It also seems like you are not actually utility indifferent as between a 90% chance of $1b and a 9% chance of $9b—the former seems far more valuable to me because once you have about ~$10m, the rest is just points on a scoreboard. So to the extent that you are more emotionally comfortable with 90%/$1b I think it’s actually because the expected utility of 90%/$1b is almost 10 times higher than the expected utility of 9%/$9b. And so to the extent you are setting your motivation based on these 2 things, you are importantly fooling yourself here.
TBH, for the equation “Util(90% chance of $1b) = Util(9% chance of $X)”, I don’t think there is any finite X that can solve that equation.
I don’t love gas but one point that hasn’t been brought up explicitly is that gas is “dispatchable”. Since the grid needs to be exactly balanced at all times, you need something that can ramp up and down easily. Obviously storage can smooth this out too but gas is often seen as a complement to solar and wind because you can turn the “gas → electricity” dial up and down pretty much at will. this is much harder to do with coal and nuclear (although nuclear is also carbon-free in it’s own right).
Found the movie hilarious and it never crossed my mind that the protagonist being a middle aged immigrant was part of what made it funny, btw.
Yes, I agree here, and I wonder how OP ended up with this impression. Her being a middle-aged immigrant is important to the family story the movie is telling because her relationships with her dad, husband, and daughter, and who each of them is, are all affected by this fact and by the protagonist’s experiences and choices. Those relationships both drive the plot and form the emotional core of the movie.
I think if you just view this movie thru the lens of one genre, it’s going to seem lacking because, what genre is this movie? It’s everything, everywhere, all at once.
A single-word change that you can observe in the wild is through → thru. It appears that way on road signs and I’ve started to see it creep into informal and even semi-formal writing, e.g. internal-only business email. Unfortunately, the related change throughout → thruout just looks bizarre, though I guess the original is kinda a bizarre word too.
it’s not that hard so long as you have an instant-read thermometer. I have done it following instructions online and had a very good success rate. If you mess it up, it’s also possible to just re-melt and try again.
I am a lawyer, and in legal writing we are taught the concept of “old to new waterfall.” This means, essentially, that every sentence, every paragraph, and the document as a whole, should start by reminding the reader of something they already know, and then proceed to new information that’s related. The important bit is that, once some “new” information is introduced, it becomes “old” information in-scope.
So you can build up complex points/arguments by building old-> new, old → new, over and over. You can see this even at the sentence level. Here’s an example:
The moon is made of green cheese. The cheese smells very bad, and the bad smell wafts through nearby space. The smell repels aliens from attacking Earth.
The above is an example of how to do this. The below is an example of how not to do this. Note that they both have the same semantic content, but the first “flows” better and is thus easier to read.
The moon is made of green cheese. Nearby space is filled with bad smell coming from the cheese. Aliens don’t attack Earth, because they are repelled by the smell.
How much of this value could be equally accomplished by a pop-up that, rather than trying to “translate” the message, reminds the user that the other person has feelings and needs that they may be having trouble expressing, and that the user should consider that, without any attempt for the AI to state its best guess of what those feelings and needs are? Because I think that alternative would address substantially all of my objections about this while possible preserving nearly all of the value.
Yes, but if you apply this concept, you still won’t be hearing their feelings and needs. You will hear some function F(feelings, original message) = output. Likely one of the following:
(A) GPT on sender’s side, guesses 90% right: You might hear that 90% and lose the 10% that GPT did not guess
(B) GPT on sender’s side, guesses 80% right: You might hear mostly the 20% because the sender, consciously or unconsciously, alters their original message after seeing GPT’s mostly-right-but-still-wrong guess to emphasize the distance between their true feelings and the guess
(C) GPT on sender’s side, guesses 80% right: Sender might tweak the output so that its close to GPT’s output but not exactly—maybe this 100% captures their feelings and needs or maybe 10% is still lost trying to wedge their amorphous vibe into GPT’s format
(D) GPT on recipient’s side: You see the original message and now GPT is totally guessing at what the other person might be feeling—if its wrong you will have no human feedback in the moment to correct it so its output is at best useless and at worst will lead you to form overconfidently wrong beliefs about the sender’s emotional state.
(E) GPT nowhere, but app is popular in your community: You see an NVCish message from someone and now have to guess at whether their message is an original message, or the output of F(feelings, original message)
Note also that a recipient definitely has no way to know if they are in world (A), world (B), world (C), and depending on your relationship might also be in world (E).
I actually think this idea is not good, and would oppose its widespread adoption. For level-setting, I also want to say that I upvoted the post as I think it’s an idea worth thinking about.
I have noticed in the last few years, many services have been rolling out more and more features like this already—not so much “anger translators” but other sorts of suggested replies, etc. My problem with these is that it feels very intrusive, the computers intruding sometimes into our most intimate lives. For example, GMail’s suggested replies and autocompletes show up on all messages if not disabled. I would be fine with this for pro forma interactions, such as customer service reps or whatever, but they also show up when I’m conversing with family and close friends.
The problem is that it feels impossible to just completely ignore the suggestions—either I want to accept one or else I feel myself changing my intended reply intentionally to avoid one. Either way, the system is changing how I would communicate with my loved ones in a way that I really dislike. It reminds me one of the most memorable lines from the Matrix, (from memory, not verbatim) “1999, the peak of your civilization. Of course, I say your civilization, because as soon as we [AIs] started thinking for you, it really became our civilization.” To the extent that AI can automate more and more work that seems great to me, but automating interactions with our friends and family sounds like dystopia.
One thing you might have missed, regarding reasons why so many people support a no-fly zone, is that a lot of people just aren’t thinking through what that actually means at all. Like, maybe you are so used to thinking about what your words mean that the inferential distance here is too great, but I think the thought process for probably a majority of poll respondents is something like:
All-out open war with Russia — that sounds scaryDoing nothing — that seems bad too
Aha, I hear about this “no-fly zone” and it sounds like kinda middle-groundy between those two things, so I guess I’m in favor of that! I think if you were to taboo “no-fly zone” and replace it with “US aircraft shoot down Russian planes, also US bombs Russian anti-aircraft emplacements in Russia, also US Navy ships sink Russian AA destroyers, etc etc” then you would see polled support plummet. So I don’t think that seeing high poll numbers for “no-fly” is really any evidence of the kind of nihilism you mention, just imprecise thinking. One piece of evidence for this is that if climate doomerism was responsible, you should expect to see the most left-wing/green new deal members of Congress pushing the no-fly zone, and you do not see this.
The unclear thinking is no excuse for the leaders who are pushing this line though, who should know better. I expect they are just playing politics, dishonestly pushing this knowing Biden will have to say no to it, so that they can accuse Biden of being weak.
Re: Augury, I feel like there’s two ways for this to close though. One is, like you said, trucking the birds from their climate-change-induced altered ranges back to their historical ranges. But, it seems much easier to alter the records of the birds’ historical ranges, a la “the birds have always been migrating over Eastasia.”
So, generally, this ROW concept needs to take account that often this sort of evidence relies on comparing two unlike phenomena, e.g. physically observed bird migrations, with historical records of bird migrations. For the ROW concept to work in practice you need to take account of the ROW on both sides of the “equation.”
I feel like legal ethics has actually come up with a reasonable policy here. Check out ABA Model Rule 1.6, governing Attorney-Client confidentiality: https://www.americanbar.org/groups/professional_responsibility/publications/model_rules_of_professional_conduct/rule_1_6_confidentiality_of_information/
Obviously one would need to analogize to the situation of a friendship, but, for example, see Rule 1.6(b)(2):
A [confidant] may reveal information relating to the [confidential conversation] to the extent the [confidant] reasonably believes necessary:
(b) to prevent the [confiding party] from committing a crime or fraud that is reasonably certain to result in substantial injury to the financial interests or property of another and in furtherance of which the [confiding party] has used or is using the [confidant]’s services
The line they draw there is that if the client/confider is, by confiding the information, involving the attorney/confidant in the misconduct or using the attorney/confidant to help perpetrate the misconduct, then the attorney/confidant is not obligated to keep the information confidential.
But it seems like a lot of this political organizing (or “collusion”) occurs through the channel of convincing people to start caring about your issue. This is a good thing! The Teamsters in the 1960s really put themselves on the line for Civil Rights, I don’t think it’s correct to say that they did it for purely Machiavellian reasons. At least now, the Teamsters seem really proud of their history on this and have a whole section of their website devoted to it.
Actually paying people to vote your way would be illegal of course. But it’s illegal now, and not widespread in the United States AFAIK, because the risk/benefit on it is so bad. To expand on this a bit, to the extent this is what you’re referring to as “collusion”, in a polity with any significant number of voters, you’d have to buy a lot of votes to influence the outcome (bare minimum thousands, most of the time). But every single person that you approach with an offer to buy their vote is a risk to turn you in, and you face serious felony charges if caught. Totally not worth it.
If what you really mean is using policy to “buy” the votes, then we’re back to this seeming good. The outcome is that a lot of people get a policy that they like, and they use their votes in a way that seems likely to them to get that. Again, seems good.
To the extent that you are using actual dollars to “vote on” (really buy) policy outcomes, I guess you have a lot of other problems with the system. To the extent you are using “voting tokens” or something as described in Vitalik’s post, then the proper, virtuous, and pro-democratic strategy is to convince more people to spend some voting tokens on your issue. Then you end up with an outcome broadly acceptable to many, which is what you are supposed to get in a democracy. Of course, it’s not that clear to me that this in practice works out all that differently than regular democracy, since convincing the mass public to support you is much more effective than spending a lot of tokens yourself.
Isn’t “collusion” here just another way of describing political organizing? Like, different interest groups which each care only about their own issues get together and decide to support each other’s issues so that they can win power and each accomplish their agendas.
This seems....at the very least not bad, and probably actively good.
A good historical example of this would be the Teamsters joining in coalition with the MLK-era civil rights movement. Are you saying this is bad collusion? Or is this good solidarity and organizing? How would one tell the difference?
The “problem” you are describing here is people making political choices to join coalitions, and exists in normal democracy just as much as it exists in QV.
I don’t see how. Maybe you can hoard without consequence but you also remove a major reason to hoard. The two reasons to hoard are either for personal consumption, or for resale.
For personal consumption: For people inclined to hoard it’s hard to see how “I’d better buy it now because they might be out later” is likely to induce a lot more hoarding than “I’d better buy it now because it might only be available for an unaffordable price later.”
For resale: This is the kicker and this is Jiro’s point—people get the idea “I’ll buy TP now before those rubes, then when its out everywhere I can resell for higher.” If its illegal to resell for higher then this excess TP is not hoarded because there’s no incentive to do so—people know approximately how much TP they actually need for personal use and there wouldn’t be much incentive to buy more. This is even more pronounced for durable goods like snow shovels, you only can use one shovel at a time but you could imagine yourself able to re-sell an unbounded number of snow shovels.
I think the gods-eye optimal solution would be something like allowance of price increases for people who are bringing goods in from outside and a ban on speculation. But in the real world it’s really hard to differentiate speculation from “legitimate” buying and selling, and, maybe it’s wrong or maybe it’s right, but it’s definitely not crazy to think that the net effect of second-order speculation induced shortages is worse than the net effect of first-order disaster induced shortages.
I find all these sorts of analyses deeply frustrating because they seem to take place in spherical-cow world and never to take account of the actual experience of virtually everyone who participates in the labor market—you can’t just choose how many of your hours you convert into dollars at an arbitrary level!
For hourly employees, one’s employer usually determines the schedule with limited input from workers.
For salaried employees, the marginal hour worked results in essentially zero wage gain. It may result in being marginally more likely to receive a bonus or promotion down the line but these effects are highly uncertain and may have all sorts of slopes and plateaus in the hours/$$ curve.
For self-employed, if you do anything other than daytrade, then you likely have to interact with suppliers, customers, etc, and there’s often pretty legit exogenous limits on your ability to translate hours in to $$.
Additionally, the experience benefits seem dubious to me for a similar reason, which is that to the extent you are selling your labor in any capacity (vast majority of workers), the potential buyers of your labor often have limited view into your experience or productivity, and to the extent they estimate these based on your experience, it’s usually going to be measured in years not in hours.
So its very unclear to me the mechanism by which most people could decide to turn their time into money at anything approaching something that looks like a marginal wage, and additionally unclear to me how, even if so, working harder now will actually increase the returns 20 years from now in the way this post proposes.
Not sure—treatment for Lyme (if caught early) is just antibiotics, and there might not be a Lyme-specific office visit code. So it might just show up in billing records as:
1 Office Visit (these are coded with levels of complexity but not with specific patient complaints/diagnoses IIRC).
0-2 days later, 1 prescription fill for a standard course of some generic antibiotic.
As someone who both lives in Massachusetts and has had Lyme, I can provide some context. When I had the classic bullseye rash pattern associated with Lyme, I called my doctor and asked about treatment. The doctor raised the possibility of getting a Lyme test, but told me that the Lyme test had a high enough false negative rate that even if I tested negative, she would still recommend I be treated for Lyme, and said that on that basis I didn’t really need to do the test. I have not independently researched the accuracy of the Lyme test, but here’s my understanding of the considerations based on what she said:
-The cost, both monetary and otherwise, of being treated for suspected Lyme is low
-The likelihood of a false negative is relatively high
-Untreated Lyme can be quite serious and can cause permanent damage if not treated promptly
-The test also takes time and hassle to obtain
Given these considerations, I (quite sensibly IMO) decided not to take the test.
It seems like the most straightforward simple fix is for everyone to guess simultaneously. Then people can’t just grab big ranges because they are big. E.g. the player who guesses 49 in your example would have to have actually had a lower estimate for the value rather than merely cutting you off at the knees. I haven’t studied in depth but I think the incentives (if each person is trying to account what the others, N-layers deep) are to just guess your true prediction.
Thanks for writing these! I often enjoy reading your game reviews. One thing I think that would be low-effort and really helpful to the reader would be to copy/paste or link to a succinct definition of the “tiers” that you always reference at the beginning of each review.