they not, unless perhaps we're Cartesians?) rather than the much more pertinent questions of whether they work and how they compare to alternatives. While reductionist projects that try to fit everything into a single framework are sometimes guilty of ignoring phenomena that are too complicated or insufficiently well understood to fit into the framework, it is equally true that sharply separating projects into distinct categories can drastically underestimate how much influence there is from factors outside a particular narrowly defined sphere.

Comment author: Fhyve 08 August 2014 11:46:58PM 0 points [-]

This is the best place to apply effort for my goals, because I think that there might be some problems underlying MIRI's epistemology and philosophy of math that is causing confusion in some of their papers.

Comment author: Protagoras 05 August 2014 06:45:25AM 5 points [-]

Pragmatists from Pierce through the positivists to Rorty have agreed with you that the goal is to avoid wasting time on theories of truth and meaning and instead focus on finding practical tools; they've only spoken of theories of truth when they thought there was was no other way to make their points understandable to those too firmly entrenched in the philosophical mainstream (or, even more often, had such theories attributed to them by people who assumed that must be what they were up to despite their explicit disavowals). I'm not saying all of those people agreed with you about everything (the positivists, for example, thought the fact/value distinction was a useful tool, although of course they didn't think it represented any fundamental truth about reality), but I think you greatly exaggerate your originality here. Of course, one might reasonably insist that originality is not as important as whether the theories actually are useful, but while I tend to be in sympathy with the pragmatist tradition, the fact that it has been around for quite a while without seeming to have radically triumphed over all rivals does provide some reason for doubt about the extent of its world-beating potential.

Comment author: Fhyve 05 August 2014 06:22:12PM *  0 points [-]

That it hasn't been radically triumphant isn't strong evidence towards its lack of world-beating potential though. Pragmatism is weird and confusing, perhaps it just hasn't been exposited or argued for clearly and convincingly enough. Perhaps it historically has been rejected for cultural reasons ("we're doing physicalism so nyah"). I think there is value on clearly presenting it to the LW/MIRI crowd. There are unresolved problems with a naturalistic philosophy that should be pointed out, and it seems that pragmatism solves them.

As for originality, I'm not sure how think about this. Pretty much everything has already been thought of, but it is hard to read all of the literature to be familiar with it. So how do you write? Acknowledge that there probably is some similar exposition, but we don't know where it is? What if you've come up with most of these ideas yourself? What if every fragment of your idea has been thought of, but it has never been put together in this particular way (which I suspect is going to be the case with us). The only reason for not appearing to be original is so not to seem arrogant to people like you who've read these arguments before.

Do you have direct, object-level criticisms of our version of pragmatism? Because that would be great. We've been having a hard time finding ones that we haven't already fixed, and it seems really unlikely that there aren't any. (I've been working on this with OP)

Comment author: ShardPhoenix 05 August 2014 08:45:56AM *  1 point [-]

If the relevant behavior of the brain is computable (which seems likely to me), isn't there then a computable algorithm that does everything that you can do, if not better? I understand if you're objecting to overly simplistic models, but the idea that there is no one single (meta-)model that is most correct seems wrong in principle if not in present-day practice.

Comment author: Fhyve 05 August 2014 06:02:25PM 0 points [-]

The computable algorithm isn't a meta-model though. It's just you in a different substrate. It's not something the agent can run to figure out what to do because it necessarily take more computing power. And there is nothing preventing such a pragmatic agent from having a universe-model that is computable, considering finding a computable algorithm approximating itself, and copying that algorithm over and over.

In response to Me and M&Ms
Comment author: StephenR 04 August 2014 02:21:04AM *  8 points [-]

There's a nice conventional categorisation of behaviour modification programmes that goes like this:

Fixed-ratio: a reward is given after a fixed number of nonreinforced responses (e.g. an M&M after every pomodoro, or even fifth pomodoro). Fixed-interval: a reward is given after a fixed interval of time (e.g. you might always set the pomodoro for 25 minutes as per convention). Variable-ratio: a reward is given after a variable number of nonreinforced responses (e.g. you flip a coin after every pomodoro to decide whether you get an M&M). Variable-interval: a reward is given after a variable time interval (i.e. you find some way to determine how long to set the pomodoro, perhaps with a lower bound).

The schedule of reinforcement you're using is left a bit vague. It looks like you're following an FR schedule but could also be doing an FI or VI schedule. But for the purposes of offering advice to people who might want to try something like this, I'll assume you're using either FR or FI.

Psychologists categorise schedules in that way because they want to study the effects of differences in reinforcement. In particular they've been interested in the effects of changes in schedules on the extinction of a behaviour. One major result from the literature (which is reported in most psych textbooks that include a chapter on learning theories) is that variable schedules (using either ratios of respondes or time intervals) are much more resistant to extinction than fixed schedules. As an example, consider a slot machine at a casino; it doesn't have a fixed ratio of 1 reward for every nth try. Instead it varies the ratio of attempts and rewarded attempts, taking advantage of the much stronger reinforcement effect.

So my first piece of advice is: do not use fixed schedules. Varying the rate of reinforcement (either as a function of time or number of completed pomodoros) will help make the good habit you're trying to build stick if your pomodoros use is ever disrupted (because you're busy, you somehow forget, or whatever).

Another result from the literature is that ratio schedules produce higher response rates. This occurs because faster responding increases the likelihood of being reward sooner, since ratio schedules don't depend on time but on attempts. In many situations you might want to take advantage of this and opt for a VR schedule (say if you wanted to encourage a child to behave). In this case, though, it would probably only lead to extinction or abuses. Extinction because if your time intervals are somewhat long (say around 30-60 minutes), then the rewards might be given too infrequently to build your motivation and give you energy. Abuses because the big spaces between rewards might encourage you to cheat the system and eat some M&Ms anyhow because you want the energy.

That leads me to my second bit of advice: don't use a VR schedule; instead vary the time interval. I suggest finding some way to randomise the selection (like rolling dice, throwing darts, or having an algorithm spit out a number) and putting a lower bound on the time intervals (to give yourself enough time to build some flow and focus).

In response to comment by StephenR on Me and M&Ms
Comment author: Fhyve 04 August 2014 09:31:48PM 1 point [-]

Intervals and ratios are going to be essentially the same thing for conventional pomodoros. They are some time on, some time off, repeat. It might be weird to have variable pomodoros since the break is for mental fatigue, not reward. Perhaps some mechanism to reward you with an M&M at some time randomly in the second half of your pomodoros?

View more: Next