Return of the Survey
[UPDATE: Survey is now closed. Thanks to everyone who took it. Results soon. Ignore everything below.]
Last week, I asked people for help writing a survey. I’ve since taken some of your suggestions. Not all, because I wanted to keep the survey short, and because the survey software I’m using made certain types of questions inconvenient, but some. I hope no one’s too angry about their contributions being left out.
Please note that, due to what was very possibly a bad decision on my part as to what would be most intuitive, I’ve requested all probabilities be in percentage format. So if you think something has a 1⁄2 chance of being true, please list 50 instead of .5.
Please take the survey now; it can be found here and it shouldn’t take more than fifteen or twenty minutes. Unless perhaps you need to spend a lot of time determining your opinions on controversial issues, in which case it will be time well spent!
Several people, despite the BOLD ALL CAPS TEXT saying not to take the survey in the last post, went ahead and took the survey. Your results have been deleted. Please take it again. Thank you.
I’ll leave this open for about a week, calculate some results, then send out the data. There is an option to make your data private at the bottom of the survey.
Thanks to everyone who takes this. If you want, post a comment saying you took it below, and I’ll give you a karma point :)
I took the survey. I will now take this opportunity to criticize it.
Severely missing items:
“Moral Views” should have had links to definitions; especially of “Eliezer’s interpretation”.
“Charity” should have had a “No opinion/I have no information on the matter” choice.
“Cryonics” should have had a “I have not yet considered the matter / No opinion” choice.
Minor missing items:
“Political Views” should have been split into questions on the particular attributes listed, for those of us who don’t bother to figure out which label applies to us.
“Time in Community” should have listed how many months ago LW and OB were created, for convenience.
Clarify your taxonomy:
If our universe is a simulation, does the containing universe have any significance to the Supernatural and God questions?
If our universe is a simulation, not of some other universe, but constructed from scratch, is the designer of it God in the sense of “Probability: God”? Does it matter whether this entity has edited the state of the simulation since the beginning of its time?
If our universe is a simulation which was designed to include ontologically basic mental entities, even though the universe in which it is simulated does not, does that count towards “Probability: Supernatural”?
I agree about the political views. The problem is any political parties I’ve come along with a huge amount of baggage in associated ideas and I’m very hesitant to claim membership with a political tribe lest I signal support for the crazy bits.
Edit: Somebody took my shiny new karma point. That was not very nice of them, since it was explicitly offered as reciprocation for survey participation.
Edit 2: It is back, or has been replaced by someone else. Thanks!
Edit 3: …I’m off to cognitive-dissonance myself into believing that I just like taking surveys.
Yay!
I just sent you a private message; please check your inbox.
Anyone who takes this survey and doesn’t post is probably trying to signal that they don’t care about karma. Anyone who does post is probably a karmawhore. So if you don’t take the survey, you are probably trying to signal that you don’t care about karma and you don’t care about signaling that you don’t care about karma. It appears that no matter what you do on this survey, you are probably trying to signal to the community that you are valuable to the community.
And, yes, I took the survey.
...and if she had led a good and proper life, this too was a proof, for witches dissemble and try to appear especially virtuous.....
Of course some people who post can effectively obscure the signal that they care about karma by trying to display an alternate valid reason they posted.
But two of the signals you mention are invisible!
my signals aren’t invisible to me… (What I tell myself three times is true, etc.)
But more to the point, anyone who didn’t keep the data private could still be signaling. I think I misread the post and the instructions.
Took it, and I was way off on the calibration question. I gave fairly low confidence, but if you’d asked for an entire confidence distribution I’d probably have give something pretty foolish...
I took it. And I decided I valued the karma point more than keeping possible anonymity. (Actually the karma point minus the probablity that someone would vote this down for being self-serving.)
Ok done… and registered for my first point (if you don’t mind!). Enjoyed the probability questions.
Issues I had were,
inadequate categorisation of political views.
inadequate categorisation for morality. I don’t think many people here would answer deontologist, I don’t know what Eliezers consequentialism is, so its basically a choice of consequentialism or other.
Most of the caterogies are well established and main stream umbrella terms that assume a good corelation between category and view set. Actually I think they should be avoided as associating with them comes down to tacitly agreeing with more than one proposition. So more single proposition (for/against) questions would be better …
for question about charity. You should reread the options, its like asking for 1-10, but listing only 3,4.5 and 9 bananas.
existential risk. Would say nuclear war/climate change are the most likely, but not the most likely to wipe out 90%+ of the population, needless to say my answer cannot represent this.
Overall I thought it was good stuff.
I took the survey and am suddenly having a memory-blank as to whether I remembered to comply with stating probabilities as percentages or not. I hope I did.
(My memory has returned. I did the probabilities correctly.)
I’ve probably missed something important. Please explain to me, how else can you obtain probabilities other than gathering statistics?
You already have an (implicit) prior probability distribution over future events without exact precedent, roughly proportional to how much you’d be surprised to see any of them happening.
If someone offered you the bet (at even odds) that a fully automated (Sebastian Thrun-style) highway would be developed before human-cat hybrids were developed, I think you might well take that bet. That’s a statement of your probability distribution.
These probabilities are much harder to calibrate than probabilities of frequencies of random events, but the important thing is that you already have rough estimates for them.
deleted
I dunnit.
I know what my opinions on these issues, but apparently I don’t practice translating them into actual probabilities enough. I wonder, should we always keep a number in the back of our mind?
Funny, ‘karmawhore’ was the 1st term that leapt to my mind, too. And yes, I did take the survey and am one....
Hal will disagree I’m sure.
So I’ve got one week to figure out the solution to Newcomb’s Problem and the Prisoner’s Dilemma...
I didn’t take the survey, but I would have done if I hadn’t been away on holiday for the entire time it was up. May I suggest that future such surveys remain open for a little longer?
The questions about longevity was a bit confusing. You asked what the chance of somebody currently living would live past 1000 is, but it’s not clear whether a person who gets lives for some time less than 1000 years, then gets frozen and lives longer has lived past 1000 years or not.
I did it. Where’s my cookie?
I took your survey, it was interesting. I’m still a little skeptical that my thoughts on various propositions take the form of probabilities in the first place. It seems absolutist to say there’s zero chance of supernatural beings, but “ontologically basic mental entities” fit so poorly into my worldview that you might as well have said “What is the probability that you are wrong about everything?” So, I said zero, although I wouldn’t ascribe such absolute confidence to myself if asked in some other way, I think. Similarly, your questions about God again pointed to this ‘basic mental entity’ idea, so I said zero again, although I’m sort of up in the air about the existence of beings beyond the scope of physical reality. It could all be a simulation, after all, and not necessarily of the same kind of universe as the simulator’s.
After you collect and release the survey data, and we all have time to stare at belief-spreads, will you do a re-survey to see if beliefs converge somewhat once we know one another’s impressions?
What does Electrical Engineering count as, under Professional Field?
EDIT: And why is not possible to select both student and for-profit work? I work as a programmer and study as an electrical engineer.
Yes, I’m a karmawhore too :-)
I took the survey! And I’m eagerly awaiting your upmod so my karma will return to an even number!
I’m skeptical about the “time well spent” part. You can probably guess which survey is mine now.
Done it. I was surprised how many questions I had difficulty nailing down a probability for.
I took the survey.
For the politics questions, it would have been nice to have been able to select answers to economic/social questions separately, rather than being forced to select a package deal. I wanted to select republican-style taxes paired with libertarian-style social freedoms, but that wasn’t an option.
(Before you blame me for not mentioning this sooner, the politics question was phrased differently on the sample survey.)
A few of these needed more caveats. Since I gave an answer less than 50 on “Existential RIsk”, I thought it was necessary to add a “conditioned on humanity living that long” rider to the Singularity question, etc.
Of course, I should have said so last week.
Also, there is a distinction (for those of us who believe MWI) between the probability that an average person frozen today will be revived (as measured in Everett branches where they get revived) and the probability of being revived from that person’s perspective. It seems to me that cryopreservation is the one clearest quantum suicide situation, and that the subjective probability of being revived should equal 1 if there is any Everett branch in which you are revived.
I took the survey. I was off by a century on the calibration question but I gave it a low confidence.
I had to put my ignorance out there, so i didn’t tick the private box.
That was the hardest survey I have ever taken.
I assume that in the political views question, under the liberal option, it should say “UK Labour Party” rather than “UK Tories”
I don’t regard the UK Labour Party as being permissive on social issues—for example they recently criminalised BDSM porn, and plan to introduce ID cards and monitor everyone’s Internet activity.
Argh, you’re right. Changed, along with one or two of kpreid’s points.
There may be left wing Conservative Party members in the UK who would be closer to Democrats in the US. I think UK centre ground is to the Left of US centre ground.
There are little checkmarks under the headings for section 2 and 3 - do those mean anything?
They mean I’m very bad at writing surveys, and I couldn’t figure out how to make a section divider without telling the code it was a “question”. All “questions” have to have at least one answer, which in the case of the section dividers was that checkmark. Anyone familiar with Google Docs and know how I can avoid that next time?
I’m sure some users will tick the boxes—I’d find it amusing if that correlated strongly with, well, anything.
I ticked the boxes, out of amusement that they existed. I’m also curious if the tickers and non-tickers systematically different.
I ticked the boxes as well. At the time, the first thought that occurred to me was “Easter Egg? Maybe if you check them you get something special at the end of the survey.” Too many video games, I suppose.
This doesn’t make any sense to me. MWI is an interpretation of QM—it’s just a piece of poetry that you put next to the maths to make yourself feel good. It doesn’t make any predictions that are not made by the maths, so it can’t be correct or incorrect. It can just be “a pleasing piece of poetry” or “a displeasing piece of poetry”.
Yes?
It makes the null prediction that the standard laws of QM apply in all situations, as opposed at least to collapse interpretations, which predict that some systems that should stay coherent won’t; this is getting more testable every day.
Oh, OK. But I thought that no-one actually believed in collapse interpretations any more? Anyway, if you take the “collapse” interpretation of QM, that’s a different mathematical theory, not merely a different interpretation pf the same maths.
Seriously, is there anyone left who actually believes in collapse any more?
Good catch.
Roger Penrose, at least (and he believes it for physical reasons, not because of his philosophy of mind).
Collapse actually looks better to me than anything else that’s not MWI, though I haven’t studied the issue in much depth – I don’t understand what any interpretations besides MWI, collapse, and Bohm/hidden-variables are even saying ontologically, and Bohm has serious zombie problems.
They all seem to reduce to many worlds, hidden variables, collapse, or gibberish.
As Mike Price of the MWI FAQ liked to say, “Bohm+Ockham=Everett”.
No. You think there’s no real difference between one world being real and all worlds being real? Lots of alternate Rokos disagree.
Thinking that the only thing that’s real is the experiences you anticipate, and all talk of where these experiences come from is “poetry”, is a very odd kind of solipsism.
I don’t understand what the word “real” means here. This may be a case where your mind contains a attribute IsReal which has no counterpart in the world.
Again, the word “real” is causing trouble here. What does it mean for (e.g.) me to think Steven is “real” vs. Steven is “not real”, subject to the condition that I have identical experience anticipations? “real” seems to be a null word—it adds nothing.
Wikipedia on Solipsism:
So, based on this I don’t think I’m adopting a solipsistic position. My knowledge about the world is contained in the mathematical laws of quantum mechanics. “interpretations” of quantum mechanics aren’t knowledge because they have no predictive power.
One day, a traveler came to Joshu’s monastery bearing an ancient map. The map showed verdant forests, majestic mountains, winding rivers. And Joshu said, “do we not live in a wondrous world, that has such terrain features in it?”
The traveler frowned. “This we do not know! All we can say is that when we travel to these places, it looks as if there are forests, it feels as if there are mountains, it sounds as if there are rivers. Geography deals not with the land, but with what we can say about the land. The rest is poetry, religion, metaphysics.”
Joshu went to the other room to fetch a scroll, and inscribed on it a crude schematic representation of the map the traveler brought. He then wound the scroll up and used it to beat the traveler over the head. At that moment, the traveler was enlightened.
Not content with the emptiness of mere talk,
he prefers to talk about talking.
Does he not know that to take one step back
is to slide into the abyss?
Strictly speaking, knowledge doesn’t need to have predictive power, as your utility may depend on a piece of knowledge, in which case you don’t expect to observe anything else differently, but you prefer to act differently. That knowledge fixed in a belief still pays rent, but not in expectation.
Survey completed. I’ve just checked the answer to the calibration question, and I’m glad I gave myself a low confidence score...
The aliens question was interesting to think about.
I realized that if I put anything other than zero for ‘probability of aliens existing within our galaxy’, then it seems like it would make little sense to put anything other than 100 for ‘observable universe’, given how many galaxies there are! Unless our galaxy is somehow special…
The probabilities for life in different galaxies aren’t independent if you’re uncertain about the fraction of stars/galaxies that spawn life. You could think that with p=.5 life is super-rare and with p=.5 life is super-common; then you’d put .5 on both questions.
Yeah. Checking back over questions can be interesting. At first, I had a greater probability for cryonics working than anti-agathics, but then I realized—hold on, if cryonics worked, wouldn’t aging have more or less been solved? Isn’t cryonics a subset of anti-agathics?
I took the survey, and am not all that interested in the Karma point… I just wanted to brag that I wasn’t far off on the calibration. I gave a pretty low confidence level, however.
...though in thinking about it now, my confidence distribution is spread pretty far towards the earlier side of things. I would not have been surprised to find that my guess was half or an entire century later than it actually was, but I would have been extremely surprised if my guess was 30 years earlier than the actual date.
Maybe that’s not too surprising, that I feel more confident estimating more recent historical events than those farther back.
I took the survey, very interesting questions. I guess I should be happy my answer to the calibration question was within his lifetime...
I didn’t check the section boxes.
The natural log of my probability of the correct answer (is there a term for that? LN score?) on the calibration question was −0.073. How did you guys do?
By the way, your LN score on this post is equal to the natural log of your probability of the fact that I took the survey.