If it's worth saying, but not worth its own post (even in Discussion), then it goes here.
Notes for future OT posters:
1. Please add the 'open_thread' tag.
2. Check if there is an active Open Thread before posting a new one. (Immediately before; refresh the list-of-threads page before posting.)
3. Open Threads should be posted in Discussion, and not Main.
4. Open Threads should start on Monday, and end on Sunday.
I wonder if the following is covered in the sequences... I could not find it.
There is a specific kind of argument which is not really an argument, because it is not just used in debates but people really seem to believe. It is a bit similar to motte and bailey, but that is a debate tactic, but this one is not, this is really believed.
The broad outline is statements that can have multiple interpretations, broader and narrower. And the broader interpretation is almost trivially true, while the narrower not and they get confused.
The latest example I saw was hedonism in the sense that everybody is a hedonist. Sure, someone working their ass off to be a champion do it because they think winning it gives them pleasure. Sure, the patriot selflessly fighting for his country and doing his duty is doing it because not doing so would give him a kind of psychological pain. This really really broad sense of hedonism is trivially true. But hedonism has a narrower, "sex and drugs and rock and roll" sense, let's call it instant gratification, and no, it is not true that everybody is chasing that.
The point I am trying to make is that I think I need to sort it out in my head whether I bel...
Idea: how NOT to use statistics for self-help
It is tempting to simply look at what seems to work for most people and try that. And when trying that costs little money and not a big time investment, neither a long waiting time, then sure, why not?
But when it is not the case, here is something to consider: why aren't you already doing what most people are doing? If you and your situation was exactly average, you would probably do it already. It is likely that there is something special about your case - and it can suggest the common solution would not work for you.
For example, most people find their romantic partners through their circle of friends. If you are social, if you have a large circle and hang out with them a lot, it would probably happen to you more or less naturally. If you are like 25 and it did not happen, it suggests something is different - maybe you are not very social, maybe you don't have much of a circle or you have the kind of circle that is not very conductive to this and so on.
The meta of this is that relatively easy to pick up popular ways of doing things. We adsorb it through our socialization, social life, upbringing, media, life in general. If you don't d...
In the spirit of asking personally important questions of LessWrong, here goes. Please be gentle with me.
Related:
Discussion post by another user on being raised by narcissists
My parent always had a number of narcissistic traits, but was never a full-blown narcissist. They (singular) supported me financially and always seemed to legitimately care about how well I was doing academically and professionally. However, they had a habit of lowering my status by verbally critiquing my actions, and sometimes made odd demands of me, such as demanding that I share some of my passwords with them, or demanding that they be present every time I go to the doctor (I'm 25).
Right now, I think that I'm either going to severely limit contact with my parent, or cut contact completely. I think that cutting contact completely is likely to be more pleasant and easier on me, but I'm really not sure about that yet. I've had a few family members tell me that I'm obligated to keep in touch with my parent. Since LW is my in-group, and since I share lots of values with the kind of people who tend to post here, I'd prefer to get advice here, rather than elsewhere. Specifically, I'm not s...
I was in a similar situation with my parents in my early 20s (although their motivations and characteristics were probably very different). Looking back I think they were not ready to deal with my independence (I was the oldest) and tried to deal with things in the same way they did when I was a child. Your mention of medical appointments really rang a bell with me - my parents did the same and this made me perhaps the most uncomfortable of all.
In my case, severely limiting contact was a highly successful approach. I didn't do this explicitly; we had no conversation "I am limiting my contact with you". I took a job in a different place, got my own (tiny, horrible) room in a shared flat and just started being more independent in my life. I must admit I used the workload of my new job as a convenient excuse to limit contact - just the occasional phone call to let them know how things were going (back in those days my parents didn't have email). I also didn't visit nearly as often as before --- I found myself reverting back to a teenage mentality, they would treat me like a child, I would get very angry/upset.
Now many years later (I'm approaching 50) I have a somewhat positive relationship with both parents - perhaps in large part because I live in a different country. After a lot of time had passed we were able to discuss the earlier issues more dispassionately (although not entirely) & get a better understanding of each other's motivations.
The Computerphile people, who are great at explaining IT, have made a video on AI risk a week ago. It already got 100K views.
https://www.youtube.com/watch?v=tcdVC4e6EV4
No new content for this crowd here, but I think it is particularly well argued and very accessible for a laypeople audience.
One lesson from the Tim Hunt affair: Always make a recording with your smart phone when you give a speech.
You want to be able to proof what you actually said.
Has anyone tried advertising existential risk?
Bostroms "End of Humanity" talk for instance.
It costs about 0.2 $ per view for a video ad on YouTube, so if 0.2% of viewers give an average of 100 $ it would break even. Hopefully people would give more than that.
You can target ads to groups likely to give much by the way, like the highly educated
"AI safety" suffers from some of the same terminology problem as "computer science".
It is written that "computer science is no more about computers than astronomy is about telescopes." The facts of computer science would be true even if there were no computers: facts such as the relative efficiency of different algorithms, or various ways to index records. If the quicksort or the hash table had been discovered in a world without computers, we would think of them as belonging to library science, or bookkeeping, or some other di...
So there's a dilemma I've been grappling with for the past year or so; I want to start a blog, since I like writing and would like the public accountability to make me write more often. The problem is, I can't think of anything worthwhile to write about. I don't want it to turn into journal entries, but anything I could possibly want to write about has already been written by someone, somewhere on the internet. How do people... get their ideas? As cliche as that question is, it's still a puzzle I can't figure out.
Any tips for boosting one's kindness? I know it's closely related to the Big Five trait of agreeableness and Big Five traits are hard to change (relative to one's age cohort). Perhaps mindfulness meditation could have a long-term effect? One could always just do a bunch of kind actions but I don't think it's that simple.
i want to easily found out about sub comments on posts I've made - comments on other people's comments but i don't get notified for these. Just now I found a really helpful subcomment and I fear ignoring something of counterfactual consequence! Any workarounds?
DAE think they look their best when they have their head lean back, chin tucked down?
Random thing that I can't recall seeing on LW: Suppose A is evidence for B, i.e. P(B|A) > P(B). Then by Bayes, P(A|B) = P(A)P(B|A)/P(B) > P(A)P(B)/P(B) = P(A), i.e. B is evidence for A. In other words, the is-evidence-for relation is symmetric.
For instance, this means that the logical fallacy of affirming the consequent (A implies B, and B is true, therefore A) is actually probabilistically valid. "If Socrates is a man then he'll probably die; Socrates died, therefore it's more likely he's a man."
Sorry, this was an useless post so now it's gone
"How useful is my contribution, given what others are already doing?What will happen if I don't do it?What are the chances of success, and how good would success be?"
- Giving What We Can newsletter
Goals
You want a status signal, a martial arts partner, a sexual partner, an emotional partner, a psychological partner, a fitness partner, AND a life coach. It seems you're making things very hard for yourself as you are narrowing the pool of possible partners to a very extreme extent. No girlfriend I've had ever had was all of these things; the best and most compatible one was maybe three of these things (sexual, psychological, and fitness partner). There's nothing wrong with having several people around you that fulfil various roles. One person can be a sexual partner and another one a non-sexual fitness partner. This is the arrangement I currently have, for instance.
Strategy
Statistically, most people find their partners through their friends. Thus it seems a good strategy for finding a girlfriend would be expanding your general circle of friends. This has certainly been the case for me - whenever I've had a larger circle of friends I'd had an easier time finding a partner.
Voice
I wouldn't take independent diagnoses of autism very seriously; they aren't doctors (I'm assuming) and 'autism' seems to be a word that gets thrown around too much these days.
Sure, I can explain. Bear in mind that this is all based on my personal experiences (male, atheist, mid-to-late 20s, college degree, lives in Seattle, WA, only interested in dating women) and that although I have developed it over around four years I'm not claiming I've found the perfect strategy so far.
First of all, filter match ratios pretty hard. Anybody below a 90% is probably not worth checking unless they checked you first, below an 80% not even then. Above that it starts being more a matter of enemy ratio; above 10% is probably not worth it, above 15% quite unlikely. 95%+ match and 5%- enemy is always worth checking out.
Next, take a quick look for dealbreakers (you do have a list of dealbreakers, right? Not things like "overweight" unless you are super opposed to that, but things like being a smoker, or "I don't really read stuff, lol", or being significantly religious). Many people also list their dealbreakers; make sure you aren't on them (sometimes it's little stuff, like having a beard or being too short; seems silly but just move on and don't waste the time). Distance or location is common one; not all 25 mile distances (or whatever threshold you set...
Agreement on CBHacking's points.
I found the match factor to be very predictive. With an ex-boyfriend of mine, the boyfriend I found via okc and a more recent one I had 99% match, though the maximum height of the match factor is constrained by amount of questions answered and the way you answer them, so you might not get that high in the first place. 95% is really decent, I never found anyone <80% interesting enough to talk to for longer.
For the enemy thing I recommend checking the answers marked "unacceptable" that go into the factor calculation. Sometimes these come merely from interpreting a question differently.
I'm open to describing which strategies would work for me (24, female, white, European), but I am not sure how much they generalise. I rely on profile text quite heavily for getting an impression of the other person and will often send the first message. I'm informed that isn't typical though.
Some types of messages I got: 1.) mass messages Just "Hi" or "Hi :)" or "Hi how r u" or similar. These are very common. I tried to talk to some of those people and the conversations tended to be extremely boring, uncreative and the people ...
looked up monotone voice on google, and found that it has a positive, redeeming side – attractiveness.
My friends tell me that my face is pretty scarred. Research shows that facial scars are attractive. By the word scar, researchers mean healed cut. My friends mean acne hole.
Not all monotone voices are created equal. I'd be really surprised if "autistic" monotone and a "high-status" monotone would refer to the same thing.
Here's a potential existential risk. Suppose a chemical is used for some task or made as a byproduct of another task, especially one that is spread throughout the atmosphere. Additionally, suppose it causes sterility, but it takes a very long time to cause sterility. Perhaps such a chemical could attain widespread use before its deleterious effects are discovered, and by then, it would have already sterilized everyone, potentially causes an existential catastrophe. I know this scenario for causing an existential catastrophe seems very small compared to other risks, but is it worthy of consideration?
On the face of it, I don't feel that this particular risk differentiates itself enough from "what if [insert subtle end-of-times scenario here]?" to be worthy of specific consideration. It's a lot of what ifs and perhapses.
Could someone point me to any existing articles on this variant of AI-Boxing and Oracle AGIs:
The boxed AGI's gatekeeper is a simpler system which runs formal proofs to verify that AGI's output satisfies a simple, formally definable. The constraint is not "safety" in general but rather is narrow enough that we can be mathematically sure that the output is safe. (This does limit potential benefits from the AGI.)
The questions about what the constraint should be remains open, and of course the fact that the AGI is physically embodied puts it i...
The only one I can think of (though I can't find the specific article) is Goertzel's description of an architecture where the guardian component is separate from the main AGI
You're probably thinking of GOLEM. The Gödel machine is another proposal along somewhat similar lines.
Some discussions more directly related to your suggestion could be:
...Our proposed protocol is based on the idea of asking what we will call ‘safe questions’ with restricted answers. We define a safe question as one for which a human being could find an answer without any help from superintelligence, all on his own merit. Why is this useful? Suppose that a scientist working on a cure for cancer is looking at two potential cures. Each one could be developed and tested without assistance from the AI, but would take about 3 years of hard work each. Which one to try first if the scientist believes that they have about an equal chance of working? If a wrong option is chosen to be tested first, humanity will still obtain a cure for cancer but it will happen 3 years later. What if we could ask the AI to suggest which option to try first? [...]
To make sure
How do I add probabilities? Say I have 23% chance of A, and 48% chance of B, what are the chances of either? I used to think I would just add the probabilities, intuitively...then I came across problems where it sums to greater than 100%, but it's not certain. If you think like I used to think, this abstract example won't help you. So I'll give a descriptive version below. For anyone who can explain it to me, feel free to skip the next part:
Say Jimmy wants to destroy an unwanted statue. From research on statue destruction, he believes there is a 95% chance...
I've started learning Machine Learning (he!), and upon reading the first chapter of the most famous textbook I was already gasping for air.
For someone like me who grew into probability with Jaynes' book, seeing in the first chapter that algorithms are trained using multiple times the same data (cross-validation) was... annoying, let's say (I actually screamed at the book).
Is there a sane textbook on machine learning? I don't demand one that starts from objective bayesianism, that would be asking too much. But at least something that assumes bayesianism as a foundation? Pretty please?
For someone like me who grew into probability with Jaynes' book, seeing in the first chapter that algorithms are trained using multiple times the same data (cross-validation) was... annoying, let's say (I actually screamed at the book).
There's two ways to train algorithms 'multiple times' on the same data. The bad one is data duplication, but cross-validation is the good one. Data duplication is the sort of thing that Jaynes would have been worried about, because it means you're counting evidence from the same piece of data twice, thus your model has illusory precision.
But what does cross-validation do? There's an issue called "overfitting," where any statistical procedure performed on a training set will fit both the noise and the signal in the training set, but while the signal on a test set will presumably be the same, the noise will be different and thus the model will do worse. Single validation is when you split your data into two parts, the training set and the test set, so that you can see how well your model trained on the training set does on the test set. When there's a tunable parameter in the training method, people will sometimes optimize the tunable para...
^ Above post is the illustration of the danger of LW's style Bayes. Below is a non-crazy discussion (e.g. one where people don't scream):
http://andrewgelman.com/2013/12/10/cross-validation-bayesian-estimation-tuning-parameters/
Suppose an agent has to choose between two main options. He can choose neither or either, but not both. His preference for each of the options is unknown, probably even by the agent himself (behaviour during the experiment signals indecision). Picture the experiment akin to a subject getting to make a pick between two useable objects such as toys, cars, gadgets etc. He is allowed to play/experiment with both of them during the experiment. Throughout the experiment, he exhibits a moderate preference for one of them, and spends more time using it.
Then, the e...
Has anyone tried advertising existential risk?
Bostroms "End of Humanity" talk for instance.
It costs about 0.2 $ per view for a video ad on YouTube, so if 0.2% of viewers give an average of 100 $ it would break even. Hopefully people would give more than that.
You can target ads to groups likely to give much by the way, like the highly educated
Youtube ads are very annoying. If someone's first acquaintance with existential risk was through youtube ads, he would get a very bad first impression.