Posts
Comments
Really neat paper, thank you!
This seems like a great question to me and I'm bummed I can't answer it. But here's a toy model that might help a bit.
Take a 2-dimensional spacetime shaped like the surface of a vertical cylinder, with space being the 1-dimensional equatorial circles, and time going vertically. Some of the straight lines in this space are slanted lines just going around and around the cylinder forever, and objects following those as world lines would sort of appear to oscillate around a point traveling along an exact vertical world line.
Anyway that model's only 2-dimensional, and the bigger problem is it's not the right type of geometry (it's Riemannian not Lorentzian). Also the cylinder is flat, not curved. But maybe it still helps.
For what's actually ideal, I would suggest (if you find it interesting) reading about technical clothing for mountaineering and winter camping and adapting that to city fashion -- but if you want some helpful more affordable tips, what works for me is many layers.
For example, long sleeve undershirt and long underwear from Walmart. T-shirt over the undershirt, cheap sweatshirt hoodie over that. Thin pajama style pants over the long underwear. For me anyway, this can be completely comfortable under an outer layer of only jeans and just an autumn jacket up to maybe -15C or -20C.
Thin cotton gloves under big mitts or heavier gloves. Thin socks under heavy socks. For items where it's more difficult to layer, such as a toque or scarf (or socks if shoes limit the room), wool is pretty good and is affordable, thick, and sturdy at Army Surplus stores, at least here in Canada.
I would also suggest, don't neglect extremities or any body parts. For example, I remember once my thighs being very cold wearing only jeans (no long underwear), even though I had an extremely warm parka and was walking briskly.
I've found personally that even fairly thin pajama type pants, if you have say two layers under jeans, can keep you pretty comfortable even up to maybe -30C. Since you have in mind a meetup, I feel that even for example students on a budget, could get extremely adequate winter clothing this way at a Walmart in Canada that would enable them to stand outside inactively for 3-4 hours at up to -25C, say.
Thanks! FWIW your high opinion of the project counts for a lot with me; I will allocate more attention to it and seriously consider donating.
Definitely interested, especially if there are more details you can give about the options bets you mentioned in this comment.
I'm only vaguely informed about options with no experience buying them, so I likely need to learn more before I can ask useful questions about this, but I don't want to miss the opportunity to say I'd definitely be interested in any thoughts you might be considering posting, since I'm interested in this type of general bet on AI progress.
Anyway, I know you can't give investment advice and don't expect you to teach me how options work, so I guess I'm just hoping for details that help figure out if I'm thinking about this correctly, specifically with respect to using options to bet on certain short AI timeline scenarios.
(For concreteness, my general understanding is something like: say I decide the bet to make is on GOOG stock. So I set aside $X per month to buy "long call" options on GOOG as you indicate in this comment. And I should be prepared to typically lose all of that $X each month, unless the bet pays off.)
I happened upon this website once previously and couldn't quickly come to an assessment of the project, before moving on. I assume from your comment you feel this is a worthwhile project? I'd be interested to hear your take on it.
I think The Economic Consequences of Noise Traders is one of those papers.
In the book Mere Thermodynamics by Don S. Lemons published in 2009, I was quite surprised to read: "While the caloric theory of heat is plausible and to this day remains useful in limited circumstances ..."
This is the only book on thermodynamics I've ever read, so I can't really elaborate on those limited circumstances, unfortunately.
I wonder if I'm understanding this correctly: is something like sweating an example of refrigeration, since it keeps the low temperature thing from heating up? And heat pumps are different, they keep a hot thing from cooling down, but otherwise the underlying thermodynamic processes are similar?
If I've got that straight, is any evaporative cooling an example of refrigeration, but the question here is specifically wondering about heat pumps not refrigerators?
On occasion I've reflected about whether I care about a beautiful statue no one would ever see. Let's say it's in a hole somewhere. I don't know whether I care or not, but it seems possible to care.
You can modify the question a bit. What about a statue not just that no one would ever see, but that no one would ever know about. Or what about a statue that's impossible to ever even know about, like in a separate universe or something.
I've got nothing concrete to add about that, just that your point 1. made me think of these questions I've asked myself.
I agree that the type of action needed to address the parenteral nutrition example would be very unlikely to arise from a hypothetical online crowd action platform.
I think much of the purpose here is to brainstorm what the most effective version of such a platform could look like, and thereby also get a picture about what class of problems such a platform might be able to address or help with.
Is your position that that class of problems is small, and/or those problems are better addressed by existing solutions anyway? Maybe that's right. We are largely brainstorming here to gauge whether a marginal improvement is possible (and probably hoping that a big improvement is).
Actually Raemon asked specifically about such possible negative outcomes here. I only point it out because I think you make a valuable point and I'd be interested to see any further discussions of it.
Interesting, what's the connection here? Are you saying that assurance contracts could be implemented using a prediction market? Or were you thinking of something more along the lines of Augur's distributed voting system for decision outcomes? Anyway I'd be interested to hear what you're thinking here.
Please do!
Is CollAction kind of what you had in mind? Wikipedia also mentions PledgeBank which closed down in 2015 but still has a site with all the archived pledges at www.pledgebank.com. Not sure but I don't think either of these sites allow versions of assurance contracts that use financial contributions.
I wonder what a good binding mechanism might be for crowdaction sites in general, because once the quorum is reached you still have to get everybody to go ahead and perform the action. Any thoughts?
A site based on dominant assurance contracts would also be interesting. It might be a bit more like Kickstarter, in that maybe the entrepreneur initiating the DAC can market and popularize the action, and be responsible for the action happening when/if the quorum is reached.
Does UDASSA include concepts like the "observer graph" and "graph machine" that Muller describes in the paper? Is Muller just filling out details that are inevitable once you have the core UDASSA concept?
Yes, that's absolutely an aspect of it.
I'm pretty agnostic about what form a solution to the original needing-to-have-a-job-to-support-myself problem might take. The hourly breakdown thing was just an attempt to ballpark some money amounts, with an emphasis on how high the value might potentially be for me. I suppose other categories of solution along the lines of tiny houses or solar panels or whatever that might ease transition to lower cost of living lifestyles would have very different prices or payment structures probably ending up a lot lower than the scenario I first outlined.
I'm not actually in option A. in real life, sorry if I gave that impression. It's just an attempt to illustrate how much I might be willing to pay, if hypothetically restricted to those two options. Not suggesting restriction to those two options is realistic or anything, nor do I have any mechanism in mind that might lead to those options.
I do know.
Same problem for me. And I agree, seems like it would take a lot of ingenuity to turn a solution to this problem into a viable business. Maybe products aiding rat-race opt-out strategies, coordination tech enabling 20 hour work weeks.... I don't really have any specific ideas.
But to answer the second question, I feel like I would pay a great deal for something like this. Hard to quantify since solutions offered might be partial, and a total solution might eliminate income, etc.; but just to illustrate (and ignoring things like extreme saving for early retirement, etc.) between the options A. and B. ---
- A. $80,000/year at 40 hours a week
- B. third party company L somehow enables me to earn $40,000/year at 20 hours a week in the same job with the same employer as above, but company L takes a 25% cut of my pay (so $10,000 a year to company L)
--- I'd gratefully choose option B. (which is just to give a sense of my intuitive preference, not saying I'd defend the merits of that choice necessarily).
Interesting idea. So I guess that approach is focused on measure across universes with physics similar to ours? I wonder what fraction of simulations have physics similar to one level up. Presumably ancestor simulations would.
It seems possible to me that after passing some threshold of metaphysics insight, beings in basement reality would come to believe that basement reality simulations have high measure.
Past a certain point, maybe original basement reality beings actually believe they are simulated. Then accurately simulated basement reality beings would mean simulating beings who think (correctly) that they are in a simulation.
I don't know how to balance such possibilities to figure out what's likely.
Eliezer ruminates on foundations and wrestles with the difficulties quite a bit in the Metaethics sequence, for example:
Michael Vassar makes some observations about this in this chat from about 37:50-40:30. He begins describing something called a "hexayurt tridome", some kind of portable desert structure, and finishes saying "for the cost of engineering the 2016 Toyota Corolla and with the level of engineering skill required to engineer the 2016 Toyota Corolla it would probably be possible to engineer a house that would cost less than a Toyota Corolla and that could be deployed more easily and be adequate for any climate pretty much anywhere in the world where there's a reasonable amount of free space".
Thanks a bunch that is the one!
I have a foggy memory of someone here (I think it was gwern) linking to an article about simulation interface design. It built up examples based on a bird's eye view of a car steering down a road. I haven't been able to find it, anyone know a link to the article?
Looks to me like Halmos does intend "one-to-one" to mean "injective". What he writes is "A function that always maps distinct elements onto distinct elements is called one-to-one (usually a one-to-one correspondence)." Then he mentions inclusion maps as examples of one-to-one functions.
Biotest is very reputable (that's my impression anyway) for supplements geared towards weightlifters.
https://www.t-nation.com/store
They have stuff in their "Health" category with broader appeal (Superfood is pretty cool). I've been very pleased with what I've used from them.
Caffeine's a strong drug for me, except I have a huge tolerance now because I consume so much coffee. One night a few years ago, after I had quit caffeine for about a month, I was picking away at a bag of chocolate almonds while doing homework, and after a few hours I noticed that I felt pretty much euphoric. So yeah, this is good info to have if you're trying to get off caffeine.
This sounds familiar to me. I'm 32 and I definitely remember hearing stuff like this. I remember in elementary school (so, late 80s early 90s) seeing the Canada food guide recommend a male adult eat something like up to 10 servings of grains a day, which could be bread or pasta or cereal. You were supposed to have some dairy products each day, maybe 2-4. And maybe 1-3 servings from Meat & Alternates.
I remember that pretty much all fat was viewed (popularly) with caution, at least until Udo Erasmus came out with his book Good Fat, Bad Fat.
But I do recall a clear message that soda and snacks were unhealthy. It wasn't as though soda was thought ok just because it was low fat / high carb.
Is the insight about free will and logical facts part of the sequences? or is it something you or others discuss in a post somewhere? I'd like to learn about it, but my searches failed.
I have a low attention span but I read through your entire document and when I reached the end I was surprised because I had the impression I was still reading the preliminary part. So, for what it's worth, I found it easy to get through.
I recommend articles by Chad Waterbury and Dan John to start, which you can find on www.t-nation.com (which is a quality resource in general). Dan John tends to recommend simple, effective strategies, and Waterbury writes a lot addressing your first and second questions.
A great place to start, this article by Mark Rippetoe: http://www.t-nation.com/free_online_article/most_recent/most_lifters_are_still_beginners
Eric Cressey and Christian Thibaudeau write great stuff, if you're into a more complicated approach.
Also I agree strongly with Zed's point 2.
I agree with Zed's point 3. about supplements sort of. I think the exceptions are (1) "peri-workout nutrition" about which there are many complex strategies but you can keep it simple, say a whey protein & gatorade shake before or after (or both) your workouts, and (2) things often lacking in a modern diet or lifestyle: fish oil (or some other way to get omega 3) and vitamin D seem to be the most frequent recommendations.
Thanks for this. I guess this goes to show how hard it can be to communicate math well. When I learned the sin(x)/x limit I accepted the "proof" by geometric intuition with no protest and was not alert to any deeper source of confusion here.
Come to think of it, the rigorous treatments of sine that I've seen probably all use power series definitions. To see that it's the same function as the one defined using triangles I expect you have to appeal to derivative properties, so that approach would not skirt the issue.
I agree the problem is even more pronounced in physics.
Also, I am interested in and would appreciate the details of the case study to which you refer.
I also find that scary/frustrating. But don't you find you can relearn those forgotten chunks much more rapidly than the first time, if you need to?
My own experience is that it is fairly easy to identify points of confusion, and the hard part is finding a book or whatever, at the right level, to address that specific point. This is a tough problem to solve with self-teaching.
If this meetup happens, I'd drop in.
I can tell you a reason I've cryocrastinated. I don't expect the reason to hold up under scrutiny. It held up under my half-hearted scrutiny, but so what? I have low confidence in my own ability to be rational. In fact, I'd be grateful if someone can eliminate this worry for me.
So, the reason is concern about dystopian or hellish scenarios. For a cartoon of such a scenario, think I Have No Mouth, and I Must Scream.
One thought I had was that these scenarios are so unlikely that if I felt they warranted avoiding cryonics, I'd also feel they warranted preventative suicide. I'm confused about where I stand on this, but in any case preventative suicide is an action I am absolutely unable to take, whereas not signing up for cryonics is easy.
Thanks if you can straighten this out for me.