If it's worth saying, but not worth its own post (even in Discussion), then it goes here.

Notes for future OT posters:

1. Please add the 'open_thread' tag.

2. Check if there is an active Open Thread before posting a new one. (Immediately before; refresh the list-of-threads page before posting.)

3. Open Threads should be posted in Discussion, and not Main.

4. Open Threads should start on Monday, and end on Sunday.

New Comment
123 comments, sorted by Click to highlight new comments since:
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

Does there exist a paper version of Yudkowsky's book "Rationality: From AI to Zombies"? I only found a Kindle version but I would like to give it as a present to someone who is more likely to read a dead-tree version.

Yes, it's entitled "Good and Real." The shadowy cabal behind LessWrong wrote it under one of their other pseudonyms, "Gary Drescher."

(Note: this is not actually true. But Good and Real recapitulates most of the points you'll see in the philosophy-related sequences, with less focus on the basics and more on elaborating philosophical arguments. If this is the content you want to share, it might be a good choice.)

Thank you.
Added to my Amazon wish list. Do you know of any other books one should be aware of?
There's probably some books by Dan Dennett that the LW articles that deal with philosophy of mind drew from, but I've mostly been exposed to Dennett through articles, like Intentional Systems and Eliminate the Middletoad! On evolution, essential reading is The Selfish Gene. On heuristics and biases, Thinking Fast and Slow (the first half, at least), and Dan Ariely's books are good reads. I'm not aware of anything similar to the sequences in terms of intersection of Bayesianism, heuristics and biases, and trying to teach how to think about confusing things. Unfortunately.
I really like Sean Carroll's The Big Picture as an intro to rationality and naturalism for the general public. It covers pretty much all the topics in RfAItZ, along with several others (esp. physics stuff). It's shorter and a lot less technical than RfAItZ, but it's readable and I thought it does a good job of laying out the basic perspectives.

no. print your own after buying the ebook. If you go to an office-supply store they should be able to print and bind it for you.

I am skeptical that an office-supply store will know how to paginate an ebook. Here is someone who produced a pdf and printed it at lulu for $12.
I suspect it could be cheaper if someone would print multiple copies and then sell them... but of course, that requires a volunteer, and an investment.
More crucially it would require permission.
It is CC noncommercial, isn't it? There is some ambiguity in that term, but there are certainly clear possibilities.

Eugine strikes again - this is really creating a great deal of noise and reversing any indications of salience for posts. Previously, he mainly did only one downvote, now he's doing ten at a time, if my -20 karma that appeared in the last hour for the two comments I made is anything to judge by. He seems to also not only be targeting posts he dislikes, but also specific people he dislikes, such as Elo and me. Makes it really hard to judge the quality of my posts, as who knows who actually downvotes them. Frustrating.

What are the differences between the 'big names' of higher education, in comparison to other places?

For example, I often hear about MIT, Oxford, and to a lesser extent, Cambridge. Either there's some sort of self-selection, or do graduates from there have better prospects than graduates of 'University of X, YZ'?

In a little bit of unintended self-reflection I noticed that I have a strange binary way of thinking of higher education. It feels that if I don't go to one of the top n, my effort is wasted. Not sure why.

I'm just becoming somewhat paranoid regardin... (read more)


I studied at Cambridge (which, btw, is definitely better than Oxford :-)).[1] Being in the Cambridge area when I got out of academia meant that there were plenty of jobs around that suited me. (Obviously that's a good thing in itself, but perhaps if I'd been somewhere else then I'd have moved to, say, London and had a different range of job opportunities.)

Pretty much every job I've taken I've found out about because someone at my new employer knew me. In some cases those were people who studied with me. Does that count as an effect of having been at a good university? I don't know -- if I'd been somewhere else, presumably other people would have known me, and maybe they'd have been even more impressed for want of strong competition :-). But there are lots of Cambridge people in Cambridge jobs, for obvious reasons.

The point here is that lots of things useful to my career have arisen from my having gone to a good university -- but not in the obvious way (people looking at my history and saying "oooh, Cambridge PhD, must be very smart").

[1] It occurs to me that there is a slight danger of that being taken more seriously than I intend it, so let me mention that Oxford and Cam... (read more)

Thanks for the long answer! I just looked at the Cambridge prices for overseas students and it made me feel poor. Might as well seen a 500,000 ILS debt in my bank account. I live in Israel and maybe I should study here. None of my family has any education though so I'm not really sure what to do. Do you know any universal things I should look for when considering higher education? ('Is it worth it?' sounds like a good question now..)

Yeah, Cambridge is pretty expensive. (I think the best US universities are a lot worse, but haven't actually looked at the numbers. Some or all of these places may have some kind of assistance available if you're very poor or very good or both.) The recent reduction in the value of the pound (because of all the "Brexit" hoohah) has made UK universities a bit cheaper for foreign students.

I'd hesitate to call anything universal, but I'd consider at least the following things. You've probably thought of them all already :-). Some of them are awfully hard to assess. You may be able to get useful answers to some of them from the universities themselves, though of course it may be in their interests to mislead you or to refuse to answer some kinds of question.

  • How well do the courses available match what I am actually interested in learning?
  • What possible career paths might I follow, and will going [wherever] and studying [whatever] help with them? (Be realistic!)
  • Will I enjoy my time there? (This depends on things like climate, culture, difficulty of course, interestingness of course, other people there, ...)
  • Will I meet plenty of people who will be friends, mentors, useful f
... (read more)
Correct me if I'm wrong, but your comment here gives me the impression that you are asking an awfully general question, but actually want the answer to a very concrete question: "Should I study X at a top uni abroad, any uni at home, or not at all, given that I'm good enough to choose myself but will have to make debts to study". This would be a much easier question for us to answer, especially if you tell us what X is, whether you'd want to continue with a postgrad, and maybe what you goals are for the time after your studies. It's perfectly ok not to know all of these yet, but some info would help.
Israel has great tech universities. Oxbridge and other UK universities are chronically underfunded because of regulations about how much they can charge domestic students, so they try to make up for it by charging foreign students big money. My guess is that elite US universities are much better value-for-money for foreign students.
You might be able to get financial aid or scholarships, so I wouldn't rule out an expensive university right away. If you apply and get accepted, which admittedly itself costs some money, then you could have some talks with the financial aid people. And their first answer may not be the final answer, or there may be alternative sources of funding, so you may have to repeat yourself and keep talking to them for a while before you together figure out something that could work. Some universities claim to have the attitude that they don't want anyone to not be able to attend because of financial reasons, but in practice it is hit and miss to get them to live up to it, and easy to get into debt. It may depend on the individual person you are talking to. If one person isn't helping much, a different person in the same department may help more. Sometimes people unfamiliar with the system get discouraged by the first thing someone in financial aid says to them and walk away, instead of advocating for themselves more, or exploring the problem from a slightly different angle. You may also want to look into universities that have a co-operative education program that involves paid work in between sessions of study. This won't completely pay for the costs of education, but it can help a lot.
I'm currently applying for jobs while finishing up my Master's degree, so I'm not technically in the group you are asking, but can hopefully still say something useful. Background: I've been studying Computer Science with a natural language focus, both at a relatively unknown university in Eastern Germany and at University of Edinburgh. The latter is definitely top n in the field, although it does not have the same nimbus and does not offer as much of regular 1-on-1 teaching time with profs like Oxford and Cambridge do (you can get it if you ask, but it is not a default teaching mode). I can't compare to the US because I haven't been there yet. Content comparison: I find that the courses at both universities is similar both according to content and quality. The focus is different of course, and workload is much higher in Edinburgh, probably because the degree program is only 1 year instead of 1.5 or 2 for roughly the same content. In both places I could get meetings with professors if I wanted to, although in Edinburgh there is additionally a lot of staff who checks up on us and reminds us about organisational things. Among students, there is a bigger share of really bright and enthusiastic people, and that is quite noticeable. The biggest difference here is that there is direct contact with the people who made major inventions and contributions to the field and are on top of things I actually care about. This is most important in a very narrow range of topics I want to go further. For the basics, it doesn't really matter who explains them. I currently also get very good dissertation supervision, but I cannot compare that to my old university because I wrote my dissertation there during an internship and largely with supervision from the company's research department. Job applications: I feel like being in Edinburgh gives a significant boost to job applications. In Germany, profs were willing to write recommendations on request, but did not offer interesting comp
http://siderea.livejournal.com/1261773.html?format=light useful thoughts about colleges.
There's a lot of self-selection, and the classes and extracurricular resources are therefore allowed to be geared towards smarter students, and that's nice. You'll also get more opportunities to learn about current research in your chosen field, which improves your grad school chances. A lot of the value is if you plan to get a job straight out of college, going to a top n school will have a name brand advantage (not without reason). However, controlling for smartness and research experience, I think that where you did your undergrad doesn't matter all that much for grad school.
Yes, there are differences. I talked to a person who was hiring for tech jobs in Silicon Valley, and he said that the Ivy League schools in the United States get a much better quality of training than other United States schools in the fields of engineering and computer science. For example, the Ivy League schools would have 3 hour exams where you have to show how you arrived at an answer as well as the answer. Most of the other schools had 1 hour multiple choice exams. The situation is different in other countries. In Canada, unlike in the US, engineering is a regulated profession. That means certain types of designs have to be approved by a Professional Engineer. There are rules about how to become a professional engineer. One path to becoming one involves graduating from an accredited program at a university. So every engineering program at a university is monitored by the Canadian Engineering Accreditation Board. No matter which university you go to in Canada, you'll get the 3 hour non-multiple choice exams, and very good training. For this reason, the person at the tech company liked to hire Canadians, since they have the good training, but not the entitled attitude of the Ivy League graduates. But he considered Ivy League graduates. He wouldn't even consider the non-Ivy League ones, unless they showed some other way that they actually have the skills and training, since they didn't get it at school. Another difference between Harvard, MIT, etc. and the typical Canadian university is that Harvard and MIT have huge endowment funds and many wealthy alumni and donors, so they have access to a lot more resources than most universities. Like about 1000x more money. The MIT endowment fund reached $13.5 billion in 2015. I also know someone who went to Harvard for a Masters degree in political science, and she said that after getting a degree from Harvard, she was taken a lot more seriously. People listened to what she said, and deferred to her, in a way that they
No, that is also true in the US.
What do you actually want to do with your life? There are careers like politics where personal connection that are gathered during university years are very important. There are other careers such as starting a startup where personal connections with high status people might not be central and a lot of the YC founders don't have them. Why "either or"?
I went to Melbourne University and did an exchange program to UCSD. So I have comparison. I think the distribution of the quality of teaching is sufficiently narrow that it should not play a major factor.. Depending on the job and your part of the world, personal connection might be a very important factor in carer success. It is more likely that you will would gain more, better personal connection in better university.
Wait what? How are you supposed to meet your co-founder / early employees without connections? College is like the ideal place to meet people to start start-ups with.
Meeting a cofounder is useful and college can help with that but I don't think that you have a huge advantage from being at Oxford compared to being at any other decent university.
Still not sure. My English sucks, and I should stop thinking in a binary format.
Yes, they do. There are basically three tiers: the elite (top 10-12 schools), the middle (top 50-100 or so), and the don't-bother (the rest).
This may be true if you want to go into research or in primarily reputation-based fields like politics and law. In engineering or technology, you'll be totally fine and get a reasonable job with a degree from other universities. Maybe in the US it's not worth the fees, but that's a different matter and does not apply in many countries.
To clarify, what I meant was: Are the famous, top n, or places for education do provide a substantially better outcome for their students on average in comparison to less exceptional ones?
Sure, but the question wasn't what would get you a reasonable job. The question was whether graduates from top schools have "better prospects" than graduates from no-name schools and yes, they do.
I think the real question is whether they have better prospects given their level of ability and that's harder to assess.
True, though I think the real question is "better by how much" :-) If, given the ability, the top schools provide no better prospects, then (a) The common advice of "go to the best school which accepts you" is misguided and (2) the top schools have been running a marvelously successful con for decades and even centuries.
The answer to the question "do I get anything valuable from going to a better university?" might of course differ according to what it is you value; e.g., it could turn out that the "best" ones do a better job of preparing you for academic research but are no better for your out-of-academia career prospects, or vice versa. (As it happens, I agree with you that they probably do have genuine advantages whether you're looking to maximize learning, future job prospects, useful contacts, or whatever.)
The question was also if the effort is wasted. I agree that the prospects are better at a top school, but that's not the same as "don't bother".
What about the top 13-49?
Is there a list?
It somewhat depends on the specialty, but standard college rankings will get you into the ballpark. Look at things like selectivity (% of applicants offered admission) and the distribution of standardized test scores for students.

so thats the evolutionary advantage of altruism, more sex !


Altruism predicts mating success in humans British Journal of Psychology DOI: 10.1111/bjop.12208

The link lacks a trailing l after "htm". Otherwise it won't work.

How likely am I to die from taking SSRI's?

This review says:

In conclusion, this review suggests the need for caution in the use of SSRI therapy, particularly among patients with other risk factors for bleeding and those taking NSAIDs. The estimated rate of hospitalization for upper GI bleeding in the US during the 1990s was ∼ 155 per 100,000 population per year [46]. The data examined suggests that the risk of GI bleeding may be increased 2 – 4-fold in patients taking SSRIs and 3 – 12-fold when combined with NSAIDs. However, one must recognise that the in

... (read more)
Well, I'd try to find a more accurate estimate of mortality and hospitalization for your age group; if you're younger than 30, I'd be very surprised to find the mortality rate that high. You could also take Acetaminophen instead, as it it is a pain reliever which is NOT an NSAID, and does not seem to cause any stomach bleeding, which should cut it down to the 4x margin. IANAD, though, so take that with a grain of salt, and speak to your GP if you have any particular questions about interactions. I can follow your maths, but I'm also not a stat major or anything.
Thank you.
In direct contradiction to your request, I am not helping with your maths but providing general (not specific) advice. This depends very much on the population you are looking at (age, source of bleeding, severity of bleeding, comorbidities at time of bleeding - both acute and chronic). I would suggest that if you are not fulfilling some of these criteria your bleeding mortality is much lower. It is certainly possible that an SSRI could have a statistically significant effect in cases of severe bleeding like major trauma, but you'd have to estimate the risk that you'll experience major bleeding. As Cariyaga has said, acetaminophen/paracetamol is an alternative, which doesn't have the same interaction with SSRIs. Individual doses of NSAIDs may be ok, but I don't know if there's any good data to support or deny this. Part of the risk with NSAIDs is that they affect coagulation AND directly cause bleeding (gastric ulceration). Past history or family history of clotting/bleeding disorders, gastric ulcers or significant gastric reflux increase the risk of this. Another alternative is non-serotonergic antidepressants. There are many classes of antidepressant and some may not feature the same level of evidence of this increased bleeding risk (although all feature potential side-effect profiles longer than my arms). To confuse matters, some are classified structurally (tricyclic antidepressants TCAs, tetracyclic antidepressants) and some functionally (SSRIs, SNRIs, NaSSAs, RIMAs, MAOIs). TCAs may behave functionally like SSRIs, and may confer a bleeding risk as well. I think the solution will depend on some research by yourself and some discussions with your doctor. Certainly avoiding NSAIDs is an easy first step. It is easy to get paralysed by the ocean of studies, but doing nothing may not be the right decision for your quality of life.
Thank you very much.

Anyone else worried about Peter Thiel's support for Donald Trump discrediting Thiel in a lot of people's eyes, and MIRI and AI safety/risk research in general by association?

No, mainly because Elon Musk's concern about AI risk added more prestige than Thiel had.
I think it's more likely to go the other way. There are FAR more people who pay attention to Trump and normal republican politics than are currently paying attention to Thiel and AI risk. A small fraction of these being interested in Thiel and looking into AI risk is probably going to outweigh any losses.
I am, and not just MIRI/AI safety, also for other topics like anti-deathism. Just today I read in a major French newspaper an article explaining how Peter Thiel is the only one from the silicon valley to support the "populist demagogue Trump" and in the same article that he also has this weird idea that death might ultimately be a curable disease... I know that reverse stupidity isn't intelligence, and about the halo effect, and that Peter Thiel having disgusting (to me, and to most French citizen) political tastes have no bearing on him being right or wrong about death, but many people will end up associating antideathism with being a Trump-supporting lunatic :/
So in which way are you different from someone who, say, thinks that Peter Thiel has disgusting (to him and a lot of other people) tastes in sex and so will end up associating antideathism with being a moral degenerate?
Well, I would consider it worrying if a major public advocate of antideathism were also publically advocating a sexuality that is considered disgusting by most people - like say pedophilia or zoophilia. It is an unfortunate state of the world, because sexual (or political) preference shouldn't have any significant impact on how you evaluate their position on non-related topics, but that's how the world works. Consider someone who never really thought about antideathism, open the newspaper the morning, reads about that person who publically advocate disgusting political/sexual/whatever opinions, and then learn in that article that he also "considers death to be a curable disease". What will happen ? The person will bundle "death is a curable disease" has the kind of opinions disgusting persons have, and reject it. That's why I'm worried about - it's bad in term of PR when the spokeperson of something unusual you support also happen to be considered "disgusting" by many. The same happens, for example, when Dawkins takes positions that are disgusting for many people about what he calls "mild pedophilia" - unrelated to whatever Dawkins is right or wrong about it, it does reflect badly on atheism, that a major public advocate of atheism also happens to be a public advocate of something considered "disgusting" by many. Except that it's even worse in the Thiel case, because atheism is relatively mainstream, so it's unlikely people will learn about atheism and that Dawkins defends "mild pedophilia" the same day. And btw, I'm not saying I've a solution to that problem - that Peter Thiel shouldn't be "allowed" to express his political view (how much I dislike them) is neither possible nor even desirable, but it's still worrying, for the cause of antideathism.
I think you overrate the impact of reading a newspaper article. It doesn't trigger strong feelings.
Not me. In general, I think hysterics over Trump are much overdone.
What hysterics are you thinking of, specifically? I've noticed that Sam Harris has been rather vocal about his deep concern about a possible Trump presidency, saying that it would be extremely dangerous and so on. Who else relevant to the rationality movement has been overdoing the hysterics? Or were you referring to the mainstream media?
Oh, I was talking about general, mainstream hysterics (go look at pretty much any mass media for examples), not about anything rationality-connected.
Oh okay, I see.

Is there a procedure in Bayesian inference to determine how much new information in the future invalidates your model?

Say I have some kind of time-series data, and I make an inference from it up to the current time. If the data is costly to get in the future, would I have a way of determining when cost of increasing error exceeds the cost of getting the new data and updating my inference?

Generally speaking, for this you need a meta-model, that is, a model of how your model will change (e.g. become outdated) with the arrival of new information. Plus, if you want to compare costs, you need a loss function which will tell you how costly the errors of your model are.
Unfortunately to pull this off you need to look closely to both your model and the model of the error, there's no general method AFAIK.

I bought a $1400 mattress in my quest for sleep, over the Internet hence much cheaper than the mattress I tried in the store, but non-returnable. When the new mattress didn’t seem to work too well once I actually tried sleeping nights on it, this was making me reluctant to spend even more money trying another mattress. I reminded myself that the $1400 was a sunk cost rather than a future consequence, and didn’t change the importance and scope of future better sleep at stake (occurring once per day and a large effect size each day).

from http://rationalit... (read more)

That's a question of psychology, not of rationality. I don't know the answer, though my prejudices say it probably isn't a great idea. But there's another reason why you might choose not to buy another in that situation: you may think it less likely than you did before that any given other mattress will solve your sleep problems -- so now the deal you're considering isn't "$1400 for better sleep" but "$1400 for one more attempt at better sleep that may well fail like the last one did". (That was really the deal you were considering all along, but you didn't know it then.) Also, now you're $1400 poorer. If that's a sizeable fraction of your wealth then $1400 is worth more to you now than it was before and that may very reasonably affect your willingness to spend that much on a mattress. If it's not a sizeable fraction of your wealth, it may still be a sizeable fraction of your readily accessible wealth (the rest being tied up in pension schemes, investments you can't liquidate very quickly, accounts or investments you could liquidate quickly but have a policy of not doing because otherwise you'd spend too much, future earnings[1], etc.) so you may not want to spend that much again soon. [1] Of course, if you consider something like the estimated net present value of your future earnings as part of your wealth it's much less likely that $1400 is a sizeable fraction of your total wealth thus reckoned. So this may be a case where letting the sunk cost fallacy do its thing produces a better decision than trying to ignore sunk costs, if you aren't very careful about it. I suspect there are quite a lot of such situations. Cognitive biases sometimes yield better approximations to optimal rationality than simple attempts at explicit unbiased reasoning...
The key question is: Why do you believe that this attempt at punishing will have this effect? Is that a theory that you came up with yourself as a person who isn't an expert in psychology and who hasn't read any of the relevant research or alternatively has done enough self experiments to have empiric data about how this effect will work on them?
Or, they could simply update on the fact that it did not work when they expected it to, and rightly conclude that it was more likely than they realized that spending even more money would not be worthwhile.

How likely is it that polls on happiness, subjective well-being, self-worth, subjective productivity etc. are influenced by the position of the date of the poll relative to school year? (School has dictated my plans for sixteen years, and with my kid enrolled in the kindergarten we enter the same pattern. That's half my life, optimistically speaking.)

I suppose in people whose work is built upon different seasonalities (like seashore resort employees, or long-distance delivery workers, probably?), ratings should differ from the rest of the population.

I have some questions on discounting. There are a lot, so I'm fine with comments that don't answer everything (although I'd appreciate it if they do!) I'm also interested in recommendations for a detailed intuitive discussion on discounting, ala EY on Bayes' Theorem.

  • Why do people focus on hyperbolic and exponential? Aren't there other options?
  • Is the primary difference between them the time consistency?
  • Are any types of non-exponential discounting time-consistent?
  • What would it mean to be an exponential discounter? Is it achievable, and if so how?
  • What a
... (read more)
If you separate utility discount into uncertainty (which isn't actually a discount of a world state, it's weighting across world-states and should be separately calculated by any rational agent anyway) and time preference, it's pretty reasonable to have no utility discount rate. It's also reasonable to discount a bit based on diffusion of identity. The thing that calls itself me next year is slightly less me than the thing that calls itself me next week. I do, in fact, care more about near-future me than about far-future me ,in the same way that I care a bit more about my brother than I do about a stranger in a faraway land. Somewhat counteracting this is that I expect further-future me to be smarter and more self aware, so his desires are probably better, in some sense. Depending on your theory of ego value, you can justify a relatively steep discount rate or a negative one. Hyperbolic discounting is still irrational, as it's self-inconsistent.
Thanks for that – the point that I’m separating out uncertainty helped clarify some things about how I’m thinking of this. So is time inconsistency the only way that a discount function can be self-inconsistent? Is there any reason other than self-inconsistency that we could call a discount function irrational?
Couple of things to throw in there. First, you can empirically observe real-life discount functions that you can, ahem, bet money on. For example, here. Second, with respect to "my intuition is not to discount at all", let's try this. I assume you have some income that you live on. How much money would you take at the end of three months to not receive any income at all for those three months? Adjust the time scale if you wish. In general, you can think of discounting in terms of loans. Assuming no risk of default, what is the interest rate you would require to lend money to someone for a particular term?
If I received an amount equal to the income I would have gotten normally, then I have no preference over which option occurs. This still assumes that I have enough savings to live from, the offer is credible, there are no opportunity costs I’m losing, no effort is required on my part, etc. This is the same question, unless I misunderstood. I do have a motivation to earn money, so practically I might want to increase the rate, but I have no preference between not loaning and a rate that will put me in the same place after repayment. With my assumptions, the rate would be zero, but it could increase to compensate - if there's an opportunity cost of X, I'd want to get X more on repayment, etc.
No, it does NOT assume that. Basically you have a zero discount rate for money which you don't need at the moment. How about money which you can use and want to use today, what rate will persuade you to go without?
That assumption is to make time the only difference between the situations, because the point is that the total amount of utility over my life stays constant. If I lose utility during the time of the agreement, then I would accept a rate that earns me back an amount equal to the value I lost. But if I only "want" to use it today and I could use it to get an equal amount of utility in 3 months, then I don't have a preference.
I don't think this is a useful approach. A major component of time preference is the fact that future is uncertain. If you set up the situation such that there is basically no difference between the future and the present (both are certain, known, you yourself don't change, etc.) then yes, reshuffling utility between two otherwise identical points on a timeline is something you could well be indifferent to. However that's very far away from the real life and I thought we were talking about something with practical applications rather than idealized abstractions.
It actually does have practical applications for me, because it will be part of my calculations. I don't know whether I should have any preference for the distribution of utility over my lifetime at all, before I consider things like uncertainty and opportunity cost. Does this mean you would say the answer is no?
I would say that before your current needs, uncertainty, opportunity cost, and changes in yourself the answer is debatable, that is, I can see it coming down to individual preferences. But I still don't see practical applications. For actual calculations you need some reasonable numbers and I don't see how you are going to come up with them.
No. I don't think people really mean anything very specific by "hyperbolic discounting." They are just considering a 3 period model and saying that anything other than exponential is inconsistent. The qualitative form of hyperbolic discounting---of falling rapidly at first, and then slowly---empirically matches human psychology, but I don't think people care very much about precisely what inconsistent preferences humans express.
Yeah, this is what I suspect, too. "Hyperbolic" is just a metaphor, a simple example of a curve that has the required properties. There is probably no hyperbolometer in the emotional center of the human brain.
Do you think you actually behave that way on a system one level?
I can think of example where I behaved both ways, but I haven't recorded the frequencies. In practice, I don't feel any emotional difference. If I have a chocolate bar, I don't feel any more motivated to eat it now than to eat it next week, and the anticipation from waiting might actually lead to a net increase in my utility. One of the things I'm interested in was whether there's anyone else who feels this way, because it seems to contradict my understanding of discounting.

Why car safety is not advertise as its main quality?

Tesla suffered its first fatal accident in self-driving mode after driving 130 million ml, while the average mileage between fatal accidents in the US is 90 million ml. This is presented as evidence of the safety of Tesla.

However, the safety of cars of different classes of security has 1000 times difference.

Kia Rio has one fatal accident at about 10 million ml, and Subaru Legacy has less than one per billion km (in fact zero).

The latest data on the risks of different car models is here: http://www.iihs... (read more)

Why car safety is not advertise as its main quality?

I think there's a major selection effect when safety comes into play; that is, there is a sizable fraction of drivers who do prioritize safety, they buy the cars that are reputed to be safest, and then those cars appear even safer in the statistics. (For example, there are some engineering differences about the Subaru compared to other cars, but the differences between Subaru drivers and the drivers of other cars are probably larger.)

But roughly only a half of accidents could be blamed on each car driver, so even safest driver would get only 50 per cent reduction in accident rate. Other reckless drivers could rear-end him or even t-bonned.

Many serious accidents are single car crashes (more than half here). And a lot of collisions that aren't officially your fault you can still avoid if you pay attention.

I think that if we include all driving habits, single car crashes, observation selection we will get probably 1 in 10 difference, but I still surprised by 1 to 1000 difference in fatalities between most dangerous and most safe cars. I think that also weight is very important factor: heavier cars are much safer.
Heavier cars are safer for the people in the heavier cars, and more dangerous for others with whom they may collide. This produces a sort of evolutionary arms race in the direction of larger heavier cars, making everyone less safe overall. (Other changes in car design have made them much safer over the years. Today's cars are much safer than those of, say, 30 years ago. But not, I think, because they are heavier.)
While at first I thought to agree with you, I have a counterexample. Bigger car is safer if it collides with a wall, small ones and bikes are more dangerous in this case, because they result in quicker deceleration of the driver. So longer cars are also safer, and longer car tends to be heavy ones. So if all cars will be large we will get safer driving environment. So it is easy to imagine "absolutely" safe car: it would be large, heavy and very slow (like 30 miles per hour).
Maybe bigger cars are safer if they hit walls, but it's not entirely obvious to me. The more massive the car, the faster it will still be going when its front crumple zone finishes crumpling and wall and windscreen make contact; isn't that bad? Slower driver deceleration isn't a good thing if the still-fast-moving driver starts colliding with bricks. Anyway, most car accidents don't involve hitting walls, and many car accidents endanger people other than the occupants of the cars, which means more mass => more danger.
A more massive car has more energy to dissipate but also more structure (crumple zones) to apply this energy to. The net balance is not obvious me, either. While most car accidents do not involve hitting walls, a lot involve hitting objects-other-than-cars (guard railings, trees, animals, etc.) where being heavy can be an advantage (because that excess energy that you have you dissipate into the object). As to pedestrians, the mass discrepancy is big enough to not matter -- the consequences to the pedestrian of being hit by a 1-tonne car are the same as being hit by a 2-tonne car.
The 2-tonne car may be harder to stop or steer so that it doesn't kill the pedestrian.
Being lighter is better in that case.
Sure, if you include when and where people drive as part of what you blame on them. (Safety conscious people might move to particular places, or spend evenings in, or so on, and so even if they're just as good at avoiding accidents conditioned on condition the total distribution is weighted by conditions, which they have some control over.)
Because people believe that the driver's skill is significantly more important factor in determining safety. I don't know if this belief is correct but it sounds plausible. Do insurance companies offer different prices for owners of different cars?

Not only driver skill, I'd have thought; also how they're used. The Dodge Caravan is a big thing with space for lots of people. I bet it's used mostly for transporting large-ish families around. Whether you're an expert driver or a very poor one, I expect you drive more slowly and more carefully when your spouse and children are on board.

I think part of the promise of a BMW is that it's a safe car. However rightly or wrongly 5-star safety ratings is the common measurement of safety and not fatal accidents per mile.
But isn't the fatal accident per mile is better predictive measure?
It might be or it might not be because observational studies don't tend to be good at analysing causation. In general marketing executives don't focus on using metrics that are good predictive measures. They also don't want to advertise less fatalities but more safety.
I don't think so; different types of car are bought by different people and driven differently. E.g. a person who buys a Lamborghini or Ferrari probably likes to drive fast and show off; a person who buys a Volvo probably drives a lot more carefully.

Lots of sleep science out this week, but this one purports that you need 6-8 hours sleep to flush the beta-amyloids out of the brain. They say that "Caverns" open in the brain during sleep, and cerobospinal flows in forcefully to flush out toxins.

Also don't use your smartfone as an alarm clock


It's worth mentioning that this doesn't mean that flushing out beta-amygloids is the only thing that sleep does. I think various processes happen during sleep.

This video discusses the most effective science-based strategies for communicating AI Risk to a broad audience. It focuses on issues such as minimizing the inference gap, using emotional engagement, avoiding pattern-matching to sci-fi narratives and instead pattern-matching to unemployment narratives and other topics that the audience would find realistic. It's unlisted, so you can watch and share it with others only if you have a link. Feel free to pass it on to those who you think might benefit from it.

Hi, I have silly question. How do I vote? It seems obvious but I cannot see any upvote or downvote button anywhere in this page. I have tried:

  1. looking at the top of the comment. Next to OP/TS is date, and then time, and then the points. At the far right is the 'minimize'
  2. looking at the bottom of the comment. I see Parent, Edit, Permalink, get notification
  3. The FAQ says: >you can vote submissions and comments up or down just like you can on Reddit but I cannot find the vote button anywhere near comments or post.
You need at least 10 karma points to vote (you currently have 2 points, according to your profile). Once you have 10 points you should be able to see the voting buttons. Incidentally, after a troll downvoted me from 12 to 4, I lost the ability to vote, and now I can no longer see the buttons.
Thank you, that is very helpful. I wish it is said in the FAQ, or I could have missed it. I would have upvoted you if I could.

Assuming the universe is infinite, is it theoretically possible to transform it into something finite?

Reason for asking: if this is possible, it would have important implications on infinite ethics and the value and potential trajectories of the far future.

Depends on what you mean by infinite precisely. Consider for example that the in some sense finite (0-1) interval can be transformed into the interval (0-inf) via e.g x -> 1/(1-x)-1. Or whether the infinite in some sense size of the universe can be described by some finite process (like here writing a finite representation 'inf' for something infinite..
Given an infinite universe, an infinite number of Mac's are stubbing their toes. Is it possible to alter/transform the universe such that only a finite number of Mac's are stubbing their toes? If so, how?
I'm not sufficiently well versed in physics but it might be the case that there aren't necessarily infinitely many Macs. There might only be one (for a suitable definition of what makes a person). There could be infinite variety. In that case there might still be a way to reduce that somehow but it might be more complicated than talking about merging all same elements.
Ignore all but one Mac?
Yes. But please how do you bound these elements? Or asked differently: What makes Mac Mac?
You could also simply transform everything into "mu".
But that isn't bijective. You can't recover the original structure.
He didn't ask for it being bijective.
Well. I kind assume that the set of answers he intended with his question didn't contain your answer either ;-)
Answering the question that's asked instead of giving the answer that someone seeks can increase the clarity about the nature of the question that's asked.
Sure. Choose a finite volume and forget everything else. If this example doesn't suit you, then I suggest there are other requirements you want such a transformation to conform.
"Infinite" is only well-defined as the precise limit of a finite process. When you say "infinite" in absolute, it's a vague notion that is very hard to manipulate without making mistakes. One of my university-level maths teacher kept saying that speaking of "infinite" without having precise limit of something finite is equivalent to dividing by zero.

I have been considering the potential for demographic changes due to mind uploading to be even more extreme than you might initially think. This may be caused by people who are both willing to create massive numbers of copies of themselves and who are better suited for an economic niche than anyone else is for that niche, or at least anyone else willing to make very large numbers of copies of themselves. In such a situation, it would be more profitable for a firm to hire such a person than it would be for them to hire others, which may result it that niche... (read more)

Robin Hanson's "The Age of Em" is a book about this sort of thing.
Well, there already exists a version of this -- a few orders of magnitude slower -- where some groups of people decide to reproduce faster than other groups. In real life the limiting factor is often that human children require resources, so too much reproduction may actually harm them by not leaving enough resources per child, so they are later unable to compete with those who received more resources. Another limiting factor is war, often over the resources. And yet another approach is to just let it happen and hope the problem will somehow solve itself, which actually sometimes happens (for example as groups of people get richer, they start valuing comfortable life more, which gets in the way of maximizing the number of their children). Also, sometimes the group actually wins, and then in the future its various subgroups have to compete against each other. Also, there is the fact of human sexual reproduction, which means that the winning group does not have to exterminate the losing groups; it can also assimilate them. Here is probably the greatest difference compared with the Em scenario, which is like a return to asexual reproduction.