I suggest having a link to the last open thread in each open thread, and similar for the quotes thread. That way, you can just follow the link to find out what people posted near the end of the last thread, so it doesn’t become pointless if there’s going to be a new thread soon.
Besides LW, what are some other online communities with very high signal/noise ratios?
Some good indicators: Lots of original content, meaningful and well-presented information, respectful conversations with a high level of discourse, low levels of trolling/strong community norms, very strong domain-specific knowledge, people know each other by username.
Another heuristic: somewhere where you would not want to share the link with a lot of people lest the quality be diluted with newcomers. (Hopefully you consider LW a strong enough pool to draw from).
The Oil Drum, is a strong web community specific to the Energy industry. Stopped updated this past September (thanks knb, I hadn’t visited for a couple months).
Niche subreddits are often a great resource, so much so that when I’m looking for information I often do a reddit search before looking over the greater internet.
Things that wouldn’t count:
Social news sites, like hacker news or the the large subreddits, which tend to have a lot of noise.
Gwern’s Google+ feed has perhaps the single highest signal/noise ratio of anywhere I’ve seen on the web, but obviously is just the one person.
The Erowid Experience Vaults is a solid collection of crowdourced drug experiences, but doesn’t have a community attached.
As far as raw content goes: I strongly suggest a “Unified standard for print and Internet” rule. Judge content against whatever book or paper you could be reading instead. Often communities have separate incentives to promote and praise original content regardless of absolute quality. There’s also a novelty bias, where mediocre content in new mediums (youtube videos, blogs) is held to a lower standard. Since I’ve implement this rule, I’ve noticed a lot of what I considered great content is really just “great for the internet” content. I read more books now.
Against this standard, I find the most raw information value comes from individuals who explicitly collate links or summarize information from other sources. Democratic collation can be great as well, but faces too many ways to die. Long-form community and blog content is almost always overrated, except for cases like niche ideologies or at the cutting edge of some technical or business topic.
Meta Warning: I think this applies to lesswrong. There are exceptions, but the long-form posts and related blogs are often either “good for the Internet” content or decision theory stuff I don’t have the background to understand. I don’t mean to gripe: I’m very happy with LW, and I regularly skim the discussion section and read the open threads. I just think that LW over-hypes it’s own original content, like any other community.
Against this standard, I find the most raw information value comes from individuals who explicitly collate links or summarize information from other sources.
Let’s say I wanted to monitor the variation in my Big 5 personality traits over time. Is there an existing way to do this, or should I handroll my own procedure, which will probably be total BS?
Debunks the common notion that hangovers are about dehydration. The reason it caught my eye is that I believed the dehydration theory, even though I should have known that extreme sensitivity to sound isn’t a normal symptom of dehydration. (I’ve never had a hangover, but at popular accounts include sensitivity to sound and light.)
One reason for the myth about dehydration would be due to “drinking plenty of water” still being one of the most effective things to do: If it’s about the liver breaking down alcohol into toxic Acetaldehyde, drinking lots of water to flush it out.
Understandable mistake to go from “more water fixes the problem” to “problem must’ve been not enough water (dehydration.)”
This was how it was discussed in my university chemistry class. Also mentioned: a similar breakdown (same enzymes or whatnot) happens with methanol, and the breakdown products (formaldehyde and then methanoic acid) are stronger / more toxic than those of ethanol (acetaldehyde / acetic acid.)
If it’s about the liver breaking down alcohol into toxic Acetaldehyde, drinking lots of water to flush it out.
How? I’m pretty sure that contrary to popular belief, water can’t simply be used to flush stuff out, the metabolite also has to be in a form that can be excreted by the kidneys, and even then extra water might have no effect whatsoever. This is basic physiology found in any textbook on the subject.
I tried to google if kidneys might excrete some of the acetaldehyde, but found no answer.
I had reservations about including that sentence, because I only have a vague idea which completely lacks details about mechanisms. And flushing seems like a folk-explanation rather than a science-explanation.
The other vague idea was that drinking more water means the toxins are more dilute, but I have even less confidence in that.
If you check out the linked study, it doesn’t do such debunking. It found no association between hangover severity and vasopressin and it wasn’t clear what other markers of dehydration it measured. Note that measuring dehydration accurately could be difficult and different people might experience different levels of dehydration differently.
It should be the default assumption that there are many mechanisms involved in hangover, because the effects of ethanol are very complex.
Interesting—I’ve had a plenty of nights of drink, including some where I felt unwell the next day, but never had either of those symptoms. I have, however, woken up recognizably dehydrated a couple of times.
I wonder whether some of the “myth” comes from the experiences of people like me, accurately reported but for some reason not defined as hangovers by the researchers behind the study cited.
I’ve mistaken caffeine withdrawal for a hangover before, partly because of similarity in symptoms (headache; nausea; photosensitivity) and partly because it tends to show up around the same time (Sunday morning). This may account for the popularity of coffee as an alleged hangover cure.
Which raises the question, if the things people say about “hangovers” are true about the things they apply the term “hangover” to, what’s left to be debunked?
That article rubs me the wrong way. I think it may be more a failure of science, rather than the author’s personal failure though. Whenever people are curious about how to reduce hangovers, all the articles you find will talk about how abstinence is the only cure (gee, this doesn’t echo any other memes), but the fact is that there is an effective treatment, even if there haven’t been sufficient scientific studies done on it. A large social group that I am involved in, which has been known to drink heavily, has started taking N-acetyl-cysteine (500-1000mg) and Source Naturals Hangover Formula (which is primarily a C & B complex), and the effect it has on hangovers is not in any way subtle. Despite the existence of this, the article only says that the best prevention is to consume alcohol with food and lots of water. There is some scientific support for this combination (see the studies and explanation referenced here or here), but even though it is a question of significant practical importance to many people, no scientists have actually gone out and done a controlled study on humans on these nutrients.
I have been interested in the phenomenon called tulpa. (interestingly, Wikipedia sheds next to no light on this issue).
According to one site, it is an “autosuggested and stable visualization, capable of independent thought and action, while possessing its own unique consciousness”. Supposedly, following the guides found on the internet, one can create a stable, persistent “imaginary friend”, with the looks and character one wants that will be real in all aspects for its creator. Some say that tulpa can provide an alternate viewpoint or help fetch information from their host’s memory, but various hosts disagree on the possibility of this.
Looks like tulpa in modern, Western definition has no connection to its Buddhist namesake (like karma on the forums). Some enthusiasts claim otherwise, but, as seems to be characteristic of this topic, there’s no evidence.
All I could find are guides and diaries of anonymous people on the Internet. It seems like the whole phenomenon, if it really exists, was invented some 1.5 years ago by some Anonymous: there’s their own slang, and absolutely no sources that connect the methods to any actual scientific research.
I suspect that the tulpa phenomenon may either:
not work (all cases of success might be explained by belief in belief; people might be adeptly fooling themselves and their peers, thinking up their imaginary friends and persuading everyone that their actions are autonomous)
be dangerous to one’s mental health (shouldn’t success here by itself be diagnosed as some kind of personality disorder?)
which is by itself a reason not to try it myself, but I have a nagging doubt: which one of the two is true, if any?
I’ve recently talked online, including by voice chat, with people who claim to have tulpas, and quickly ran out of useful questions to ask. I asked for interesting questions to a member of LW meetup group and he said
“it would be interesting to check tulpa’s autonomy, but because only its host can report its actions, I don’t see a way to do it. Maybe if I understood psychology of personality better, I couuld think up something”
Really, what else can one do? The people that I talked to seemed surprisingly unconcerned with using methods written by Anonymouses, lack of scientific evidence, status of the phenomena with regard to psychiatry or what comes for “evidence”, “proof” or “source”.
Can anyone help me to find answers for the following questions:
What interesting questions can I ask people who claim to have tulpas?
Are methods grounded in science in any way?
What is the position of science with regard to tulpas? Surely, if the phenomena exists, science must have encountered it and have a stance wrt. it, right?
It could be a good occasion for me to learn scientific research; any help with it would be appreciated: where should I start figuring out what, if any, does science know about this?
Luhrmann wrote a book, When God Talks Back, about her experiences with evangelicals, which might be useful. She also succeeded in inducing tulpa-like visions of Leland Stanford, jr. in experimental subjects.
You may also be interested in the earlier LW discussion about tulpas. Like I mentioned in the thread, they seem like an intentionally developed version of a thing that many writers have naturally—my guess would be that they use the normal circuitry that we have for emulating and predicting the behavior of other people, only they’re modelling a non-existent person and the outputs of that modeling get fed back in to be used as new input.
Edit: Still, at a glance, the 3 questions I’m asking this time, were not exactly asked in the linked discussion. So I welcome everyone to share their thoughts here.
What is the position of science with regard to tulpas? Surely, if the phenomena exists, science must have encountered it an d have a stance wrt. it, right?
There no such thing as a position of science. Science is a process.
As far as Tuplas go, Tulpas are mainly about qualia. Qualia are by their nature but directly measureable. As a result a lot of scientists do shun research of qualia. Orthodox reductionists do try to explain qualia away whenever they can instead of trying to investigate them.
Western scientific education doesn’t train it’s scientists do be in control of their minds and that means that most they can’t run experiments that require mental control and good awareness of their own minds.
They can invite some buddhist monk and investigate how that monk does his thing, but that doesn’t allow for easy controlled experiments.
In that area even the easy questions seem to have little scientific investigation. Take heat development during particular types of meditation.
Ten years ago a scientists would have looked strangly at you for suggesting for heat development without movement but now we do know that the body can in principle do this in brown fat tissue.
There some small experiments that illustrate that a specific Tibetian technique can consistently produce heat but there no general science based theory that predicts when you would expect someone who meditate develops heat. It’s even not clear whether it’s really produced in brown fat tissue because adults have little of it.
Present thinking in medicine is rather: “How can we develop a drug that stimulates brown fat tissue to be active to help people lose weight?”
Tulpa need around 6 months of hard focused mental practice of 1 hours per day. That not something that you usually study in scientific studies.
If you look at another recent controversy on Lesswrong we even disagree whether mainstream science knows what losing weight is about. There a lot more scientific effort going into that question but it still isn’t conclusively answered.
If you are looking at far out mental phenomena there no reason to expect that they are well investigated by scientists.
Finding a person who hallucinates is pretty easy. You go to your nearest asylum and go through the patients and you will usually find someone who has hallucinations.
As luck would have it the patients are also bound for years to a specific location and might have no possibilty to opt out of your study.
Finding people as test subjects who spent halve a year doing a hard mental practice is harder. Experiments that require that you have test subjects who spend a lot of time on a hard mental practice are much easier to do.
At a glance, science seems pretty well informed about hallucinations.
The page you linked to doesn’t provide evidence that indicates that science is well informed about the issue. It doesn’t illustrate that scientific theories are able to make reliable predictions about hallucinations.
One of the examples about which the wikipedia article talks is a unreplicated 13 person experiment with 5 days duration. It talks about is as “strong support” for an idea.
It says “There are few treatments for many types of hallucinations.” You can translate that into the acknowledgement that the phenomena isn’t well enough understood to effectively modify it in the way you want.
The third way to check whether someone understands something is to check with your own empirical experience. I unfortunately don’t have much experience with hallucinations that go beyond things like the optional illusion where every normal viewer hallucinates that wheels turn.
I do have some experiences I had after spending 5 days in an artificial coma. One of them is a state where what I see visually doesn’t change when I close my eyes. I know of descriptions of other people who experienced the same thing.
Can you find a mainstream science description of that visual hallucination?
Someone who works at google told me the company is working on trolley problems because self-driving cars may have to make that sort of decision, and google will be responsible.
Dumb reinforcement question: How do I reward the successful partial-completion of an open ended task without reinforcing myself for quitting?
Basically I’m picking up the practice of using chocolates as reinforcement. I reward myself when I start and when I finish. This normally works very well. Start doing dishes → chocolate → do dishes → finish doing dishes → chocolate. It seems viable for anything with discrete end states.
Problem—I’ve got a couple long term tasks (fiction writing and computer program I’m making) that don’t have markers, and I can put anywhere from 30 minutes to 3 days into them without necessarily seeing a stopping point. I’m worried that rewarding chocolates whenever I get up from working will (in the long run) reinforce me to quit more frequently. I don’t want to end up with a hummingbird work ethic for these tasks.
How should I reinforce to maximize my time-on-task?
(So far my best plan is to write a smartphone app that creates a hidden random timer between 5-55 minutes (bell curve) that goes off, and I reward myself chocolate if I’m on task when the alarm activates. But there’s logistical hurtles and it seems like quite a bit of work for something that might be solved easily otherwise. Plus, I don’t know what possible bad behavior that might incitivize.)
Why does it need to be a hidden random timer? Reward yourself if you stayed on task for the past 30 minutes. (Hmm, I think we’ve just reinvented the Pomodoro Technique.)
Incidentally, have you (or others who use schemes like this) considered using intermittent reinforcement? Like, instead of just rewarding yourself upon meeting the victory condition, you flip a coin to see if you get the reward. It seems the obvious thing to do if you’re going for the whole inner pigeon thing.
Hmm, reward myself after a fixed interval of 30 minutes? That’s just crazy enough to work! (I have heard of the Pomodoro technique before, and I’m not quite sure why I didn’t just go for that at the start.)
The hidden random timer is to make myself resilient to extinction and ingrain the habit even without reward. Although, randomly choosing to reward at the end of pomodoros would work too. IIRC, intermittent time interval is the reward structure that survives the longest without extinction, whereas a variable ratio reward structure creates the most vigorous workers.
Also, I think what you describe is a conditional reinforcer and not an intermittent one. What I mean by that is after a long enough time, they subject would become attached to the coin flip itself as a partial reward. Kind of like clicker training for animals, or how a shot at a jackput pull is a reward even when it doesn’t payout. Then you could use the stronger conditional training systems...
Your suggestion is brilliant. Aaaand now I’ve got “write a gamblerdoro app” on my to-do list.
You might like http://HabitRPG.com—there are already a few of us from LW on there. If you join it, tell us your user ID so we can invite you to our party :)
I’ve noticed that I have low priority in at mid-large group conversations. What I mean is that in situations where I’m one of two people talking, I’m (generally) the one who stops and the attention of the “audience” (people-who-aren’t-speaking) is predominantly on the other person even before I stop speaking.
This used to cause me considerable distress, but no longer. I’ve accepted it as a fact of the social universe. But I’m still curious and would like to change it, if possible.
I suspect that this is something that varies by social group, and more strongly suspect that some behavior of mine is key.
I’m interested in (being pointed to) discussion of this type of thing, especially if it contains actionable advice.
I can’t really offer anything more than a personal anecdotes, but here is what I usually do for when I try to grab attention of a group of my peers:
If you are talking to several people gathered in circle, and it is my turn to say something important, I make a small step forward so that I physically place myself in the center of the group.
When I am speaking, I try to mantain eye contact with all people gathered around, If I focus too much only on the person I am speaking to, everyone else turns their attention towards them as well.
I rarely do it myself, as I suppose it is a technique more tailored for public speeches, but conservative use of hand gestures to signify what you are talking about, probably won’t hurt.
I probably sound like a self absorbed jerk writing this, but if I want the attention to focus on myself, and not my interlocutor I often use “me” language. Compare and contrast [“What you say about vegans is true, but you may conisder...”—now everybody looks at the person who said something about vegans] [“I think that I agree with what was said about vegans, but I also think...”—now everybody looks at me as I explain my position].
But those are all just little little tricks, when the surest way of attracting attention of the audience is simply to have something important and interesting to say.
I probably sound like a self absorbed jerk writing this, but if I want the attention to focus on myself, and not my interlocutor I often use “me” language.
While we are at that topic many people use “you” when talking about themselves. They say sentence like: “Yesterday I thought: You should go to gym.”
I once even listened to someone who used “he” to when speaking about himself a few years ago. The language was German and he was an Austrian, but it still signified how little he identified with his self in the past.
After a bit of prodding he changed to “I”. That also changed subtle things about his body language did change. It was interesting to watch the effect. Identifying with oneself helps to be more charismatic.
It’s one of those nontrivial aspects of: “Just be yourself.”
Thanks. These are things I’ve learnt or tried learning in the past. I’d guess there are good odds that I’m reverting to past (shyer) behaviors in some situations.
I’ll make an effort to be aware of my body language and focus next time.
You’re really not giving enough specific information. There are countless reasons why that might happen so any advice you take here could lead to erratic behavior.
In addition to what other people said, this sounds like you might be too verbose or bad at gauging which topics interest other people and to what extent. They might look at the other person because they wish them to interrupt you and move on.
Fair enough. At this stage I’m curious as to which specifics I should be looking at. Or what kinds of things are key (to speaker priority in groups of 5-10).
The various elements of body language given, and your notes on content (I can be too verbose, for sure) have given me what I need to go on for now.
If you start talking loudly, perhaps a little louder than you think you should be talking, attention will usually shift to you. If you say something relatively short and relatively interesting, you gain some credit, and then later people will be more likely to listen to you. This requires some timing—you have to start talking just as the previous speaker is finishing, or even a little before. Mistiming could make you look like (and be) a shouting idiot.
If two or more people use the same strategy (starting off loud), then you can use the opportunity to appear (and be) gracious by telling the other person to go ahead. This also subtly puts you in a position of control; they got to talk, but you decided who should talk.
(This is all based on personal experience, of course; YMMV.)
The first would be that it’s just an issue of perception. There a plenty of people with low self esteem who hold inaccurate beliefs about how much attention other people pay to them.
A second would be generally proxies for low testosterone. If you have a louder and deeper voice people are more likely to listen to you.
There are also body language changes and various other things that are hard to fake.
When someone speaks in a way that strikes you as unusually ambiguous, consider that it may be intentionally so to avoid placing emphasis on irrelevancies.
When someone speaks in a way that strikes you as unusually ambiguous, consider that it may be intentionally so to avoid placing emphasis on irrelevancies.
What makes you think I didn’t? If he’s in the habit of intentionally sabotaging himself that increases the value of having it being pointed out to him.
I wouldn’t highlite it if he wouldn’t have specifically asked for ways to draw more attention on himself. The more of your attention you concentrate in withholding information the less likely people are going to give you attention.
It’s not like it’s wrong to do so, but you should be aware of the tradeoffs that you are making.
It’s not obvious that attention is being diverted to obscuring information. Maybe “partner” is the lower-effort term. It is for me: “boyfriend” and “girlfriend” have weird connotations and baggage that make me want to pause before using them and wonder if I’m really being accurate.
It’s not obvious that attention is being diverted to obscuring information.
It makes it harder for me to build a mental image of the situation that he’s describing. If it’s harder than I will put less attention on the situation.
Here it’s a subtle choice. But it points to a pattern. A common general piece of advice on telling stories which draw listeners attention is to provide a lot of adjectives to make it more easy for the audience to picture what you are describing.
The kind that most people don’t put to practice when they hear it. Instead of describing the principle abstractly, I pointed to an example.
It takes effort to provide listeners with details. It’s still one of the pure white hat strategies to getting peoples attention when you speak.
As far as the information being relevant, when I give someone recommendation about testosterone, the gender of the person I’m interacting with matters.
I have a good idea that more testosterone in males will help with given attention. I’m less certain, that trying to increase testosterone is a good strategy for females.
I don’t know whether you refer to a business partner, you are male and refer to a girlfriend or you are female and refer to a boyfriend.
You omit at least two possibilities: that he is male and referring to his boyfriend or that she is female and referring to her girlfriend. In these cases, word “boy/girlfriend” would have you interpreting the situation wrongly.
As others have commented, the fact that we do not know these unnecessary details is a feature, not a bug, of ungendered words.
In this case, it’s deliberately non-gendered language. Lower-effort, as kalium says. In my case because I cultivated the habit, in years past.
As both you and Douglas_Knight point out, there are tradeoffs involved. In the case of not gendering pronouns I expect I’ll continue thinking it worthwhile.
But it’s a helpful thing to consider- I’ll bet there are other habits I’ve developed that I’ve never considered if it’s worth the costs. Especially when I contrast my teenaged self – “I don’t care what anyone thinks of me” + “I’ll choose my words for my own aesthetical pleasure” – with the me of today – who does care, and on balance values communication higher than self-expression. I doubt my conversational habits have shifted as far as my preferences have.
I’m also interested by what you say about details. It’s not something I’d’ve thought of, worrying I tend more to being too verbose. But I like to write, and concrete detail / description is the area I’d consider my weakest there. I can think of some ways to practice this (started playing tabletop RPGs recently, for one.)
When it comes to details, the thing that count is whether the details help the person you are talking with to form a picture of the situation that you are describing in their mind.
If you are talking in “I” using colorful language that describes the qualia you perceive is nearly always good. That doesn’t mean that sentences should be long. If you can transform one sentence into two, that’s often good.
If you are talking in “you” it better to be a bit more vague. If you tell someone: “When you go to work on Monday morning at 7 o’clock and drive though rush hour, you know the feeling where you wish, you could just take a day off?”, the might be irritated if they aren’t in the habit of going to work at 7 o’clock or don’t drive.
I”m okay with non-gendered language. In text it’s nearly impossible to see what the words mean for you.
If your teenage self trained a bunch of separate word choices I would look carefully at them and judge whether they are real self-expression or whether they are more of a mask that’s supposed to provide shelter.
Authentic self-expression draws attention, wearing a mask reduces it.
Usually when people intentionally manipulate emphasis like this, they are trying to change relevance, not merely acknowledging existing relevance. Maybe it is worth the personal cost to achieve the political end, but it is not possible to make the tradeoff without being honest about the cost.
I think body language tends to be pretty relevant here; it’s possible you are not noticing the other person’s body language indicating that they’re about to start speaking. (And of course, if you want to have it be your turn to speak, using body language to signal this is also relevant.)
Cosmologist and science popularizer Sean Carroll debates Christian apologist William Lane Craig on Feb 28. The topic is God and Cosmology. My prediction: while Sean Carroll is very good, I don’t expect him to beat a professional (and a very successful) debater.
I don’t expect him to beat a professional (and a very successful) debater.
The word we’d need to better define is “beat”.
Will WLC appear to win the debate? Likely. Will he appear to overwhelm his opponent? Likely. Will he succeed at framing the debate in a way that suits his position? Overwhelmingly likely.
The debate takes place in front of his home crowd (an Evangelical theological seminary). WLC will gallop his way to “victory” again, methinks.
Because he won’t actually win anything, anymore than a rabbit actually disappears inside a top hat or a woman actually gets sawed in half and restored.
Debates are of limited value in this way: The correct argument can rather easily lose. WLC is a great example of why and how.
If he follows Chris Hallquist’s advice of insisting on speaking first and gives a talk as good as this one with edits to suit Lane Craig, I predict it would go down pretty well.
However, I assign a low probability to this happening.
I teach high school English to underclassmen who skew towards “totally unmotivated”. I have been using spaced repetition principles for years (using games, puzzles, and other spaced reviews) to help with vocabulary and terminology. These do effectively engage many of the poorly motivated.
But recently, I feel like smartphones have become ubiquitous enough among students that I’m looking for software I could use as a quasi-official SRS companion app with my students. I think many of them would use it, but only if they experience very minimal frustration setting it up and running it. My wishlist:
(1) Free app on both Android and iPhone (I’d say it’s about 50⁄50 with my students)
(2) Companion web app with cloud sync to mobile apps.
(3) Very easy to use and update with new cards regularly. I would like to be able to post weekly deck additions on my teacher web page that students can add to their deck.
Anki, which I use for my personal learning, seems to come closest—but the $25 cost of the iPhone app is a problem, and I worry that using the web app on the iPhone be too much of a hassle. I also worry that the “add external cards to your deck” procedure is a bit too hairy as well.
Has anyone seen anything that comes closer to my needs than Anki? Thanks!
It has an app, it has a lot of the bells and whistles that Anki lacks (like a scoring/gamification system) that could be helpful with the population you are teaching, and it is all around a solid SRS system. The only thing I think it lacks are those Easy/Good/Hard buttons that Anki has to differentiate between how well you know the answer, but that’s something I can live without. I use both it and Anki on a day to day basis.
I just played around with Memrise, and it does indeed look perfect for my audience. I had begun my SRS search with gwern’s excellent exploration of the topic, where Memrise does not appear. Thank you so much!
I’m glad you like my recommendation. After you have used it for a while, perhaps consider writing up a post about your experiences teaching using an SRS. It’s a topic which could be very interesting, and I’m sure that many would wish to read such a report. I certainly would.
I’ve started making heavy use of archive.is. You give them a link, or click their super-handy bookmarklet, and that page will be archived. I use it whenever I’m going to be saving a link, now, to ensure that there will be a copy if I go looking for it years later (archive.org is often missing things, as I’m sure we’ve all run in to).
Today in Nature, evolutionary biologist Owen Jones and his colleagues have published a first-of-its-kind comparison of the aging patterns of humans and 45 other species. For folks (myself included) who tend to have a people-centric view of biology, the paper is a crazy, fun ride. Sure, some species are like us, with fertility waning and mortality skyrocketing over time. But lots of species show different patterns — bizarrely different. Some organisms are the opposite of humans, becoming more likely to reproduce and less likely to die with each passing year. Others show a spike in both fertility and mortality in old age. Still others show no change in fertility or mortality over their entire lifespan.
That leftmost of the five graphs, the one that shows in its shaded area the proportion of the hydra sample that still lives, confuses me. What could possibly cause a linear decrease in population? A constant probability of death given previous survival (podgps) would produce a decreasing exponential curve, but this requires a podgps for each individual that is exactly hyperbolically growing over time, for example through an internal candleclock that is at birth set to a uniformly distributed time between 0 and 1400 years.
(Yes, hyperbolically. As the time interval lengths during which we record whether the hydra died go to 0, the ratio of podgps between the last time segment and the, say, 3rd goes to infinity. The crux of the matter lies in “given previous survival” and if we leave that part out, the result would indeed be a constant line.)
Maybe the researchers found by experiment or hearsay that the hydra has a constant podgps, the graphmakers failed to record that last bit and simply took the negative integral of that absolute probability of death to generate that ridiculous population curve and then the authors found that that integral curve hits 0 at 1400 years and postulated that as their livespan.
(By the way, it’s also troubling that they also left out the bit of the curve that would have indicated infant mortality.)
See the label for Figure 1. I’m not sure why army1987 retracted his comment, but he is correct: the y-axis is logarithmic for the survivorship curve. So the graph actually confirms your expectation and shows an exponential decrease in population.
(Unfortunately, the graph label in the National Geographic article is just wrong—there is no reasonable interpretation under which the logarithmic survivorship curve can be interpreted as a raw proportion.)
On the last panel (that for hypericum) of the figure on the NatGeo page the red curve doesn’t look like the negative derivative of the gray curve, so I assumed I was missing something.
Is the Iron law of oligarchy essentially a Goodhart’s Law applied to humans? Like: You want a group of humans to accomplish something useful, so you create a system to resolve conflicts, e.g. a democratic majority vote. Sooner or later people learn how to win the majority vote by optimizing for winning the majority vote, without accomplishing much of what you originally wanted them to do. -- And if you try to fix this by adding some safety mechanism X to the democratic vote, then people will simply optimize for the majority vote plus X. For example in addition to elected politicians known to optimize for popularity, you add unelected bureaucrats who are supposed to be the experts, but somehow those just entrench themselves in the bureaucratic system regardless of their level of expertise.
If so, then essentially there is no safe way to solve this. If we measure something, then Goodhart’s Law attacks. If we don’t measure something, then… well, just because you are not looking at something, it doesn’t mean it’s not there… in the absence of explicit rules, the implicit rules will decide; the most popular people will simply be the most popular people.
All we can do is to use are some heuristics, and remember the nameless virtue; i.e. to change or abandon the heuristics when they stop being reasonable. We must keep thinking and updating, and thinking and updating, again and again.
Specifically, I have already noticed the Goodhart’s Law in action on StackExchange. Instead of helping other people it’s more and more about getting more points than other contributors. For example, you start writing your answer before you even think about it completely, because posting the partial answer and editing it later is better than thinking about it, posting it, and finding that someone else posted their very similar answer 1 minute sooner. So it’s “use google, post the first information found, use google more, edit your answer to include the additional information” cycle. And if the question cannot be solved by googling, nominate it for deleting, as off-topic or something. If no one can answer the question fully, then criticize the partial answers of other people as incomplete and downvote them; if you couldn’t get points by using your strategy, they don’t deserve them either.
Improving measurements is one of the boring but massive levers we have at our disposal, e.g. givewell, the technical details of how voting schemes capture preferences etc.
Stackoverflow seems to succeed in growing despite the issues that Michael brings up. Yes it’s not the place it was three years ago but it’s okay that communities change.
Ways to deal with the fact that fanfiction.net has sabotaged your computer:
Turn off javascript (I’m not going to tell you how to do that)
Select all (and copy and paste elsewhere). It will look like it only selected the header, but it actually works.
Tell google to translate the whole page from from Chinese to English by putting the URL into the left-hand box and choosing the language to translate from (the default is “detect language,” which of course detects English)
Sorry if stupid question. Let’s assume that the universe (mathematical multiverse?) gives us observations sampled from some simplicity-based distribution, like the universal distribution in UDASSA. Can that explain the initial low entropy of our universe (fewer bits to specify), and also the fact that we’re not in a tiny ordered bubble surrounded by chaos?
ETA: I see Rolf Nelson made the same point in 2007. This just makes me more puzzled why Eliezer insists on using causality, given that the causal arrow of time comes from initial low entropy of the universe in the first place, so mathematical simplicity seems to be the more fundamental thing.
A low entropy microstate takes fewer bits to specify once you’re given the macrostate to which it belongs, since low entropy macrostates are instantiated by fewer microstates than high entropy ones. But I don’t see why that should be the relevant way to determine simplicity. The extra bits are just being smuggled into the macrostate description. If you’re trying to simply specify the microstate without any prior information about the macrostate, then it seems to me that any microstate—low or high entropy—should take the same number of bits to specify, no?
If you can encode microstate s in n bits, that implies that you have a prior that assigns P(s)=2^-n. The set of all possible microstates is countably infinite. There is no such thing as a uniform distribution over a countably infinite set. Therefore, even the ignorance prior can’t assign equal length bitstrings to all microstates.
It seems to me that macrostates should take very few bits to specify (e.g. temperature and pressure), compared to microstates within a macrostate (positions of each molecule). So there would still be a difference overall.
Equilibrium macrostates (such as the homogeneous soup at the end of the universe) can be described using very few macroscopic predicates. Non-equilibrium low-entropy states cannot. No need to go back all the way to the beginning of the universe; just look at the present macrostate. No need to look at the entire universe even, just look at the room you’re in. The amount of information you’d need to specify to pick out the current macroscopic state of your room out of all the other macroscopic states of similar entropy it could be in is in fact quite high. A few predicates isn’t going to do it. The lower entropy you get, the more structure there is in the macrostate and the more information you’ll need to specify it.
Now maybe you think that the number of bits required to specify this structure, high though it may be, absolutely pales in comparison to the number of extra bits required to pick out an equilibrium microstate. But think about what you’re doing when you specify a macrostate. You’re basically just picking out a region of phase space. Once you’ve picked out the region, you then pick out a particular point within that region to specify the microstate. In terms of the number of bits required, it makes no difference whether you start our by picking a larger region and then picking a microstate within it, or if you start out by picking a smaller region. Either way, you need to rule out all points in state space but one.
Of course, you might get the illusion of simplicity if you give your small region a special name, like “Bill”. You could say, “All I need to do to pick out this region is say ‘Bill’. It’s so simple.” But then you’ve just hidden the complexity away by choosing a gerrymandered description language. I’ll do you one better by naming the precise microstate of the early universe ‘Bob’, and now I can pick out that exact state just by saying that three-letter word. We have simple names for macro-properties that are salient to us, and that we can easily pick out perceptually, but that doesn’t mean those properties are actually simple from a fundamental physics point of view. The properties of classical thermodynamics correspond to gross constraints that we can control. They are relative to our epistemic and causal capacities. They are not nature’s joints.
Either way, you need to rule out all points in state space but one.
Are you sure that all points in state space require the same number of bits to describe, if descriptions are computer programs? It seems to me that some states are more ordered and can be written out by a shorter program. For example, if all particles have zero velocity, that can be written out by a pretty short program. Equilibrium vs non-equilibrium doesn’t really come into it, e.g. a pair of very ordered objects about to collide at high speed could have a very short description and still lead to a big bang.
I agree that the constants depend on the choice of programming language, but that’s a problem for K-complexity in general. I’d love to know the solution to that...
Can anyone suggest visual symbols for Reason? I need one for a project I’m working on.
Specifically, I’m looking for something that could represent the concept of Reason, but isn’t associated with any modern politics, and doesn’t rely on an understanding of modern science. i.e. the first thing that came to mind was a sketch of an atom, but that won’t suit. It should be recognizeable to a pre-industrial scientist or philosopher.
On googling, the best fit I’ve found so far was a lit candle; Reason as a light in the darkness. That should give an idea of where my head’s at.
I donated a small amount of money to CFAR and Archive.org when there were matching campaigns because both organisations accept bitcoin as a payment option. I liked the feeling but as a student it is quite an expensive experience though I will do it again in the future. One thing I’d like though is to be notified about is when my favorite charities have matching campaigns going on so that my contribution goes a longer way. Is there a way to get emails on those occasions without being spammed every other day about stuff I really do not care about?
It’s one short email a month. While not exactly a notification for a campaign, it should always inform you in time when the matching campaign is running.
A new paper gives a much better algorithm for approximating max flow in undirected graphs. Paper is here. Article for general readers is here. Although the new algorithm is asymptotically better, it remains to be seen if it is substantially better in the practical range. However, this is an example of discovering a substantially more efficient algorithm where one might not have guessed that substantial improvements were possible.
Is anyone out there irritated with the concept of anti-fragility? It seems to me to unnecessarily merge three different concepts: evolution by selection, learning, going out from local optimum by means of random stimulations. The claim that there are systems that benefit from disorder, exactly as stated, to me has only one instance, not even a real thing: Project Orion.
That’s a fun challenge. It was hard to try to summarize the motivation behind the idea of antifragility in such a restricted vocabulary. Here is my attempt:
Things sometimes break. This happens when outside forces cause bad changes to it and to the world it acts in. Things that this can happen to are not things you can put much trust in. It would be a lot better to have something that does not change because of things happening to it, or even better, one that gets better the more those bad things happen to it. It is a good idea to make the things you have be of this sort, but that can not be done all of the time. To find things that match this idea is not easy, but it is true that there are some situations in the real world where this does happen.
I think you want to separate out “bad” from change.Things sometimes break, this happens when outside forces cause changes to it and the world it acts in. We call a thing fragile when most changes are bad from the perspective of the thing. EG ice is fragile because changes in temperature, motion, etc. will cause it to break. Anti-fragility is when the thing is designed such that the biggest possible changes do not break it.
I see what you are saying, but the whole point behind anti-fragility is that change is for good, not bad. By default, in fragile things, change is bad. But in antifragile things, that change is harnessed for good.
Hm. The best way to clearly demarcate that would probably to move the word “bad” from describing the word “change”, and put it as part of the first sentence.
Things sometimes break, and that is a bad thing that you do not want happening. It happens when outside forces cause changes to it and to the world it acts in. …
I’d like to be able to watch posts. Currently, I can see if someone replies to a post of mine, but not any other post. Sometimes there are posts where someone asks an interesting question, but nobody has answered it, and that person isn’t me. There’s no way for me to tell when someone replies to it so I can read the reply.
Every post and every comment has an envelope icon in the lower right, just to the right of the permalink chain icon, just to the right of the speech bubble reply icon. Click on the envelope and it will be highlighted; click again to go back to gray. While highlighted, replies will show up in your inbox. It is only for immediate replies, not replies to replies.
I have recently seen this article on a drug, which is as yet not approved by FDA due to lack of controlled study. However, the proponent doctor considers the drug a life saver for babies and the existing evidence sufficient to skip the trial phase and go ahead saving lives.
I am usually fond of evidence based medicine. However, in this case, I am shaken by the fact, that the drug is already in use in Europe for some time, with no scandals. Additionaly, it is basically a mixture of lipids from fish oil, which sounds like normal nutrition to me, not a novel chemical with unpredictable effects. (Although the application is intravenous, which is not “natural”).
I believe no finding will be found, which will, in retrospect, justify the current state of denying access to it to fataly ill children patients. No side effects of the kind of “Oh, I wish I let my baby die back then !” (Disability ? Chronic pain ? Mental retardation? - I do not expect any of that… After not seeing it in Europe so far.)
I want to transform this belief into a bet in prediction book. Can anybody give me advice, how to formulate the bet ? What should be a deadline for the bet evaluation ? It will be my first bet in prediction book...
Theoretically, that could make a difference.
In my country omegaven is available, but the pack insert gives warning about insufficient experience in children.
Anyway, I still want to bet. What is a fair time span ?
(BTW, do I understand it correctly, that in the USA, omegaven is approved neither for infants nor for adults ?)
In a recent comment, II expected that my question might have been already answered so I wrote this:
I’m just now seeing this discussion, and don’t have time to read earlier posts.
I knew this was arrogant, so I appreciated the humor of this reply:
Maybe you can hire someone to read them for you and prepare an executive summary :)
I wanted to explain here why I did not read the previous posts.
There are roughly three nested reasons:
first, it was an experiment because I am often tempted to write something like this (in fact, I have in less egregious cases) and I was wondering what kind of norm-violation response I would get. I was “testing”, because this seems like a nice thing to get away with. The fact is I don’t have time, but if i want to participate in a discussion sometimes I take a gamble and ask a question anyway. I don’t provide a disclaimer that I haven’t read all (or any) of the background, but it would be nice to be able to.
it was symmetric that it was suggested that I ‘hire’ someone to write a summary, because when I am more conscientious and research something just-to-be-conscientious, I often self-criticize that Less Wrong is not “paying me by the hour”. (That is, it’s not supposed to be work.)
finally, it’s not just me that needs to do all this ‘work’. How many people have read all the previous posts? I think it would be nice if executive summaries were provided more often (I provided one here at the top, as well as links) but if that’s too much work (sure it can be) then it could be more of a norm that a person can just plead ignorance / request a relevant recap.
(Ironically, has an open thread like this already been posted?)
I know there are some R Scott Bakker fans on here, and I was thinking recently about the Second Darkness series. Rot13d for spoilers:
Vg’f n funzr gur pbafhyg ner rivy. Vs gurl jrera’g fb pbzzvggrq gb rivy npgf, gurl pbhyq ratvarre n jnl gb tvir rirelbar n unccl raqvat.
Jr ner gbyq gung fbepreref ner qnzarq, naq gung gur hygvzngr tbny bs gur Vapubebv vf gb erqhpr gur ahzore bs yvivat fbhyf ba gur cynarg gb srjre guna 144,00 va beqre gb frny gur cynarg sebz gur Bhgfvqr naq rfpncr qnzangvba. Gur Pbafhyg pbhyq erpehvg nf znal fbepreref nf cbffvoyr, genva gurz gb terng cbjre naq gura rkgraq gurve yvsrfcnaf jvgu gurve Grxar, juvyr hfvat gurve vasyhrapr gb zbir gur phygherf bs gur jbeyq gbjneqf n zber uhznar naq pvivyvmrq zbenyvgl (be, tvira gung gurer ner nccneragyl jnlf gb trg qngn ba ubj gb orpbzr fnirq be qnzarq, ng yrnfg gbjneqf n fbpvrgl gung vf uvtuyl rssrpgvir ng qvfpbhentvat qnzavat npgf). Bapr n fhssvpvrag ahzore bs fbepreref (fbzrjurer whfg orybj gur 144,000 svther) unir orra npphzhyngrq naq gur jbeyq pbafvfgf cevznevyl bs aba-qnzarq vaqvivqhnyf, gur Pbafhyg pbhyq hfr gurve uhtr zntvpny cbjre gb dhvpxyl trabpvqr gur ragver cynarg, oneevat gurve bja zrzoref. Zbfg fbhyf jvyy or fnirq, naq gur qnzarq fbepreref jvyy erznva ba n frnyrq cynarg, univat nibvqrq qnzangvba.
Guvf ng yrnfg nccrnef gb or n jnl gb trg gur znkvzhz ahzore bs fbhyf n unccl raqvat—juvyr jr qba’g xabj jung unccraf ba gur cynarg bapr vg vf frnyrq sebz gur Bhgfvqr, gur Vapubebv frrz vagrag ba npuvrivat fb gurl zhfg oryvrir vg’f qrfvenoyr.
People who post probability estimates of anything should explain in details how they arrived at them. Otherwise it should be called not probability estimate but pulling it out of your ass.
Seriously, stuff like “1% FAI success by 2100”? When there is no clear definition of AI in sight? Just stop.
All beliefs are probability estimates, although it can be hard to trace how a particular belief got to the degree of confidence it’s at, and while it might be a nice norm to have in a perfect world I think it’s unreasonable to demand that every time someone expresses how confident or unconfident they are in a belief, they should also clarify the entire precise history of that belief’s presence in their mind.
Apologies for the curmudgeonliness, but it really bugs me when people say things like this. The actual version of this statement that is true is
All coherent actions can be modeled as arising from beliefs that correspond to probability estimates
which is different and much weaker, as now we can argue about how important coherence is relative to other desiderata. One such desideratum is correspondence to reality, which I believe is Locaha’s point above. Personally, I would much rather have incoherent beliefs that correspond to reality than coherent beliefs that do not correspond to reality.
Whenever I post a probability estimate, it is solely for the purpose of making my position more clear, not as something anyone should use to actually update their beliefs. You should always consider probability estimates as rough information about how the person who made the estimate thinks, not as a factual bit of information about the prediction itself.
I don’t think you make anything clearer by translating your intuition’s unlikely to N%, while my unlikely is M%, where M!=N. You just make a false impression of having done a calculation (which, unlike intuition, can be confirmed).
Suppose passive_fist translates “unlikely” as 2% and Locaha translates “unlikely” as 12%. This could mean either of two things (or some combination of them). (1) passive_fist applies the word “unlikely” to things that feel more unlikely, corresponding to lower probability estimates when forced to quantify. (2) Both actually think much the same about the event in question, as shown by their use of the same word, but they have quite different processes (at least one of them very inaccurate) for translating those thoughts into numbers.
In case 1, quantifying helps to clarify that the two people involved mean quite different things by “unlikely”. There may be a lot of fuzziness about the numbers, but once we have them we can see that passive_fist will likely be much more surprised if something s/he calls “unlikely” happens, than Locaha will be if something s/he calls “unlikely” happens.
In case 2, quantifying just adds confusion and error.
I would expect that (especially for analytical quantitative types like most of LW’s readership) the truth is something like this. We think, mostly, in fuzzy terms that don’t correspond directly either to numbers or to words. There will be some region of subjective likelihood-feeling space that corresponds (e.g.) to the number 2% or 12%. There will be some region that corresponds (e.g.) to the word “unlikely”. These correspondences will all work differently for different people, but (a) there will generally be more consistency between one person’s “10%” and another’s than between one person’s “unlikely” and another’s, and (b) the finer-grained information you get by asking for probability estimates does have some value, provided you’ve wit enough not to imagine that everything expressed numerically is known accurately.
People who post probability estimates of anything should explain in details how they arrived at them. Otherwise it should be called not probability estimate but pulling it out of your ass.
Probabilities are useful for being precise about the claims that you are making. There no reason why one shouldn’t be precise about the claim one is making even when one doesn’t use a formal method to arrive at them.
If you don’t use a precise method to arrive at your claim, you have no business making a precise claim. Remember significant figures from high school chemistry? Same principle.
I think this is an error. (And so are “significant figures” as commonly used.) 2.4 +- 2 and 2.0 +- 2 are quite different estimates even though you wouldn’t (according to conventional wisdom) be justified in giving more than one “significant figure” for either.
Using the number of digits you quote to indicate how accurately you think you know the figure as well as to say what the number is is a hack. It’s a convenient hack sometimes, but that’s all it is. Everyone knows not to round intermediate results even when starting with low-precision numbers. Well, your final result might be used by someone else as an intermediate result in some bigger calculation.
The same goes for probabilities. It is very important to know when your estimate of a probability is very inaccurate—but that’s no reason to refuse to estimate an actual probability. Even if you just pulled it out of your arse: doing that makes it a very unreliable probability estimate but it’s still a probability estimate.
I won’t deny that significant figures are a crap implementation of the principle I’m talking about. But you have to propagate the uncertainty and include it, in some way, in your final answer, either numerically or via some explanation that might let me figure out how precise your answer is.
Don’t say “1% probability of FAI success by 2100.” Say ”.01-10% probability of FAI success by 100 based on XYZ.” Or if there’s no numerical process behind it that can support even a range like that, just say “FAI success by 2100 seems unlikely.”
Agreed. Though in the latter case you might still do best to give numbers: “Somewhere around 1%, but this is a wild guess so don’t take it too seriously.” This is not the same statement as the corresponding one with 2% instead of 1%, even though both might be reasonably accurately paraphrased as “unlikely” or even “very unlikely”.
If you don’t use a precise method to arrive at your claim, you have no business making a precise claim. Remember significant figures from high school chemistry? Same principle.
That assumes that someone isn’t calibrated. If someone calibrates his intuition via frequent usage of prediction book and by always thinking in terms of probability he might be able to make precise claims without following a precise method.
If someone would claim “1.21% chance of FAI” success by 2100 I would agree with you that the person didn’t learn the lesson about significant figures from high school chemistry. I don’t the that issue with someone claiming 1% chance.
If you want to get calibrated it’s also useful to start putting numbers on a lot of likelihoods that you think about, even if the precision is sometimes to high. It allows you to be wrong and that’s good for learning.
I think it’s likely that calibration is domain-specific, so I’m not sure I buy this unless the calibration has occurred in the same domain, which is rare/impossible for the domains we’re talking about.
I think it’s likely that calibration is domain-specific, so I’m not sure I buy this unless the calibration has occurred in the same domain, which is rare/impossible for the domains we’re talking about.
I think you can argue that the probability is inherently unknowable but I don’t see how a detailed process is much better than an intuitive process.
It’s very useful to have a mental ability to distinguish between 0.01, 0.001 and 0.0001 when it comes to thinking about XRisk events. I don’t think that it’s a good practice to call all of those events unlikely and avoiding to make semantic distinctions between them.
It’s very useful to have a mental ability to distinguish between 0.01, 0.001 and 0.0001 when it comes to thinking about XRisk events. I don’t think that it’s a good practice to call all of those events unlikely and avoiding to make semantic distinctions between them.
But how do you arrive at them? Intuition doesn’t deal with 0.01 and 0.00001. Intuition deals with vague notions of likely and unlikely, which also change depends on what you ate for lunch and the phase of the moon. IOW, your intuition is useless to me unless I can confirm it myself. (But then it’s not intuition anymore.)
But how do you arrive at them? Intuition doesn’t deal with 0.01 and 0.00001. Intuition deals with vague notions of likely and unlikely, which also change depends on what you ate for lunch and the phase of the moon.
I think there are plenty of cases where I can give you a intuitive answer that won’t change from 0.01 to 0.00001 depending on what I ate for lunch.
The chance that I die in the next year is higher than 0.00001 but lower than 0.01.
If you don’t have an intuition that allows you to do so, I think it’s because you don’t have enough exposure to people making distinctions between 0.01 and 0.00001.
If there’s a 0.01 chance that something happens tomorrow, then if everything stays the same you’d expect that thing to happen about three or four times this year, whereas if it’s 0.00001 you’d be quite surprised if it ever happens (EDIT: during your lifetime, assuming no cryonics/antiagathics/uploads). (Of course with stuff like x-risk intuition will be much less reliable.)
And that’s a good assumption, since by my estimate, 99.9537% of people are not calibrated.
Part of lesswrong mission is about moving to a world where more people are calibrated. I don’t think it’s helpful to declare calibration a lost course.
But the “50% probability of Situation A (2% probability of FAI in 100 years) and 50% probability of Situation B (0% probability of FAI in 100 years)” is much more informative to the reader than “1% probability of FAI in 100 years.” It exposes more about which parts of the estimate are pulled out of the writer’s ass. If I know something the writer doesn’t about any one of these component probabilities, I can update my own beliefs, or discuss the estimate, more usefully this way than if I’m just given a flat “1%.”
Probabilities are useful for being precise about the claims that you are making. There no reason why one shouldn’t be precise about the claim one is making even when one doesn’t use a formal method to arrive at them.
How is the belief of some random person X in some vague-defined event many years into the future is useful for anything nut the research into person X state of mind? Even if it’s defined to 1000 significant figures?
If you are reading the text of a person who presumably care about that person state of mind and what this person believes.
If you don’t why do you read the text in the first place?
I do think there a difference between someone thinking an event is unlikely with p=0.2, p=0.01 or p=0.0001. It worthwhile to put a number on the belief to communicate the likelihood.
If people frequently provide likelihoods you can also aggregate the data.
I suggest having a link to the last open thread in each open thread, and similar for the quotes thread. That way, you can just follow the link to find out what people posted near the end of the last thread, so it doesn’t become pointless if there’s going to be a new thread soon.
Besides LW, what are some other online communities with very high signal/noise ratios?
Some good indicators: Lots of original content, meaningful and well-presented information, respectful conversations with a high level of discourse, low levels of trolling/strong community norms, very strong domain-specific knowledge, people know each other by username.
Another heuristic: somewhere where you would not want to share the link with a lot of people lest the quality be diluted with newcomers. (Hopefully you consider LW a strong enough pool to draw from).
Examples I can think of:
The Straight Dope Message Boards is a great general topic forum with low rates of trolling.
The Oil Drum, is a strong web community specific to the Energy industry.Stopped updated this past September (thanks knb, I hadn’t visited for a couple months).Niche subreddits are often a great resource, so much so that when I’m looking for information I often do a reddit search before looking over the greater internet.
Things that wouldn’t count:
Social news sites, like hacker news or the the large subreddits, which tend to have a lot of noise.
Gwern’s Google+ feed has perhaps the single highest signal/noise ratio of anywhere I’ve seen on the web, but obviously is just the one person.
The Erowid Experience Vaults is a solid collection of crowdourced drug experiences, but doesn’t have a community attached.
Any others?
Are you after interaction or information content?
As far as raw content goes: I strongly suggest a “Unified standard for print and Internet” rule. Judge content against whatever book or paper you could be reading instead. Often communities have separate incentives to promote and praise original content regardless of absolute quality. There’s also a novelty bias, where mediocre content in new mediums (youtube videos, blogs) is held to a lower standard. Since I’ve implement this rule, I’ve noticed a lot of what I considered great content is really just “great for the internet” content. I read more books now.
Against this standard, I find the most raw information value comes from individuals who explicitly collate links or summarize information from other sources. Democratic collation can be great as well, but faces too many ways to die. Long-form community and blog content is almost always overrated, except for cases like niche ideologies or at the cutting edge of some technical or business topic.
Meta Warning: I think this applies to lesswrong. There are exceptions, but the long-form posts and related blogs are often either “good for the Internet” content or decision theory stuff I don’t have the background to understand. I don’t mean to gripe: I’m very happy with LW, and I regularly skim the discussion section and read the open threads. I just think that LW over-hypes it’s own original content, like any other community.
Would you care to name some of them?
The Oil Drum isn’t updated anymore. It also never struck me as particularly good.
Let’s say I wanted to monitor the variation in my Big 5 personality traits over time. Is there an existing way to do this, or should I handroll my own procedure, which will probably be total BS?
What’s scientifically known about hangovers
Debunks the common notion that hangovers are about dehydration. The reason it caught my eye is that I believed the dehydration theory, even though I should have known that extreme sensitivity to sound isn’t a normal symptom of dehydration. (I’ve never had a hangover, but at popular accounts include sensitivity to sound and light.)
I’m wondering how I can become skeptical enough.
One reason for the myth about dehydration would be due to “drinking plenty of water” still being one of the most effective things to do: If it’s about the liver breaking down alcohol into toxic Acetaldehyde, drinking lots of water to flush it out.
Understandable mistake to go from “more water fixes the problem” to “problem must’ve been not enough water (dehydration.)”
This was how it was discussed in my university chemistry class. Also mentioned: a similar breakdown (same enzymes or whatnot) happens with methanol, and the breakdown products (formaldehyde and then methanoic acid) are stronger / more toxic than those of ethanol (acetaldehyde / acetic acid.)
How? I’m pretty sure that contrary to popular belief, water can’t simply be used to flush stuff out, the metabolite also has to be in a form that can be excreted by the kidneys, and even then extra water might have no effect whatsoever. This is basic physiology found in any textbook on the subject.
I tried to google if kidneys might excrete some of the acetaldehyde, but found no answer.
I had reservations about including that sentence, because I only have a vague idea which completely lacks details about mechanisms. And flushing seems like a folk-explanation rather than a science-explanation.
The other vague idea was that drinking more water means the toxins are more dilute, but I have even less confidence in that.
More dilute compared to the (cellular) mass of a person? That’s a rather lot of water.
If you check out the linked study, it doesn’t do such debunking. It found no association between hangover severity and vasopressin and it wasn’t clear what other markers of dehydration it measured. Note that measuring dehydration accurately could be difficult and different people might experience different levels of dehydration differently.
It should be the default assumption that there are many mechanisms involved in hangover, because the effects of ethanol are very complex.
Interesting—I’ve had a plenty of nights of drink, including some where I felt unwell the next day, but never had either of those symptoms. I have, however, woken up recognizably dehydrated a couple of times.
I wonder whether some of the “myth” comes from the experiences of people like me, accurately reported but for some reason not defined as hangovers by the researchers behind the study cited.
I’ve mistaken caffeine withdrawal for a hangover before, partly because of similarity in symptoms (headache; nausea; photosensitivity) and partly because it tends to show up around the same time (Sunday morning). This may account for the popularity of coffee as an alleged hangover cure.
Which raises the question, if the things people say about “hangovers” are true about the things they apply the term “hangover” to, what’s left to be debunked?
The belief that the things they’re talking about are caused narrowly by overuse of alcohol?
But isn’t the claim being “debunked” that hangovers are mainly dehydration, not the direct effects of alcohol?
That article rubs me the wrong way. I think it may be more a failure of science, rather than the author’s personal failure though. Whenever people are curious about how to reduce hangovers, all the articles you find will talk about how abstinence is the only cure (gee, this doesn’t echo any other memes), but the fact is that there is an effective treatment, even if there haven’t been sufficient scientific studies done on it. A large social group that I am involved in, which has been known to drink heavily, has started taking N-acetyl-cysteine (500-1000mg) and Source Naturals Hangover Formula (which is primarily a C & B complex), and the effect it has on hangovers is not in any way subtle. Despite the existence of this, the article only says that the best prevention is to consume alcohol with food and lots of water. There is some scientific support for this combination (see the studies and explanation referenced here or here), but even though it is a question of significant practical importance to many people, no scientists have actually gone out and done a controlled study on humans on these nutrients.
I think “only known cure” and related phrases like “no known cure” should be added to the list of semantic stopsigns and lullaby words.
I have been interested in the phenomenon called tulpa. (interestingly, Wikipedia sheds next to no light on this issue).
According to one site, it is an “autosuggested and stable visualization, capable of independent thought and action, while possessing its own unique consciousness”. Supposedly, following the guides found on the internet, one can create a stable, persistent “imaginary friend”, with the looks and character one wants that will be real in all aspects for its creator. Some say that tulpa can provide an alternate viewpoint or help fetch information from their host’s memory, but various hosts disagree on the possibility of this.
Looks like tulpa in modern, Western definition has no connection to its Buddhist namesake (like karma on the forums). Some enthusiasts claim otherwise, but, as seems to be characteristic of this topic, there’s no evidence.
All I could find are guides and diaries of anonymous people on the Internet. It seems like the whole phenomenon, if it really exists, was invented some 1.5 years ago by some Anonymous: there’s their own slang, and absolutely no sources that connect the methods to any actual scientific research.
I suspect that the tulpa phenomenon may either:
not work (all cases of success might be explained by belief in belief; people might be adeptly fooling themselves and their peers, thinking up their imaginary friends and persuading everyone that their actions are autonomous)
be dangerous to one’s mental health (shouldn’t success here by itself be diagnosed as some kind of personality disorder?)
which is by itself a reason not to try it myself, but I have a nagging doubt: which one of the two is true, if any?
I’ve recently talked online, including by voice chat, with people who claim to have tulpas, and quickly ran out of useful questions to ask. I asked for interesting questions to a member of LW meetup group and he said
Really, what else can one do? The people that I talked to seemed surprisingly unconcerned with using methods written by Anonymouses, lack of scientific evidence, status of the phenomena with regard to psychiatry or what comes for “evidence”, “proof” or “source”.
Can anyone help me to find answers for the following questions:
What interesting questions can I ask people who claim to have tulpas?
Are methods grounded in science in any way?
What is the position of science with regard to tulpas? Surely, if the phenomena exists, science must have encountered it and have a stance wrt. it, right?
It could be a good occasion for me to learn scientific research; any help with it would be appreciated: where should I start figuring out what, if any, does science know about this?
There should really be full discussion post about this, since it keeps getting brought up.
EDIT: So I made one. If there isn’t interest, well, at least it’s spurred me to consolidate a bibliography
In addition to what I wrote in the other thread:
Luhrmann wrote a book, When God Talks Back, about her experiences with evangelicals, which might be useful. She also succeeded in inducing tulpa-like visions of Leland Stanford, jr. in experimental subjects.
The tulpa community also seems to have a fondness for amateur psychological research, although I imagine there’ll be a lot of chaff and unfinished projects in there.
I don’t know what questions you’ve already asked, but how about
What has your tulpa done that’s surprised you? What does your tulpa do that you can’t do when you’re in your default state?
You may also be interested in the earlier LW discussion about tulpas. Like I mentioned in the thread, they seem like an intentionally developed version of a thing that many writers have naturally—my guess would be that they use the normal circuitry that we have for emulating and predicting the behavior of other people, only they’re modelling a non-existent person and the outputs of that modeling get fed back in to be used as new input.
Oops, I didn’t see that. Thanks.
Edit: Still, at a glance, the 3 questions I’m asking this time, were not exactly asked in the linked discussion. So I welcome everyone to share their thoughts here.
There no such thing as a position of science. Science is a process.
As far as Tuplas go, Tulpas are mainly about qualia. Qualia are by their nature but directly measureable. As a result a lot of scientists do shun research of qualia. Orthodox reductionists do try to explain qualia away whenever they can instead of trying to investigate them.
Western scientific education doesn’t train it’s scientists do be in control of their minds and that means that most they can’t run experiments that require mental control and good awareness of their own minds.
They can invite some buddhist monk and investigate how that monk does his thing, but that doesn’t allow for easy controlled experiments.
In that area even the easy questions seem to have little scientific investigation. Take heat development during particular types of meditation.
Ten years ago a scientists would have looked strangly at you for suggesting for heat development without movement but now we do know that the body can in principle do this in brown fat tissue.
There some small experiments that illustrate that a specific Tibetian technique can consistently produce heat but there no general science based theory that predicts when you would expect someone who meditate develops heat. It’s even not clear whether it’s really produced in brown fat tissue because adults have little of it.
Present thinking in medicine is rather: “How can we develop a drug that stimulates brown fat tissue to be active to help people lose weight?”
Tulpa need around 6 months of hard focused mental practice of 1 hours per day. That not something that you usually study in scientific studies.
If you look at another recent controversy on Lesswrong we even disagree whether mainstream science knows what losing weight is about. There a lot more scientific effort going into that question but it still isn’t conclusively answered.
If you are looking at far out mental phenomena there no reason to expect that they are well investigated by scientists.
Aren’t hallucinations about qualia? At a glance, science seems pretty well informed about hallucinations. What’s the important difference?
What makes tulpas significantly more far out than hallucinations?
Finding a person who hallucinates is pretty easy. You go to your nearest asylum and go through the patients and you will usually find someone who has hallucinations.
As luck would have it the patients are also bound for years to a specific location and might have no possibilty to opt out of your study.
Finding people as test subjects who spent halve a year doing a hard mental practice is harder. Experiments that require that you have test subjects who spend a lot of time on a hard mental practice are much easier to do.
The page you linked to doesn’t provide evidence that indicates that science is well informed about the issue. It doesn’t illustrate that scientific theories are able to make reliable predictions about hallucinations. One of the examples about which the wikipedia article talks is a unreplicated 13 person experiment with 5 days duration. It talks about is as “strong support” for an idea.
It says “There are few treatments for many types of hallucinations.” You can translate that into the acknowledgement that the phenomena isn’t well enough understood to effectively modify it in the way you want.
The third way to check whether someone understands something is to check with your own empirical experience. I unfortunately don’t have much experience with hallucinations that go beyond things like the optional illusion where every normal viewer hallucinates that wheels turn.
I do have some experiences I had after spending 5 days in an artificial coma. One of them is a state where what I see visually doesn’t change when I close my eyes. I know of descriptions of other people who experienced the same thing. Can you find a mainstream science description of that visual hallucination?
Someone who works at google told me the company is working on trolley problems because self-driving cars may have to make that sort of decision, and google will be responsible.
...I’d love to know what that means, and who in google makes the decision of what it means.
Google will at least be at risk of being held legally responsible.
https://twitter.com/slatestarcodex/status/410199029243342848
Dumb reinforcement question: How do I reward the successful partial-completion of an open ended task without reinforcing myself for quitting?
Basically I’m picking up the practice of using chocolates as reinforcement. I reward myself when I start and when I finish. This normally works very well. Start doing dishes → chocolate → do dishes → finish doing dishes → chocolate. It seems viable for anything with discrete end states.
Problem—I’ve got a couple long term tasks (fiction writing and computer program I’m making) that don’t have markers, and I can put anywhere from 30 minutes to 3 days into them without necessarily seeing a stopping point. I’m worried that rewarding chocolates whenever I get up from working will (in the long run) reinforce me to quit more frequently. I don’t want to end up with a hummingbird work ethic for these tasks.
How should I reinforce to maximize my time-on-task?
(So far my best plan is to write a smartphone app that creates a hidden random timer between 5-55 minutes (bell curve) that goes off, and I reward myself chocolate if I’m on task when the alarm activates. But there’s logistical hurtles and it seems like quite a bit of work for something that might be solved easily otherwise. Plus, I don’t know what possible bad behavior that might incitivize.)
Why does it need to be a hidden random timer? Reward yourself if you stayed on task for the past 30 minutes. (Hmm, I think we’ve just reinvented the Pomodoro Technique.)
Incidentally, have you (or others who use schemes like this) considered using intermittent reinforcement? Like, instead of just rewarding yourself upon meeting the victory condition, you flip a coin to see if you get the reward. It seems the obvious thing to do if you’re going for the whole inner pigeon thing.
Hmm, reward myself after a fixed interval of 30 minutes? That’s just crazy enough to work! (I have heard of the Pomodoro technique before, and I’m not quite sure why I didn’t just go for that at the start.)
The hidden random timer is to make myself resilient to extinction and ingrain the habit even without reward. Although, randomly choosing to reward at the end of pomodoros would work too. IIRC, intermittent time interval is the reward structure that survives the longest without extinction, whereas a variable ratio reward structure creates the most vigorous workers.
Also, I think what you describe is a conditional reinforcer and not an intermittent one. What I mean by that is after a long enough time, they subject would become attached to the coin flip itself as a partial reward. Kind of like clicker training for animals, or how a shot at a jackput pull is a reward even when it doesn’t payout. Then you could use the stronger conditional training systems...
Your suggestion is brilliant. Aaaand now I’ve got “write a gamblerdoro app” on my to-do list.
You might like http://HabitRPG.com—there are already a few of us from LW on there. If you join it, tell us your user ID so we can invite you to our party :)
I just joined via facebook as misha gurevich
Question about a low-level social thing:
I’ve noticed that I have low priority in at mid-large group conversations. What I mean is that in situations where I’m one of two people talking, I’m (generally) the one who stops and the attention of the “audience” (people-who-aren’t-speaking) is predominantly on the other person even before I stop speaking.
This used to cause me considerable distress, but no longer. I’ve accepted it as a fact of the social universe. But I’m still curious and would like to change it, if possible.
I suspect that this is something that varies by social group, and more strongly suspect that some behavior of mine is key.
I’m interested in (being pointed to) discussion of this type of thing, especially if it contains actionable advice.
I can’t really offer anything more than a personal anecdotes, but here is what I usually do for when I try to grab attention of a group of my peers:
If you are talking to several people gathered in circle, and it is my turn to say something important, I make a small step forward so that I physically place myself in the center of the group.
When I am speaking, I try to mantain eye contact with all people gathered around, If I focus too much only on the person I am speaking to, everyone else turns their attention towards them as well.
I rarely do it myself, as I suppose it is a technique more tailored for public speeches, but conservative use of hand gestures to signify what you are talking about, probably won’t hurt.
I probably sound like a self absorbed jerk writing this, but if I want the attention to focus on myself, and not my interlocutor I often use “me” language. Compare and contrast [“What you say about vegans is true, but you may conisder...”—now everybody looks at the person who said something about vegans] [“I think that I agree with what was said about vegans, but I also think...”—now everybody looks at me as I explain my position].
But those are all just little little tricks, when the surest way of attracting attention of the audience is simply to have something important and interesting to say.
While we are at that topic many people use “you” when talking about themselves. They say sentence like: “Yesterday I thought: You should go to gym.”
I once even listened to someone who used “he” to when speaking about himself a few years ago. The language was German and he was an Austrian, but it still signified how little he identified with his self in the past.
After a bit of prodding he changed to “I”. That also changed subtle things about his body language did change. It was interesting to watch the effect. Identifying with oneself helps to be more charismatic.
It’s one of those nontrivial aspects of: “Just be yourself.”
Thanks. These are things I’ve learnt or tried learning in the past. I’d guess there are good odds that I’m reverting to past (shyer) behaviors in some situations.
I’ll make an effort to be aware of my body language and focus next time.
You’re really not giving enough specific information. There are countless reasons why that might happen so any advice you take here could lead to erratic behavior.
In addition to what other people said, this sounds like you might be too verbose or bad at gauging which topics interest other people and to what extent. They might look at the other person because they wish them to interrupt you and move on.
Fair enough. At this stage I’m curious as to which specifics I should be looking at. Or what kinds of things are key (to speaker priority in groups of 5-10).
The various elements of body language given, and your notes on content (I can be too verbose, for sure) have given me what I need to go on for now.
If you start talking loudly, perhaps a little louder than you think you should be talking, attention will usually shift to you. If you say something relatively short and relatively interesting, you gain some credit, and then later people will be more likely to listen to you. This requires some timing—you have to start talking just as the previous speaker is finishing, or even a little before. Mistiming could make you look like (and be) a shouting idiot.
If two or more people use the same strategy (starting off loud), then you can use the opportunity to appear (and be) gracious by telling the other person to go ahead. This also subtly puts you in a position of control; they got to talk, but you decided who should talk.
(This is all based on personal experience, of course; YMMV.)
There are a variety of plausible explanation.
The first would be that it’s just an issue of perception. There a plenty of people with low self esteem who hold inaccurate beliefs about how much attention other people pay to them.
A second would be generally proxies for low testosterone. If you have a louder and deeper voice people are more likely to listen to you. There are also body language changes and various other things that are hard to fake.
Thanks. Perception I thought was the main contribution in the past. But after a recent party my partner commented to me about people speaking over me.
And testosterone, that makes me curious. I wonder if I can get levels of that tested without too much hassle...
The interesting thing about this sentence is that you communicate a minimum of information by using the word partner.
As a listener I don’t know whether you refer to a business partner, you are male and refer to a girlfriend or you are female and refer to a boyfriend.
Increasing the amount of details that you communicate can increases the amount of attention that other people pay towards yourself.
When someone speaks in a way that strikes you as unusually ambiguous, consider that it may be intentionally so to avoid placing emphasis on irrelevancies.
What makes you think I didn’t? If he’s in the habit of intentionally sabotaging himself that increases the value of having it being pointed out to him.
I wouldn’t highlite it if he wouldn’t have specifically asked for ways to draw more attention on himself. The more of your attention you concentrate in withholding information the less likely people are going to give you attention.
It’s not like it’s wrong to do so, but you should be aware of the tradeoffs that you are making.
It’s not obvious that attention is being diverted to obscuring information. Maybe “partner” is the lower-effort term. It is for me: “boyfriend” and “girlfriend” have weird connotations and baggage that make me want to pause before using them and wonder if I’m really being accurate.
It makes it harder for me to build a mental image of the situation that he’s describing. If it’s harder than I will put less attention on the situation.
Here it’s a subtle choice. But it points to a pattern. A common general piece of advice on telling stories which draw listeners attention is to provide a lot of adjectives to make it more easy for the audience to picture what you are describing.
The kind that most people don’t put to practice when they hear it. Instead of describing the principle abstractly, I pointed to an example.
It takes effort to provide listeners with details. It’s still one of the pure white hat strategies to getting peoples attention when you speak.
As far as the information being relevant, when I give someone recommendation about testosterone, the gender of the person I’m interacting with matters.
I have a good idea that more testosterone in males will help with given attention. I’m less certain, that trying to increase testosterone is a good strategy for females.
You omit at least two possibilities: that he is male and referring to his boyfriend or that she is female and referring to her girlfriend. In these cases, word “boy/girlfriend” would have you interpreting the situation wrongly.
As others have commented, the fact that we do not know these unnecessary details is a feature, not a bug, of ungendered words.
In this case, it’s deliberately non-gendered language. Lower-effort, as kalium says. In my case because I cultivated the habit, in years past.
As both you and Douglas_Knight point out, there are tradeoffs involved. In the case of not gendering pronouns I expect I’ll continue thinking it worthwhile.
But it’s a helpful thing to consider- I’ll bet there are other habits I’ve developed that I’ve never considered if it’s worth the costs. Especially when I contrast my teenaged self – “I don’t care what anyone thinks of me” + “I’ll choose my words for my own aesthetical pleasure” – with the me of today – who does care, and on balance values communication higher than self-expression. I doubt my conversational habits have shifted as far as my preferences have.
I’m also interested by what you say about details. It’s not something I’d’ve thought of, worrying I tend more to being too verbose. But I like to write, and concrete detail / description is the area I’d consider my weakest there. I can think of some ways to practice this (started playing tabletop RPGs recently, for one.)
When it comes to details, the thing that count is whether the details help the person you are talking with to form a picture of the situation that you are describing in their mind.
If you are talking in “I” using colorful language that describes the qualia you perceive is nearly always good. That doesn’t mean that sentences should be long. If you can transform one sentence into two, that’s often good.
If you are talking in “you” it better to be a bit more vague. If you tell someone: “When you go to work on Monday morning at 7 o’clock and drive though rush hour, you know the feeling where you wish, you could just take a day off?”, the might be irritated if they aren’t in the habit of going to work at 7 o’clock or don’t drive.
I”m okay with non-gendered language. In text it’s nearly impossible to see what the words mean for you.
If your teenage self trained a bunch of separate word choices I would look carefully at them and judge whether they are real self-expression or whether they are more of a mask that’s supposed to provide shelter.
Authentic self-expression draws attention, wearing a mask reduces it.
May I suggest that you not consider, but just experiment?
Usually when people intentionally manipulate emphasis like this, they are trying to change relevance, not merely acknowledging existing relevance. Maybe it is worth the personal cost to achieve the political end, but it is not possible to make the tradeoff without being honest about the cost.
I think body language tends to be pretty relevant here; it’s possible you are not noticing the other person’s body language indicating that they’re about to start speaking. (And of course, if you want to have it be your turn to speak, using body language to signal this is also relevant.)
Cosmologist and science popularizer Sean Carroll debates Christian apologist William Lane Craig on Feb 28. The topic is God and Cosmology. My prediction: while Sean Carroll is very good, I don’t expect him to beat a professional (and a very successful) debater.
The word we’d need to better define is “beat”.
Will WLC appear to win the debate? Likely. Will he appear to overwhelm his opponent? Likely. Will he succeed at framing the debate in a way that suits his position? Overwhelmingly likely.
The debate takes place in front of his home crowd (an Evangelical theological seminary). WLC will gallop his way to “victory” again, methinks.
If you predict that he’ll win under every definition of “beat” that you suggest, why do you need your initial disclaimer that the definition matters?
Because he won’t actually win anything, anymore than a rabbit actually disappears inside a top hat or a woman actually gets sawed in half and restored.
Debates are of limited value in this way: The correct argument can rather easily lose. WLC is a great example of why and how.
Other than status.
If he follows Chris Hallquist’s advice of insisting on speaking first and gives a talk as good as this one with edits to suit Lane Craig, I predict it would go down pretty well.
However, I assign a low probability to this happening.
I need some advice on spaced repetition software.
I teach high school English to underclassmen who skew towards “totally unmotivated”. I have been using spaced repetition principles for years (using games, puzzles, and other spaced reviews) to help with vocabulary and terminology. These do effectively engage many of the poorly motivated.
But recently, I feel like smartphones have become ubiquitous enough among students that I’m looking for software I could use as a quasi-official SRS companion app with my students. I think many of them would use it, but only if they experience very minimal frustration setting it up and running it. My wishlist:
(1) Free app on both Android and iPhone (I’d say it’s about 50⁄50 with my students) (2) Companion web app with cloud sync to mobile apps. (3) Very easy to use and update with new cards regularly. I would like to be able to post weekly deck additions on my teacher web page that students can add to their deck.
Anki, which I use for my personal learning, seems to come closest—but the $25 cost of the iPhone app is a problem, and I worry that using the web app on the iPhone be too much of a hassle. I also worry that the “add external cards to your deck” procedure is a bit too hairy as well.
Has anyone seen anything that comes closer to my needs than Anki? Thanks!
Look into memrise.
It has an app, it has a lot of the bells and whistles that Anki lacks (like a scoring/gamification system) that could be helpful with the population you are teaching, and it is all around a solid SRS system. The only thing I think it lacks are those Easy/Good/Hard buttons that Anki has to differentiate between how well you know the answer, but that’s something I can live without. I use both it and Anki on a day to day basis.
I just played around with Memrise, and it does indeed look perfect for my audience. I had begun my SRS search with gwern’s excellent exploration of the topic, where Memrise does not appear. Thank you so much!
I’m glad you like my recommendation. After you have used it for a while, perhaps consider writing up a post about your experiences teaching using an SRS. It’s a topic which could be very interesting, and I’m sure that many would wish to read such a report. I certainly would.
I’ve started making heavy use of archive.is. You give them a link, or click their super-handy bookmarklet, and that page will be archived. I use it whenever I’m going to be saving a link, now, to ensure that there will be a copy if I go looking for it years later (archive.org is often missing things, as I’m sure we’ve all run in to).
Aging doesn’t necessarily resemble the human pattern
That leftmost of the five graphs, the one that shows in its shaded area the proportion of the hydra sample that still lives, confuses me. What could possibly cause a linear decrease in population? A constant probability of death given previous survival (podgps) would produce a decreasing exponential curve, but this requires a podgps for each individual that is exactly hyperbolically growing over time, for example through an internal candleclock that is at birth set to a uniformly distributed time between 0 and 1400 years.
(Yes, hyperbolically. As the time interval lengths during which we record whether the hydra died go to 0, the ratio of podgps between the last time segment and the, say, 3rd goes to infinity. The crux of the matter lies in “given previous survival” and if we leave that part out, the result would indeed be a constant line.)
Maybe the researchers found by experiment or hearsay that the hydra has a constant podgps, the graphmakers failed to record that last bit and simply took the negative integral of that absolute probability of death to generate that ridiculous population curve and then the authors found that that integral curve hits 0 at 1400 years and postulated that as their livespan.
(By the way, it’s also troubling that they also left out the bit of the curve that would have indicated infant mortality.)
Full paper: Jones et al. (2013)
See the label for Figure 1. I’m not sure why army1987 retracted his comment, but he is correct: the y-axis is logarithmic for the survivorship curve. So the graph actually confirms your expectation and shows an exponential decrease in population.
(Unfortunately, the graph label in the National Geographic article is just wrong—there is no reasonable interpretation under which the logarithmic survivorship curve can be interpreted as a raw proportion.)
On the last panel (that for hypericum) of the figure on the NatGeo page the red curve doesn’t look like the negative derivative of the gray curve, so I assumed I was missing something.
Maybe the y axis is logarithmic or something.
This write-up: http://michael.richter.name/blogs/why-i-no-longer-contribute-to-stackoverflow/
And also one of the main issues he discusses:
http://en.wikipedia.org/wiki/Iron_law_of_oligarchy
Seem to be relevant to LessWrong as well, to some degree. How can we avoid the problem of ‘Creeping Authoritarianism’?
Is the Iron law of oligarchy essentially a Goodhart’s Law applied to humans? Like: You want a group of humans to accomplish something useful, so you create a system to resolve conflicts, e.g. a democratic majority vote. Sooner or later people learn how to win the majority vote by optimizing for winning the majority vote, without accomplishing much of what you originally wanted them to do. -- And if you try to fix this by adding some safety mechanism X to the democratic vote, then people will simply optimize for the majority vote plus X. For example in addition to elected politicians known to optimize for popularity, you add unelected bureaucrats who are supposed to be the experts, but somehow those just entrench themselves in the bureaucratic system regardless of their level of expertise.
If so, then essentially there is no safe way to solve this. If we measure something, then Goodhart’s Law attacks. If we don’t measure something, then… well, just because you are not looking at something, it doesn’t mean it’s not there… in the absence of explicit rules, the implicit rules will decide; the most popular people will simply be the most popular people.
All we can do is to use are some heuristics, and remember the nameless virtue; i.e. to change or abandon the heuristics when they stop being reasonable. We must keep thinking and updating, and thinking and updating, again and again.
Specifically, I have already noticed the Goodhart’s Law in action on StackExchange. Instead of helping other people it’s more and more about getting more points than other contributors. For example, you start writing your answer before you even think about it completely, because posting the partial answer and editing it later is better than thinking about it, posting it, and finding that someone else posted their very similar answer 1 minute sooner. So it’s “use google, post the first information found, use google more, edit your answer to include the additional information” cycle. And if the question cannot be solved by googling, nominate it for deleting, as off-topic or something. If no one can answer the question fully, then criticize the partial answers of other people as incomplete and downvote them; if you couldn’t get points by using your strategy, they don’t deserve them either.
Improving measurements is one of the boring but massive levers we have at our disposal, e.g. givewell, the technical details of how voting schemes capture preferences etc.
Stackoverflow seems to succeed in growing despite the issues that Michael brings up. Yes it’s not the place it was three years ago but it’s okay that communities change.
This very short xmas story was mentioned by Eliezer on Facebook: https://www.fanfiction.net/s/9915682/1/The-Last-Christmas. I wonder what the ending means (speculations and spoilers welcome, rot13, if you like).
If someone who reads Chinese could translate the line on the final page (rot13-ed), please do.
Santa Claus, according to Google.
For some reason I couldn’t copy-paste from that page, so I couldn’t Google it myself.
That’s odd considering the earlier mention that she didn’t believe in Santa.
Ways to deal with the fact that fanfiction.net has sabotaged your computer:
Turn off javascript (I’m not going to tell you how to do that)
Select all (and copy and paste elsewhere). It will look like it only selected the header, but it actually works.
Tell google to translate the whole page from from Chinese to English by putting the URL into the left-hand box and choosing the language to translate from (the default is “detect language,” which of course detects English)
fanfiction.com doesn’t let you anymore.
Sorry if stupid question. Let’s assume that the universe (mathematical multiverse?) gives us observations sampled from some simplicity-based distribution, like the universal distribution in UDASSA. Can that explain the initial low entropy of our universe (fewer bits to specify), and also the fact that we’re not in a tiny ordered bubble surrounded by chaos?
ETA: I see Rolf Nelson made the same point in 2007. This just makes me more puzzled why Eliezer insists on using causality, given that the causal arrow of time comes from initial low entropy of the universe in the first place, so mathematical simplicity seems to be the more fundamental thing.
A low entropy microstate takes fewer bits to specify once you’re given the macrostate to which it belongs, since low entropy macrostates are instantiated by fewer microstates than high entropy ones. But I don’t see why that should be the relevant way to determine simplicity. The extra bits are just being smuggled into the macrostate description. If you’re trying to simply specify the microstate without any prior information about the macrostate, then it seems to me that any microstate—low or high entropy—should take the same number of bits to specify, no?
If you can encode microstate s in n bits, that implies that you have a prior that assigns P(s)=2^-n. The set of all possible microstates is countably infinite. There is no such thing as a uniform distribution over a countably infinite set. Therefore, even the ignorance prior can’t assign equal length bitstrings to all microstates.
It seems to me that macrostates should take very few bits to specify (e.g. temperature and pressure), compared to microstates within a macrostate (positions of each molecule). So there would still be a difference overall.
Equilibrium macrostates (such as the homogeneous soup at the end of the universe) can be described using very few macroscopic predicates. Non-equilibrium low-entropy states cannot. No need to go back all the way to the beginning of the universe; just look at the present macrostate. No need to look at the entire universe even, just look at the room you’re in. The amount of information you’d need to specify to pick out the current macroscopic state of your room out of all the other macroscopic states of similar entropy it could be in is in fact quite high. A few predicates isn’t going to do it. The lower entropy you get, the more structure there is in the macrostate and the more information you’ll need to specify it.
Now maybe you think that the number of bits required to specify this structure, high though it may be, absolutely pales in comparison to the number of extra bits required to pick out an equilibrium microstate. But think about what you’re doing when you specify a macrostate. You’re basically just picking out a region of phase space. Once you’ve picked out the region, you then pick out a particular point within that region to specify the microstate. In terms of the number of bits required, it makes no difference whether you start our by picking a larger region and then picking a microstate within it, or if you start out by picking a smaller region. Either way, you need to rule out all points in state space but one.
Of course, you might get the illusion of simplicity if you give your small region a special name, like “Bill”. You could say, “All I need to do to pick out this region is say ‘Bill’. It’s so simple.” But then you’ve just hidden the complexity away by choosing a gerrymandered description language. I’ll do you one better by naming the precise microstate of the early universe ‘Bob’, and now I can pick out that exact state just by saying that three-letter word. We have simple names for macro-properties that are salient to us, and that we can easily pick out perceptually, but that doesn’t mean those properties are actually simple from a fundamental physics point of view. The properties of classical thermodynamics correspond to gross constraints that we can control. They are relative to our epistemic and causal capacities. They are not nature’s joints.
Are you sure that all points in state space require the same number of bits to describe, if descriptions are computer programs? It seems to me that some states are more ordered and can be written out by a shorter program. For example, if all particles have zero velocity, that can be written out by a pretty short program. Equilibrium vs non-equilibrium doesn’t really come into it, e.g. a pair of very ordered objects about to collide at high speed could have a very short description and still lead to a big bang.
I agree that the constants depend on the choice of programming language, but that’s a problem for K-complexity in general. I’d love to know the solution to that...
Cultivating anti-fragility
Can anyone suggest visual symbols for Reason? I need one for a project I’m working on.
Specifically, I’m looking for something that could represent the concept of Reason, but isn’t associated with any modern politics, and doesn’t rely on an understanding of modern science. i.e. the first thing that came to mind was a sketch of an atom, but that won’t suit. It should be recognizeable to a pre-industrial scientist or philosopher.
On googling, the best fit I’ve found so far was a lit candle; Reason as a light in the darkness. That should give an idea of where my head’s at.
A geometric pattern or construction. Maybe a golden spiral. Or something that looks like it’s out of Da Vinci’s notebooks.
A map, or a star map.
A perspective drawing with very obvious foreshortening.
Something based on the allegory of the cave or the divided line of Plato? The latter probably wouldn’t stand without explanation, though.
A version of what you think: the title image from this blog post: http://www.studentguidewebdesign.com/be-rationale-why-creative-rationality-is-important/ , or maybe from this one: http://rationallyspeaking.blogspot.ca/2012/08/the-community-of-reason-self-assessment.html
The first link is broken, and the second one doesn’t seem too relevant. How about the Thinker? Or a picture of a brain?
Fixed first link (stray comma at the end)
Knowledge that the brain is the seat of intelligence might be too high-tech, but I think The Thinker is a good idea.
I donated a small amount of money to CFAR and Archive.org when there were matching campaigns because both organisations accept bitcoin as a payment option. I liked the feeling but as a student it is quite an expensive experience though I will do it again in the future. One thing I’d like though is to be notified about is when my favorite charities have matching campaigns going on so that my contribution goes a longer way. Is there a way to get emails on those occasions without being spammed every other day about stuff I really do not care about?
You can subscribe to a monthly newsletter at intelligence.org.
It’s one short email a month. While not exactly a notification for a campaign, it should always inform you in time when the matching campaign is running.
So it would take about 15 favorite charities before on average Metus gets “spammed every other day.”
A new paper gives a much better algorithm for approximating max flow in undirected graphs. Paper is here. Article for general readers is here. Although the new algorithm is asymptotically better, it remains to be seen if it is substantially better in the practical range. However, this is an example of discovering a substantially more efficient algorithm where one might not have guessed that substantial improvements were possible.
Is anyone out there irritated with the concept of anti-fragility?
It seems to me to unnecessarily merge three different concepts: evolution by selection, learning, going out from local optimum by means of random stimulations.
The claim that there are systems that benefit from disorder, exactly as stated, to me has only one instance, not even a real thing: Project Orion.
Can you explain the concept of antifragility using only 1000 most common English words?
That’s a fun challenge. It was hard to try to summarize the motivation behind the idea of antifragility in such a restricted vocabulary. Here is my attempt:
I think you want to separate out “bad” from change.Things sometimes break, this happens when outside forces cause changes to it and the world it acts in. We call a thing fragile when most changes are bad from the perspective of the thing. EG ice is fragile because changes in temperature, motion, etc. will cause it to break. Anti-fragility is when the thing is designed such that the biggest possible changes do not break it.
I see what you are saying, but the whole point behind anti-fragility is that change is for good, not bad. By default, in fragile things, change is bad. But in antifragile things, that change is harnessed for good.
Hm. The best way to clearly demarcate that would probably to move the word “bad” from describing the word “change”, and put it as part of the first sentence.
This is my attempt:
I’d like to be able to watch posts. Currently, I can see if someone replies to a post of mine, but not any other post. Sometimes there are posts where someone asks an interesting question, but nobody has answered it, and that person isn’t me. There’s no way for me to tell when someone replies to it so I can read the reply.
Every post and every comment has an envelope icon in the lower right, just to the right of the permalink chain icon, just to the right of the speech bubble reply icon. Click on the envelope and it will be highlighted; click again to go back to gray. While highlighted, replies will show up in your inbox. It is only for immediate replies, not replies to replies.
I DID NOT KNOW THIS. THANK YOU.
It’s fairly recent. I want to say it showed up a few months ago?
test
I have recently seen this article on a drug, which is as yet not approved by FDA due to lack of controlled study. However, the proponent doctor considers the drug a life saver for babies and the existing evidence sufficient to skip the trial phase and go ahead saving lives.
I am usually fond of evidence based medicine. However, in this case, I am shaken by the fact, that the drug is already in use in Europe for some time, with no scandals. Additionaly, it is basically a mixture of lipids from fish oil, which sounds like normal nutrition to me, not a novel chemical with unpredictable effects. (Although the application is intravenous, which is not “natural”).
I believe no finding will be found, which will, in retrospect, justify the current state of denying access to it to fataly ill children patients. No side effects of the kind of “Oh, I wish I let my baby die back then !” (Disability ? Chronic pain ? Mental retardation? - I do not expect any of that… After not seeing it in Europe so far.)
I want to transform this belief into a bet in prediction book. Can anybody give me advice, how to formulate the bet ? What should be a deadline for the bet evaluation ? It will be my first bet in prediction book...
Not for infants.
Theoretically, that could make a difference. In my country omegaven is available, but the pack insert gives warning about insufficient experience in children.
Anyway, I still want to bet. What is a fair time span ?
(BTW, do I understand it correctly, that in the USA, omegaven is approved neither for infants nor for adults ?)
Does anyone else find the free will solution sequence totally unconvincing?
Donations to one of GiveWell’s top rated charities (GiveDirectly) are being matched 1:1 up to January 31st. If the $5 million available for matching isn’t used up then there is a good chance that it won’t go to GiveDirectly, so this genuinely is an especially good time to give.
EDIT: I just donated.
In a recent comment, II expected that my question might have been already answered so I wrote this:
I knew this was arrogant, so I appreciated the humor of this reply:
I wanted to explain here why I did not read the previous posts.
There are roughly three nested reasons:
first, it was an experiment because I am often tempted to write something like this (in fact, I have in less egregious cases) and I was wondering what kind of norm-violation response I would get. I was “testing”, because this seems like a nice thing to get away with. The fact is I don’t have time, but if i want to participate in a discussion sometimes I take a gamble and ask a question anyway. I don’t provide a disclaimer that I haven’t read all (or any) of the background, but it would be nice to be able to.
it was symmetric that it was suggested that I ‘hire’ someone to write a summary, because when I am more conscientious and research something just-to-be-conscientious, I often self-criticize that Less Wrong is not “paying me by the hour”. (That is, it’s not supposed to be work.)
finally, it’s not just me that needs to do all this ‘work’. How many people have read all the previous posts? I think it would be nice if executive summaries were provided more often (I provided one here at the top, as well as links) but if that’s too much work (sure it can be) then it could be more of a norm that a person can just plead ignorance / request a relevant recap.
(Ironically, has an open thread like this already been posted?)
I know there are some R Scott Bakker fans on here, and I was thinking recently about the Second Darkness series. Rot13d for spoilers:
Vg’f n funzr gur pbafhyg ner rivy. Vs gurl jrera’g fb pbzzvggrq gb rivy npgf, gurl pbhyq ratvarre n jnl gb tvir rirelbar n unccl raqvat.
Jr ner gbyq gung fbepreref ner qnzarq, naq gung gur hygvzngr tbny bs gur Vapubebv vf gb erqhpr gur ahzore bs yvivat fbhyf ba gur cynarg gb srjre guna 144,00 va beqre gb frny gur cynarg sebz gur Bhgfvqr naq rfpncr qnzangvba. Gur Pbafhyg pbhyq erpehvg nf znal fbepreref nf cbffvoyr, genva gurz gb terng cbjre naq gura rkgraq gurve yvsrfcnaf jvgu gurve Grxar, juvyr hfvat gurve vasyhrapr gb zbir gur phygherf bs gur jbeyq gbjneqf n zber uhznar naq pvivyvmrq zbenyvgl (be, tvira gung gurer ner nccneragyl jnlf gb trg qngn ba ubj gb orpbzr fnirq be qnzarq, ng yrnfg gbjneqf n fbpvrgl gung vf uvtuyl rssrpgvir ng qvfpbhentvat qnzavat npgf). Bapr n fhssvpvrag ahzore bs fbepreref (fbzrjurer whfg orybj gur 144,000 svther) unir orra npphzhyngrq naq gur jbeyq pbafvfgf cevznevyl bs aba-qnzarq vaqvivqhnyf, gur Pbafhyg pbhyq hfr gurve uhtr zntvpny cbjre gb dhvpxyl trabpvqr gur ragver cynarg, oneevat gurve bja zrzoref. Zbfg fbhyf jvyy or fnirq, naq gur qnzarq fbepreref jvyy erznva ba n frnyrq cynarg, univat nibvqrq qnzangvba.
Guvf ng yrnfg nccrnef gb or n jnl gb trg gur znkvzhz ahzore bs fbhyf n unccl raqvat—juvyr jr qba’g xabj jung unccraf ba gur cynarg bapr vg vf frnyrq sebz gur Bhgfvqr, gur Vapubebv frrz vagrag ba npuvrivat fb gurl zhfg oryvrir vg’f qrfvenoyr.
People who post probability estimates of anything should explain in details how they arrived at them. Otherwise it should be called not probability estimate but pulling it out of your ass.
Seriously, stuff like “1% FAI success by 2100”? When there is no clear definition of AI in sight? Just stop.
All beliefs are probability estimates, although it can be hard to trace how a particular belief got to the degree of confidence it’s at, and while it might be a nice norm to have in a perfect world I think it’s unreasonable to demand that every time someone expresses how confident or unconfident they are in a belief, they should also clarify the entire precise history of that belief’s presence in their mind.
Apologies for the curmudgeonliness, but it really bugs me when people say things like this. The actual version of this statement that is true is
which is different and much weaker, as now we can argue about how important coherence is relative to other desiderata. One such desideratum is correspondence to reality, which I believe is Locaha’s point above. Personally, I would much rather have incoherent beliefs that correspond to reality than coherent beliefs that do not correspond to reality.
I don’t think this belief has much correspondence to reality.
If It’s Worth Doing, It’s Worth Doing With Made-Up Statistics
Whenever I post a probability estimate, it is solely for the purpose of making my position more clear, not as something anyone should use to actually update their beliefs. You should always consider probability estimates as rough information about how the person who made the estimate thinks, not as a factual bit of information about the prediction itself.
I don’t think you make anything clearer by translating your intuition’s unlikely to N%, while my unlikely is M%, where M!=N. You just make a false impression of having done a calculation (which, unlike intuition, can be confirmed).
Suppose passive_fist translates “unlikely” as 2% and Locaha translates “unlikely” as 12%. This could mean either of two things (or some combination of them). (1) passive_fist applies the word “unlikely” to things that feel more unlikely, corresponding to lower probability estimates when forced to quantify. (2) Both actually think much the same about the event in question, as shown by their use of the same word, but they have quite different processes (at least one of them very inaccurate) for translating those thoughts into numbers.
In case 1, quantifying helps to clarify that the two people involved mean quite different things by “unlikely”. There may be a lot of fuzziness about the numbers, but once we have them we can see that passive_fist will likely be much more surprised if something s/he calls “unlikely” happens, than Locaha will be if something s/he calls “unlikely” happens.
In case 2, quantifying just adds confusion and error.
I would expect that (especially for analytical quantitative types like most of LW’s readership) the truth is something like this. We think, mostly, in fuzzy terms that don’t correspond directly either to numbers or to words. There will be some region of subjective likelihood-feeling space that corresponds (e.g.) to the number 2% or 12%. There will be some region that corresponds (e.g.) to the word “unlikely”. These correspondences will all work differently for different people, but (a) there will generally be more consistency between one person’s “10%” and another’s than between one person’s “unlikely” and another’s, and (b) the finer-grained information you get by asking for probability estimates does have some value, provided you’ve wit enough not to imagine that everything expressed numerically is known accurately.
[EDITED to fix formatting screwup.]
Plus, some people here use stuff like PredictionBook to check whether the intuition they call “10%” is actually correct 10% of the time.
Probabilities are useful for being precise about the claims that you are making. There no reason why one shouldn’t be precise about the claim one is making even when one doesn’t use a formal method to arrive at them.
If you don’t use a precise method to arrive at your claim, you have no business making a precise claim. Remember significant figures from high school chemistry? Same principle.
I think this is an error. (And so are “significant figures” as commonly used.) 2.4 +- 2 and 2.0 +- 2 are quite different estimates even though you wouldn’t (according to conventional wisdom) be justified in giving more than one “significant figure” for either.
Using the number of digits you quote to indicate how accurately you think you know the figure as well as to say what the number is is a hack. It’s a convenient hack sometimes, but that’s all it is. Everyone knows not to round intermediate results even when starting with low-precision numbers. Well, your final result might be used by someone else as an intermediate result in some bigger calculation.
The same goes for probabilities. It is very important to know when your estimate of a probability is very inaccurate—but that’s no reason to refuse to estimate an actual probability. Even if you just pulled it out of your arse: doing that makes it a very unreliable probability estimate but it’s still a probability estimate.
I won’t deny that significant figures are a crap implementation of the principle I’m talking about. But you have to propagate the uncertainty and include it, in some way, in your final answer, either numerically or via some explanation that might let me figure out how precise your answer is.
Don’t say “1% probability of FAI success by 2100.” Say ”.01-10% probability of FAI success by 100 based on XYZ.” Or if there’s no numerical process behind it that can support even a range like that, just say “FAI success by 2100 seems unlikely.”
Agreed. Though in the latter case you might still do best to give numbers: “Somewhere around 1%, but this is a wild guess so don’t take it too seriously.” This is not the same statement as the corresponding one with 2% instead of 1%, even though both might be reasonably accurately paraphrased as “unlikely” or even “very unlikely”.
That assumes that someone isn’t calibrated. If someone calibrates his intuition via frequent usage of prediction book and by always thinking in terms of probability he might be able to make precise claims without following a precise method.
If someone would claim “1.21% chance of FAI” success by 2100 I would agree with you that the person didn’t learn the lesson about significant figures from high school chemistry. I don’t the that issue with someone claiming 1% chance.
If you want to get calibrated it’s also useful to start putting numbers on a lot of likelihoods that you think about, even if the precision is sometimes to high. It allows you to be wrong and that’s good for learning.
I think it’s likely that calibration is domain-specific, so I’m not sure I buy this unless the calibration has occurred in the same domain, which is rare/impossible for the domains we’re talking about.
I think you can argue that the probability is inherently unknowable but I don’t see how a detailed process is much better than an intuitive process.
It’s very useful to have a mental ability to distinguish between 0.01, 0.001 and 0.0001 when it comes to thinking about XRisk events. I don’t think that it’s a good practice to call all of those events unlikely and avoiding to make semantic distinctions between them.
But how do you arrive at them? Intuition doesn’t deal with 0.01 and 0.00001. Intuition deals with vague notions of likely and unlikely, which also change depends on what you ate for lunch and the phase of the moon. IOW, your intuition is useless to me unless I can confirm it myself. (But then it’s not intuition anymore.)
I think there are plenty of cases where I can give you a intuitive answer that won’t change from 0.01 to 0.00001 depending on what I ate for lunch.
The chance that I die in the next year is higher than 0.00001 but lower than 0.01.
If you don’t have an intuition that allows you to do so, I think it’s because you don’t have enough exposure to people making distinctions between 0.01 and 0.00001.
If there’s a 0.01 chance that something happens tomorrow, then if everything stays the same you’d expect that thing to happen about three or four times this year, whereas if it’s 0.00001 you’d be quite surprised if it ever happens (EDIT: during your lifetime, assuming no cryonics/antiagathics/uploads). (Of course with stuff like x-risk intuition will be much less reliable.)
And that’s a good assumption, since by my estimate, 99.9537% of people are not calibrated.
Part of lesswrong mission is about moving to a world where more people are calibrated. I don’t think it’s helpful to declare calibration a lost course.
Shalizi had a nice post about that.
But the “50% probability of Situation A (2% probability of FAI in 100 years) and 50% probability of Situation B (0% probability of FAI in 100 years)” is much more informative to the reader than “1% probability of FAI in 100 years.” It exposes more about which parts of the estimate are pulled out of the writer’s ass. If I know something the writer doesn’t about any one of these component probabilities, I can update my own beliefs, or discuss the estimate, more usefully this way than if I’m just given a flat “1%.”
Anna and Steve had a nice post about that.
How is the belief of some random person X in some vague-defined event many years into the future is useful for anything nut the research into person X state of mind? Even if it’s defined to 1000 significant figures?
If you are reading the text of a person who presumably care about that person state of mind and what this person believes. If you don’t why do you read the text in the first place?
I do think there a difference between someone thinking an event is unlikely with p=0.2, p=0.01 or p=0.0001. It worthwhile to put a number on the belief to communicate the likelihood.
If people frequently provide likelihoods you can also aggregate the data.
I recommend operationalizing this by recommending that people ask “Why do you think so?” when they see a probability estimate.