PokerPirate
6d ago
Medicine is hard, and answers to important medical questions can't fit in the length of a tweet. I have a phd in machine learning, so I'm confident I could form an opinion on your questions if I tried really hard and read a bunch of papers and thought about the problem for a week. But I don't care to do that, and so at some point I have to trust other people's judgements.
that the absence of certain design features tends to mean a study's finding will tend to be overturned, if done properly?
Part of the parachute study's point is that RCTs are not enough! And you are placing too much faith in RCTs! It's very easy to design a RCT that "looks good from the outside" but has a fatal flaw that makes it not applicable to the real world. In the parachute example, the fatal flaw is that the plane was grounded the whole time. Downthread, people are pointing out a bunch of fatal flaws in hypothetical RCTs for gender transition that would undermine any possible conclusion.
No matter what the methods are of an experiment, you can't get around having to sit down carefully and examine all of the assumptions.
PokerPirate
6d ago
There is not a single randomized control study of gender transition, in either children or adults.
You're overstating the importance of randomized controlled trials (RCTs) in medical research.
As a famous parody of your point, this 2003 study found that no RCTs had been done of parachute use when jumping out of airplanes and concluded that there is insufficient evidence to conclude that parachutes are effective. As a follow up, this 2018 study did implement a RCT for parachute use when jumping from airplanes and concluded that parachutes do not in fact prevent injury. (Participants jumped from an airplane on the ground.)
Less facetiously, we have no RCTs demonstrating that HIV causes AIDs, but we can still be pretty confident about the link between the virus and the disease. Recognizing this relationship has led to a lot of good medical progress both for the populationis affected by AIDs and those not affected by AIDs (by for example keeping HIV out of blood transfusions to prevent the spread of AIDs).
I happen to also be skeptical of the benefits of transition. But your explanation of the science is not good here and at best leading you to the "right belief for the wrong reason".
PokerPirate
9d ago
That's a good point, thanks. I'm guessing that insurance costs are up, and it'll be hard to get a reliable quote before actually purchasing.
PokerPirate
10d ago
I'm going to buy a house in Southern California (LA area) sometime in the next 3 years. I'd like to time the market a bit, and I'm wondering if anyone has any insights to what housing prices might look like in the Trump admin?
My naive analysis is that prices will trend downward for a couple of reasons:
- if immigration goes down then there will be more supply and prices will go down,
- interest rates have gone up recently, and that drives house prices down, and the market is still catching up to this trend (I'm more concerned about the overall price than the monthly payment for various reasons).
PokerPirate
12d ago
I'm an outsider to the legal system, not to the US :)
PokerPirate
13d ago
I like these court opinions that you (and others) post. For an outsider like me, it's nice to get some insight into how the legal system works and to not only be exposed to "activist judge did XXX" type cases.
Thanks :)
PokerPirate
20d ago
There's a reason communism is the bourgeois ideology par excellence.
As a linguist, this is one of the best examples of linguistic drift I've ever read.
The 1800s communists and bourgeois would have obviously disagreed with this sentence (because communism was about stripping the bourgeois of their power and giving it to the working class).
But you're not using these terms how Marx and his contemporaries used them. The way I'm reading you is that: the bourgeois is idealized by the DINK couple who works an email job and got a degree in "gender-studies"; communism may-or-may-not be the traditional purely economic theory, but it likely has incorporated a lot of generic social leftism (that we would expect to be taught in a gender-studies program).
PokerPirate
27d ago
It's certainly not a golden ticket. @cjet79's original claim is that veterans from a hypothetical European-Russian conflict will have outsized political influence in their respective governments. I'm just saying I think we've seen that trend in the US where the political class has a much larger fraction of veterans than ordinary citizens.
PokerPirate
29d ago
FWIW your point seems obvious to me about the US. There have been 14 US presidents since WWII, and 8 of them have military experience.
PokerPirate
1mo ago
For the parents: How do you introduce/talk about yourself and your spouse to your kids friends? Are you on first name basis? Do you go by Mr/Mrs so-and-so?
My sense is that Mr/Mrs so-and-so makes it easier for young kids to understand their relationship to you and that they need to respect your decisions. Some of my peer group, however, goes by a first name basis. Most parents never even broach the subject with their kids' friends, and so the friends are in an awkward limbo where they don't know how to address parents.
PokerPirate
1mo ago
The song came first, but I doubt Stallman or any of the fellow GNU folk had that song as an explicit motivation.
PokerPirate
1mo ago
Whoa... you're blowing my mind...
But do you really pronounce something like ISO/IEC 27000:2018 as "ey-so slash eye ee cee ..."? I guess I never refer to the name of the org without a bunch of other numbers after it.
I maintain a git repo about the "programmer dialect of English" that I use to teach my computer science/data science students how to not sound like a n00b/PHB. It's niche, but many people here are likely to fit into the niche.
The document goes over phonology / lexicon / grammar / discourse differences in programmer English vs American standard English, but to give you all a taste, here's the top of the phonology section:
-
RAG (as in retrieval augmented generation with LLMs) is pronounced "rag" not "R-A-G".
NOTE: The word "rag" has only one sylable and is faster/easier to pronounce than the 3-sylable "R-A-G". The practice of pronouncing abbreviations as acronyms stems from the programmer's desire for efficiency in all things.
-
ICLR (a famous machine learning conference) is pronounced "I clear".
-
PNG is pronounced "ping" and not "pee-en-gee". This pronunciation is specified in the standard.
The G in GIF is pronounced like the G in GIGANTIC.
PokerPirate
1mo ago
I've enjoyed reading everyone else's advice here, so I'll add in my own: It's your responsibility to manage your own parents (and protect your spouse from their inlaws).
In the extreme case: my parents traumatized my wife ~1 year into our marriage by getting into a yelling match over politics/religion. I had to tell them that they were out of line, and that I'd be cutting contact with them if they continued to behave that way.
Less extreme: my wife's relatives regularly give too much sugar/presents to our kids, and it's my wife's job to let them know when to stop.
PokerPirate
1mo ago
This is my exact relationship with my kids/wife. I suspect it's overall healthy for the kids to have one "strict" parent and one "soft" parent, but I agree that modern mothers have gone off the deep end. I often find myself having to be harder on the kids than I'd like to be in order to strike a better balance.
PokerPirate
1mo ago
I've got a 1, 2, 3, and 6yo. All were sleeping through the night by 6 months. I (father) was responsible for the night shifts, and did basically the Ferber method.
PokerPirate
1mo ago
We've got an all-in-one Narnia that I got used for $2. It's fantastic for an adult, but for kids I'd go with the smaller size. The font on our all-in-one is a bit too small to be comfortable for a kid to read (or to point with you're finger at what you're reading to show a kid), and the weight is too heavy for them to comfortably carry it around. Mine is ~400 pages but they are very large (bigger than US letter paper size).
PokerPirate
1mo ago
That's a great idea, I'm going to try it :)
I started my kids off with Harry Potter. After watching the movies, it's much easier for them to visualize what's going on when I read. I read books 1 and 2 to my 6 year old that way.
I thought about doing the same with the Hobbit or Narnia, but the movies aren't quite as friendly to young kids. (I've also got a 1, 2, and 3 yo right now that even Narnia would be too much for. Harry Potter is full of funny scenes in between the scary ones that these other "kids" movies don't have.)
PokerPirate
1mo ago
I like your shared thoughts.
But I also really want the AI to go away and would be supportive of more extreme moderation in the future. I haven't seen a single thread with AI content that I thought was productive.
PokerPirate
2mo ago
in the United States it is illegal to use surveillance aircraft and NSA SIGINT assets to hoover up reams of data and then act without warrants in response.
This isn't really true in the sense that you're implying. Ordinary police regularly use surveillance aircraft to establish probable cause for an arrest. Ordinary police don't quite have NSA SIGINT assets, but they do still have quite a bit of SIGINT assets they use regularly. If there was political will to arrest/imprison drugees, it would certainly be possible for the police to do so legally.
PokerPirate
2mo ago
When Bush walks over, Obama stands up to shake his hands like an old friend. Trump stays seated and neither he nor bush acknowledge each other. Then Obama and Trump are back to cracking jokes with each other.
PokerPirate
2mo ago
LLPSI is pretty intense for an elementary schooler. There's much better material for young learners. Minecraftium is a youtube playlist of a Latin teacher playing minecraft while talking latin, and is a good example of something in the same "natural" style but geared for the younger audience. It's not a complete course by itself, but there's lots of other similar material out there.
PokerPirate
2mo ago
I think the phrase "quantum woo" vastly understates the potential impact of quantum computing on machine learning. The quantum algorithm zoo, for example, lists a number of quantum machine learning algorithms. Several of these get exponential speed up from classical algorithms, but even a quadratic speedup of grover's algorithm would be game changing at the scale frontier models operate on.
I agree that most normie use of quantum in the brain is "woo". And I also agree that it's not been established that the brain relies on any quantum effects. But there is actual legitimate research in these directions and it seems wrong to offhandedly dismiss it.
For those (like me) wondering why "female" should be related to "sucking", it's because the babies suck on their mothers breast. (See https://en.wiktionary.org/wiki/female and follow the etymology links.)
Another fun note is that the old-english for male was "weaponed" because their was a weapon between their legs: https://en.wiktionary.org/wiki/w%C3%A6pned#Old_English
All I'm trying to say is that your original post overemphasized the importance of RCTs in medicine. I'm not trying to make any claim about gender dysphoria or its treatment.
More options
Context Copy link