jimrandomh

[jimrandomh]

The first thing to realize is that some of the recipients of spending are automated systems which are selling a different form of money in return, eg another cryptocurrency, or USD. The second thing to realize is that "double spend" does not mean "Scott sends a Bitcoin to Alice, and also to Bob", it means "Scott sends a Bitcoin to Alice, and also to Scott-wearing-a-disguise". So if Alice rejects the transaction, then either the system declares the coins permanently unspendable, or Scott gets to try again. The third thing to realize is that if someone had the ability to make coins permanently unspendable for too long after those coins were supposed to have left their possession, that would be a problem. The fourth thing to realize is that attackers can create temporary communication disruptions pretty much whenever and wherever they want, and it's so easy kids are using this capability to win online games by kicking out their opponents. So an attacker could create a partition in the network, send coins to a target in one partition and to themself in another partition, and hope the target pays out USD or another cryptocurrency before the partitions reconnect. And if they don't, the attacker can just try again with the same coins.

Proof-of-work solves this by making it really obvious to potential targets whether their communications have been partitioned off: you're either getting proof-of-work from the rest of the network, or you aren't. Proof-of-stake does the same thing: you're either getting proof that you're in contact with the owners of the most of the coins, or you aren't.

God help me I’m trying to figure out how Bitcoin works.

Here’s where I’m stuck.

Imagine something much simpler than Bitcoin. Each node just maintains a spreadsheet of who has how many Bitcoins. If Alice wants to send a Bitcoin, she sends a cryptographically signed message saying “please transfer one Bitcoin from Alice to Bob”, and then everyone does this. What goes wrong?

I think the answer is something like: if one of the people with the spreadsheet dishonestly changes the numbers, then there’s no good way to know which spreadsheet is right. So you include a ledger. You have a record of every transaction, cryptographically signed, and then if I try to dishonestly pass off a spreadsheet that says you gave me money when you didn’t, you can point out that there’s no record of that transaction, signed by you.

Okay, fine, so spreadsheet with a really long ledger at the bottom. Now what goes wrong?

I keep reading people saying “double-spending”, but I don’t get how this works. If I try to spend a bitcoin I don’t have - if I send the message “Scott sends ten bitcoins to Bob” when in fact I have zero bitcoins - presumably someone maintaining the spreadsheet sends Bob a message saying “no, we checked and he can’t do this”. But suppose I do have ten Bitcoins. I tell Yolanda “Scott sends ten bitcoins to Alice” and Zachary “Scott sends ten bitcoins to Bob”, and potentially Yolanda could tell Alice “yeah, this checks out” and Zack could tell Bob the same thing. But my understanding is that the solution to this is to wait a few seconds before confirming transactions, during which time Yolanda and Zack can talk to each other, realize something is wrong, and take some action - either cancelling both transactions, or using whichever one has the earlier timestamp.

My understanding is that actually, for some reason, this doesn’t work, and you need mining and proof-of-work. What is that reason, and why do mining and proof-of-work solve it?

[jimrandomh]

It's a way of balancing visibility between nested and top-level comments. If they start out fully expanded, most people will read the first top-level comment, then read a bunch of replies to that without ever getting to the second top-level comment, which probably has more karma and is better.

On LessWrong, Cmd+F (or Ctrl+F) expands every collapsed thing on the page, so you can search and also so you can use it as a hotkey for getting rid of collapsing if you don't like it.

Does anyone actually like comments sections that work like this?:

Are there people out there who get angry when a page displays all the comments at once? Who smash their keyboards, saying “No! I want to have to click ‘see more comments’ before I can see any comments other than the top two! I want to keep having to do this for every two or three more comments I want to see! And it shouldn’t show me any replies to comments until I press 'see replies’ for each individual comment I want to see replies to!“

It seems weird that every website in the world would switch to a commenting style nobody likes, but I also can’t imagine anyone liking this, so it’s just a huge mystery to me.

[jimrandomh]

Different PDF readers will do a better or worse job at this; Chrome's built in default PDF reader seems to do a better job than what you showed here. It's one part OCR problem, one part dealing with the corner cases of the worst file format design ever created. In a summer job in 2003, I worked with the PDF file format professionally, and I consider the PDF format's widespread adoption to have been a minor crime against humanity.

When I try to copy-paste from PDFs, the text usually ends up really mangled. For example, https://sci-hub.tw/10.1017/s0264180100000412 becomes

A  number   of  studies  with  human   subjectsreport  that  despite  the  accumulation  of  drug  inthe  plasma  and  higher  actual  drug  levels,  initialdrowsiness decreases after  a few days of repeatedtreatment  with  benzodiazepines  (see e.g.  Scharf& Jacoby,  1982; Hillestad etal.  1974; Greenblatt&  Shader,  1978;  Greenblatt  et  al.  1979;  Bondet al. 1983; Tedeschi et al. 1985). Tolerance to theanticonvulsant  activity  of  oral  benzodiazepinesdevelops  so rapidly  that  they  are  regarded  as oflittle  benefit  in prophylaxis  (Anonymous,  1986).Psychophysiological   measures  of  the  effects   ofbenzodiazepines    also   decrease   with   chronicadministration.  Lader et al. (1980) demonstratedtolerance   to   clorazepate   in   auditory   evokedresponse (AER)  amplitude and  latency changes,as  well  as to  the  fast-wave  increase  on  the  EEG(see  also  Bond  et  al.  1983; Oswald  et  al.  1982;Hoehn-Saric   &  MacLeod,   1986;  Kales  et  al.1978,  1979)

…ie half the words have too many spaces and the other half have none at all.

Does anyone know an easy solution to this?

[jimrandomh]

You probably have an HTMLCollection, which live-updates when the DOM changes, when you thought you had an array.

what the fuck what the fuck what the fuck

So I’m parsing the tumblr DOM right?

And the only change - THE ONLY CHANGE - I make is logging the children of a particular node.

Doing this somehow… completely changes the list I’m iterating over?

[jimrandomh]

The US trade deficit is about equal to growth in the M2 money supply. So, it's coming out of complex banking rules that de-facto print money.

Econ question: the US has a trade deficit, right? Where does the extra money come from? Presumably some of it is borrowed? Is the rest just that fractional lending or whatever produces enough wealth to make up for it?

@dataandphilosophy @rusalkii @other econ people

[jimrandomh]

You're a cuck. You're welcome.

Don’t feed the trolls. I feel like this is a bit of lost wisdom. It’s somewhat difficult to talk about specifics because of the virtue of silence, but I’ll try to discuss it in broad strokes. If someone is deliberately trying to hurt you, telling them that they are hurting you will not make them stop. It will egg them on.

If you get a nasty anon, delete it. If you get vicious reblogs, block the perpetrators. If someone sets up a community to talk about how evil your community is, at the minimum, don’t link to them and don’t mention them in public by their true name. If you’re dealing with a stalker, the recommendation of experts on the matter is to not respond to any of their emails or letters, but to quietly document them. 

[jimrandomh]

No, I think the difference between a restricted hero pool and the full hero pool says something very important about what they (haven't) achieved: it means for their AI to learn an additional hero requires either hero-specific programmer effort, or superlinear computation. Which means it's not very general.

The latest incarnation of OpenAI Five is going to play against the human team OG (the world champions in Dota 2), today at 18:30 UTC. Place your bets :)

[jimrandomh]

Actually, the problem is more fundamental. If you want to parse your XML into a treelike structure, you need pointers and memory allocations. A pointer is 8 bytes, and you need a lot of them: for the tag name, for each attribute key, for each attribute value, for the sibling pointers... and that's before you take into account the overhead of the allocated objects themselves, which is also substantial.

Why does XML parsing use so much ram? I’m reading in an 8GB file and the parsed tree is blowing up at least 5 or 6x.

I’m doing an iterparse now and it works, but I’m still confused.

[jimrandomh]

When people have against-consensus views, they tend to split pretty neatly into two further categories: people who have reasonably-nuanced contrarian positions, and... people who didn't read or didn't understand the arguments that led to the consensus forming in the first place. Distinguishing these is actually pretty easy, and the latter category is much more common. In practice, my social experience has been that nuanced-contrarians do well, and it's the didn't-get-it-enough-to-realize-they're-being-contrarians who get filtered out.

Supply/Demand Blues

I want to trace out a general phenomenon that you’ve probably thought about before, and then a specific subspecies of it that you might not have.

Then, spoilers, I’ll admit that I have worries specifically about the effective altruism community, and a (plausibly moderate) threat to its ideological representativeness.

General phenomenon: When there’s a ridiculous oversupply of some kind of labor, the effective price people will pay for that labor gets really low.

Example: Approximately a zillion people want to write fiction short stories for some reason. Therefore, the majority of places that publish short stories don’t pay for them at all. Places that do pay for them are extremely competitive, and random noise/reviewer taste plays at least some role, such that consistent success is almost impossible.

Unfortunately, sometimes instead of the price getting low, it bottoms out at some floor and is offset by horrible externalities.

Keep reading

[jimrandomh]

Nope! The source code to FAI might get burnt or otherwise fail to be executed, if you use it as a pizza topping. The best pizza topping is a colection of heat-resistant FAI-constructed nanomachines.

“Pepperoni is the best pizza topping” is, like, the most boring take imaginable. But smallish slices of pepperoni which have curled upward while baking to form little cups, with a little bit of grease pooling in the bottom of the cups, and some char on the outer rims is, as it happens, the best pizza topping.

[jimrandomh]

According to Quora, fire ladders typically reach ten stories, though rare specialized ones may go as high as 40. So on the eighth floor, a window is definitely relevant. (Also a window can be used to vent a room after you've sealed the door to a smoke-filled hallway, which is important.)

"the only acceptable housing regulations are safety regulations" you realize this is why bedrooms with no windows are illegal right? so you can get out I'd there's a fire?

I asserted that this was the reason for the window regulations the last time I was talking about the SF housing crisis but a bunch of people correctly pointed out that the window in a 20th floor apartment, or even an eighth floor apartment, is zero protection in a fire. 

I am in favor of safety regulations which actually make us safer, but on the 20th floor of an apartment building, that’s ‘the sprinklers are well-installed, regularly tested and up to date’ and ‘the emergency stairwell is correctly designed’ and maybe “each room has two points of egress”, not “each bedroom has a window, from which you could technically jump though you would definitely die”. 

I am strongly in favor of requiring access to the external fire escape in buildings which only have an external fire escape, and I’m just as strongly in favor of requiring sprinkler systems and regular maintenance in buildings where windows are not a realistic egress. 

And like - being housed instead of homeless is a safety concern too. Feeling able to leave an abusive partner because you can house your children on one income is a safety concern too. Having savings is a safety concern too. We should absolutely prioritize safety, but when it comes to housing that means identifying the most important regulatory concerns, regulating those strictly, transparently and fairly, and then encouraging the building of tons and tons of housing so that no one is houseless or insecurely housed and so that people can meaningfully afford to leave bad situations. A city with median rents over $1k/bedroom/month is unsafe for poor people no matter how many safety regulations they pile on, and in fact past a point the regulations, even if well-intentioned, are making people less safe.

[jimrandomh]

Counterexample: This cat. Also, you are a cat.

Cats

Cats have no positive qualities.

[jimrandomh]

I argued against trying to have discussions with explicit tagged claims and arguments in https://arbital.com/p/7gc/ . It's been tried, and it doesn't work even a little bit. The problem is that the stuff that goes on in typical textual discussions is too varied to structure formally; there's a very thick tail of types, such that the minimum viable ruleset is too copmlex for people to keep track of while also keeping track of an object-level topic.

Hot take: the rationality community would be significantly improved if everyone were forced to write blog posts in the form of bullet points. A bullet point can be a claim or a supporting argument. It cannot be an anecdote. It *certainly* can’t be an analogy. No adjectives, adverbs, or other modifiers are allowed. Claims must be labeled as such, and must be accompanied by prediction markets.

He said, not doing any of that. (Actually in the brief period when I had a real blog I did do the prediction market thing but oh well)

[jimrandomh]

And with with our beliefs and actions we try to optimise the one thing $U-KL(Q,P)$. That is (how good we think life is) - (how wrong we are about how the world works)

That model seems trivially falsified by the existence of pessimists.

I’m going to judge my attempt to understand free energy as basically a failure.

At this point, I don’t think I even understand the most basic point about how a probability distribution equals a utility function. What’s the probability distribution equal to “maximize paperclips”? Is it “state of the world with lots of paperclips - 100%, state of the world with no paperclips, 0%”? How do you assign probability to states of the world with 5, 10, or 200 paperclips?

http://slatestarcodex.com/2018/03/04/god-help-us-lets-try-to-understand-friston-on-free-energy/#comment-607698 seems like the best comment I got, and the one that should explain it to me, but I can’t understand it any better than the original <s>Greek</s> Friston. I think (1) and (2) are just saying that you’re trying to create a model that predicts the data you get given your prior and what you know about the position of your body. I think I lose the plot around (3) - I know it’s something about using motor stuff to enact your predictions, but I can’t get more than that.

Nobody’s thoughts on how drives like sex minimize uncertainty, or solutions to the dark room problem, seem consistent with each other or really sound very well thought through.

I wish I was better at math, or other people were better at explaining things without using math.

[jimrandomh]

I think the problem with modern forensics is more of an incentives problem; governments will pay for forensic analysis but they aren't paying for quality control, or checking that they're getting what they pay for. Consider https://www.washingtonpost.com/news/true-crime/wp/2017/12/28/massachusetts-prosecutors-to-throw-out-8000-convictions-in-second-drug-lab-scandal/?utm_term=.c8240b8a071a for example; that sort of thing would be impossible if even a small number of control samples were mixed into the test stream.

Terra Ignota security is *hilariously* bad. The Saneer-Weeksbooths *answer the door*. The physical prowess (and swordedness) of their visitors is *relevant at all*.

Ockham Saneer you *do not* have good precautions and you should be ashamed of yourself.

[jimrandomh]

I wrote a review of the Impossible Burger, from a nutrition perspective, as compared to beef. Overall: mostly positive, but there are a few things I'm hesitant about. https://conceptspacecartography.com/a-nutrition-focused-review-of-the-impossible-burger/

[jimrandomh]

I'm fat, and I have a restrictive eating disorder

This post may be triggering. I’m about to talk about being fat, and my eating disorder, and an exchange I had with a new therapist at my treatment center. It’s an assumption that has become so common place it hardly phases me anymore. But it should. It needs to. Because it’s an example that even within a treatment facility with a weight set point mind set and a health at every size governing philosophy, when the staff see me, and see my body, and my size, there’s an assumption that I either have binge eating disorder or compulsive eating disorder.

I’ll share the interaction, and then why it’s so important. I sat down to an orientation for a new group I’m joining. The therapist and I have met in passing, but have had no direct interaction before this. She asked how my recovery was going. I told her I’ve had a lot of set backs over the last six months, but I’m four weeks in to eating AND sleeping, which is a huge accomplishment. I’m really starting to feel true hunger and my dietitian is thrilled that I’m actually eating. And the therapist said, “Oh, so, you mean feeling hunger and eating in a good way?”

And I just stared at her for a moment. I’ve gone for so long not knowing what actual hunger felt like, doing anything and everything to busy myself the moment my body twinged in that direction, suddenly becoming desperately busy and oh look, it’s too late now, I’ll just eat at the next meal, or perhaps the next…

When I first entered treatment, I was nauseated all the time. The amount of food on my meal plan felt impossible. Every meal was a nearly unbearable experience. My meal plan’s been raised multiple times since then. Now, I wake up hungry and cranky until I feed my body. It’s still hard to eat. But I’m doing it. Every day.  So many things have changed. 

But not that I’m fat.

Every now and then, I get to forgot that I’m fat for a little while. It’s usually not long before I’m reminded, before an assumption, or a seemingly innocuous question brings it back to the forefront. Right. I exist in a large body. There are a lot of assumptions that go along with that. 

 And my answer was a laugh, and a clear explanation, because I’ve had to do it so many times before, ticking off each thing on my fingers. “Oh! I have a restrictive eating disorder. My behaviors are restricting, purging, starvation, diet cycling, and using exercise as a way to punish my body. My meal plan is a minimum I have to eat each day, going over that is great but I can’t go below. And I have restrictions on water, because I have a history of water loading.” 

 "Oh!“ was her reply. 

I followed immediately with the example I always use, preempting the inevitable bingeing question. “When I started here, I thought I binged. I was like, ‘I totally binge. I’ll eat an ENTIRE cheeseburger!’ And the staff was like ’…That’s not a binge.’”

And she stepped in and finished the sentence with, “…right, that’s a portion.”

And suddenly we were back on the same page. She had adjusted her internal diagnostic criteria of me, and what I struggle with, and the rest of the orientation went great. I’m excited about the group, and I think she’ll be a great facilitator. I do feel a bit heartsick though that she didn’t take a moment to look at my file, or speak to my current treatment team before our orientation. It would have saved me from that moment, where I have to explain, again, that while yes, I’m fat, the disorder that I’ve been struggling with for over 30 years is restrictive, based in control, anxiety, and trauma, as well as genetics and a solid dose of disordered modeling. The reactive eating I’ve done, what feels like bingeing to me, falls squarely in the realm of a portion, or at its most extreme, a large meal. 

Now there’s some great stuff over at www.youreatopia.com and @bigfatscience & @everythingeatingdisordered if you’d like to understand how you can not lose weight, and even gain weight, while being in a restrictive state. This is especially true while diet cycling, which has been my path in between ED behaviors for the past 15-ish years. To quote @heavyweightheart “It turns out human bodies are a little more complicated than lawn mowers.” We’re not just fuel-in, energy-out robots. There are all kinds of processes that occur making us very different than, say, a car’s engine. And each of us has unique needs. 

 And here’s where we veer into the why: Why these underlying assumptions matter so very much, why they must be examined and challenged, and why there needs to be more open and honest dialogue around the reality of eating disorders occurring in individuals of all body sizes. 

I went to my neurologist recently to discuss the increase in my migraines in both frequency and severity. I told him in clear and blunt language that I’m in the early stages of eating disorder treatment for a restrictive eating disorder and on exercise restriction. He asked why. Not kindly, or with compassion. But very bluntly. “Why?” And I explained. 

He then went on to bring up weight loss three times and walking an hour a day four times in the remaining time of the appointment. Did he say these things over and over because he thought this was what was best for my overall health, taking into consideration the serious damage I’ve done to multiple organ systems with my decades of starvation and purging? No. He said these things because of the theoretical damage that could occur based on correlation between these issues and obesity. My blood pressure, pulse, blood sugar, and other metabolic measures of health look pretty damn good for someone with as many health issues as I have, none of which are weight related, but many of which are ED related. 

It’s so ingrained. You’re fat. Get healthy. Lose weight. Exercise. My psychiatrist put it really well. She said, “If you had gone in for a well check, checking in and making sure you’re eating a variety of food and getting some exercise makes sense when discussing migraines. But it’s like you went in and said, ‘I’ve been through hell. And I’m still there.’ And his response was, ‘Let’s get some walks in!’ He didn’t actually listen to you, or hear you.”

I was asked recently whether I thought there was any point when inpatient treatment would have been appropriate. And I said yes. When I was 19, and passing out daily in class, and my movement teacher told me I couldn’t return until I’d seen the doctor. They diagnosed me as hypoglycemic. I was at a “normal” BMI, the smallest I’d ever been, but still not underweight. I had never missed a period, and the fact that I worked out 2-8 hours a day showed how really dedicated I was to my health. I used the diagnosis as an excuse to become a vegetarian and cut even more food out of my diet, and learned how to eat just enough to not pass out in class. 19 years after that, I finally sought treatment. 

We say it over and over and over. Eating disorders impact all manner of people. Eating disorders are mental health disorders first, and cause a myriad of health issues, further psychological issues, and relationship and social issues. And not everyone who has them is underweight. Yes, the pictures of skeletal women are the most shocking, attention grabbing, and likely to get people to raise both money and awareness, which ED treatment desperately needs. But the cost is that the rest of us, those in average size bodies, and yes, those of us in fat bodies, are receiving extremely dangerous messages, repeatedly, about our bodies, and our health. 

I’m not referring to our theoretical some day health, that health that my neurologist is so focused on. I’m talking about our health right now. My biggest wake up call in treatment was when my dietitian sat me down, looked at me seriously, and said, “We need to talk about your weight.” I flinched, sure that she was going to say that I’d gained a huge amount of weight adding in more food, that I was messing up my meal plan, that I was failing at recovery. 

Instead, she said, “You’re losing weight alarmingly fast. You have to start eating again. Everyday. Your full meal plan. All that long term damage you thought you wouldn’t have to worry about?” She pulled out some charts and showed me some very scary things. “You’ve been doing this for 30 years. It’s now. The long term damage is happening now.” I was shocked. Wasn’t I supposed to lose weight? Wasn’t I supposed to shrink away into nothingness? Or at the very least, into a “normal” BMI?

My restriction, dieting, and exercise addiction were supported and even recommended to me by my doctors. And what’s so scary is that when I tell doctors outside my treatment team I’m in recovery, and on exercise restriction, it’s still being recommended. My dietitian was horrified. My therapist was furious. My psychiatrist gently but firmly suggested I find a different neurologist. 

It’s hard enough to confront that what you’re doing is hurting yourself. It’s excruciating to confront that your eating disorder also hurts the people you love, and who love you. There’s no break from recovery. There’s no day off. And when you’re fat, you also deal with a daily onslaught from every direction suggesting, prodding, lecturing, even screaming at you that you are invalid in your body. That you are not worthy as a human being until you reach a certain weight. And that your well being is dependent on you losing weight.  

We live in a culture where being thin is so closely aligned with being healthy, that that goal is worth anything. It’s worth a doctor telling me to actively do things that will harm me. He’s not the first. He’s just the most recent this month. Not this year. Not this season. This month. It causes assumptions so deeply ingrained that even amazing, supportive therapists at the front line of this work still catch themselves making a judgment based on the size of the body of the person sitting in front of them.  

I’m having to unlearn a lifetime of information while it’s still being forcefully given to me by doctors, friends, biological family, the media, random strangers on social media, random strangers on the street, and every single packaged food loudly proclaiming it has less fat, less calories, and more all natural ingredients!!! I have had to narrow my focus. I listen to my treatment team. I do my own research. I trust the chosen few who are supporting me with everything they have. 

I’m fat, and I have a restrictive eating disorder. These things can, and do, exist in the same body, at the same time. If you take anything away from reading this, know that I am not the only person who exists in a large body who has a serious eating disorder. And it may not be the kind you think.