It isn't one cause, but the vast majority is interrelated causes, stricter building codes, stricter zoning and more bureaucracy making it harder to build more. The fact that parts of the US which make it easier to build see housing prices go up slower is one of a variety of pieces evidence that this is the largest aspect. It may help to read the links I gave earlier.

Absolutely not. Housing prices are up because regulation has made building more housing more expensive if not outright impossible. See e.g https://www.cato.org/policy-analysis/zoning-land-use-planning-housing-affordability https://www.urban.org/urban-wire/how-communities-are-rethinking-zoning-improve-housing-affordability-and-access-opportunity https://www.noahpinion.blog/p/market-rate-housing-will-make-your. Inflation is not up because people are competing for products. In fact, inflation is due to a bunch of factors including Trump's tariffs, but more people means more economies of scale which keeps prices down. https://en.wikipedia.org/wiki/Economies_of_scale Pollution is not up due to too many people driving. In fact, most forms of pollution are declining in the long-term. US CO2 production has been trending down https://www.statista.com/statistics/183943/us-carbon-dioxide-emissions-from-1999/ . And doctors and nurse shortages are part of why medical costs have gone up, but that's because the US has a massive bureaucracy to handle insurance, an aging population https://www.pgpf.org/article/why-are-americans-paying-more-for-healthcare/ and because the US has not had the number of doctors and nurses increase with population. The US has a very low percentage of doctors to population compared to many other developed countries https://data.who.int/indicators/i/CCCEBB2/217795AThis occurred in part because in the 1980s people were worried about a doctor surplus and so we spent 30 years deliberately reducing how many new doctors there were https://thedailyeconomy.org/article/how-congress-created-the-doctor-shortage/. Now a critical way of getting more doctors in the US is immigrants. https://www.aamc.org/news/1-5-us-physicians-was-born-and-educated-abroad-who-are-they-and-what-do-they-contribute.

Fair; I should have said something like "primarily" about, since it does mention the reaction, but the actual focus is on films about AI. I overstated the issue.

I know reading TFA is hard, but at least read the summary before commenting. The article is about people not wanting to watch films about AI. It is not at all about using AI to write or animate films.

That's not remotely true. A murderer is a murderer whether or not they've been caught. What is true is that someone who is not convicted of a crime should not be punished by the government. That's a radically different statement.

Or just a habit of how to keep things that you want all in one place. I email myself reminders, and have only been gradually transitioning to being more disciplined about using things like Google's task list or putting even very rough drafts in Google docs or Overleaf. The Petraeus thing was interesting because by keeping it in a draft it deliberately evaded things, and also was interesting because of where Petraeus got it from; it was a technique used by Al-Quaeda.

A combination of culture and large-scale market factors. There's no single individual making a conscious decision. Asking who created the incentives is a bit like asking who made evolution favor crab-like crustaceans.

ChatGPT has a version which is FERPA compliant, and high school and university teachers are told explicitly not put any student names or anything else sensitive into personal ChatGPT or other AI accounts. I don't use the teacher version of ChatGPT, in part because I've never had need of any interaction with an AI where having an AI would be useful and where a student's name or other identifying info would show up. I honestly struggle at seeing what the reasonable use cases are in that intersection. (Also, I'm slightly cynical/skeptical about their promises to not use anything from those accounts for new training data.) But the bottom line is that this sort of thing is something which would get stern rebukes up to being fired if it happened in a high school environment. And this is the nominal head of cyber security for the US government. So the question is how much of this is part of the Trump's administrations general tendency to not hire people who are competent, how much is the tendency for people (in any administration) who are powerful to just ignore the rules, and how much of this is that there are some people who are really into credulously using LLM AIs and are just idiots?

Lots of people like row houses. In places like New York and Boston, those sort of houses are in high demand. it turns out that lots of people want to live near other people and actually like the density. As for high population, we all benefit from the high population. More people mean more ideas, more comparative advantage, and more economies of scale, which translate into better standards of living in general. There's some point beyond which very high populations would cause standards of living to get worse, but we're not near that point, and given technological improvement, we're likely moving farther from that point.

Meanwhile, even if you put everyone into individual suburban homes, even with a lower population, you do tremendous environmental damage because suburbs spread out more, and because they require more roads, pipes and other infrastructure, they end up costing a tremendous amount of resources. Functionally, cities are often subsidizing the suburbs https://www.strongtowns.org/journal/2023-7-6-stop-subsidizing-suburban-development-charge-it-what-it-costs .

Not at all. That would require a multi-country, massive conspiracy over a hundred years. Instead the results are much more prosaic; the same basic economic incentives as well as the same basic desires (against change, neighborhood expansion, or new people who look or sound different) all come together badly. Never propose a nefarious conspiracy when simple incentives will do.

Huh? No you are missing multiple things here. Did you see the words "in practice" in the bit you quoted? That's the important; of the size of statements which humans care about, the vast majority of statements are decidable. (The reasons for this become more apparent if you unpack the details of the construction I linked to.) And the last sentence isn't really accurate, or at least it is missing context: Operating under any given axiomatic system, Godel's theorem limits humans as much as it does any machine.

Yes, you do misunderstand. There are two different theorems frequently referred to as Godel's theorems which are closely related, but in this case, the relevant one you are referring to is the statement that says roughly that any axiomatic system which is strong enough to model the natural numbers must have statements that cannot be proved or disproved within that system. This is a rough summary of a somewhat subtle result. Note that one needs to be really careful about what one means by axiomatic system otherwise one could just take as one's axiomatic system all true statements. Also, although the theorem is normally phrased as needing to model Peano Arithmetic https://en.wikipedia.org/wiki/Peano_axioms one really only needs to model the much weaker Robinson arithmetic https://en.wikipedia.org/wiki/Robinson_arithmetic.

So, why isn't this relevant here? Well, the simplest reason is that empirically speaking, undecidable statements are in practice extremely rare. So the vast majority of the time, the existence of such statements being undecidable doesn't matter. We have good reason for thinking that as statements get very large, a positive fraction of all statements of a given length are going to be undecidable but this likely doesn't kick in until one is frequently looking at long, convoluted statements that humans don't normally think about https://mathoverflow.net/questions/4454/how-many-of-the-true-sentences-are-provable/7902#7902. So for math that we care about this doesn't really matter. But note that even when thinking about an undecidable statement, all the aspects of the problem are still in the problem itself, it just is that there may be no solution to the problem in the axiomatic system you care about. Closely connected to this last observation is the fact in so far as computers are limited by Godel's theorem, so are humans.

Does that help?

What utter nonsense. First, Israel is a tiny country in the Middle East. It isn't anywhere near Russia or China. So, blaming Israel for what superpowers are doing is pretty off the wall, and I'll let others here speculate on why one would even think that it is remotely reasonable.

And while it is true that the United States is no condition for sainthood, Russia and China have been highly aggressive for their own aims. Russia has a long history of cutting off parts of neighboring countries. They took part of Georgia, they took Transnistria from Moldova, and they took Crimea from Ukraine. The Ukraine invasion is simply their most blatant version of that, a war of aggression to capture territory, of the sort that was common before the end of the 2nd World War. Putin has decided for his own reasons which are essentially to increase the size of Russia and make Russia larger to engage in this war.

In fact, this war is so blatantly a war of aggression that Putin started it when after Zelenskyy was elected. Zelenskyy is a native Russian speaker, and when he was elected, he was seen as someone who would be willing to work with Russia. In fact, many in the West were concerned he would be too pro-Russia. And even given that Russia still invased.

The basic point, that the world's situation is getting more precarious is a valid one. The entire Doomsday Clock itself is a pretentious and somewhat silly way of expressing concern about the world. It gives an illusion of careful precision (what makes it 4 seconds closer and not 3 seconds closer or 5 seconds closer) to what is largely a political decision.

The summary is missing some pretty important words in TFA. Here's the context of where he said the models aren't there yet:

Notice the important part before the sentence; he's talking about "game-changing new discovery" not regular nuts and bolts of scientific work. And in at least some fields, we are seeing these used to accelerate work. He talks about in the article a bit. The article correctly notes that there's been inaccurate hype from his group in the past, including claiming to have solved open math problems where it really just found solutions in obscure literature. But it is important to note that there has been successful work using LLMs to do math, especially when the systems are paired with more traditional theorem proving software and proof checkers like Lean. There's a long list compiled on Mathoverflow of successful examples https://mathoverflow.net/questions/502120/examples-for-the-use-of-ai-and-especially-llms-in-notable-mathematical-developme and this list is now getting long enough and this is getting common enough that the Meta Mathoverflow has had discussion about stop expanding the list https://meta.mathoverflow.net/questions/6348/use-of-llms-in-notable-mathematical-developments. Right now, the most successful system seems to be Aristotle, which is a software which combines LLM style informal reasoning with automated proof checking.

Now, math has some major advantages over other fields here: all the content is in the paper. In a lot of other fields, raw data and the like is not in the papers. Moreover, the universe is weird and complicated, and LLMs cannot just run experiments. In contrast, in pure math, all the problem aspects are in the problem statement itself. I am pretty skeptical that LLMs will ever get to the point of making "Einstein-level" discoveries since by nature they are still working off their training data. Without some major breakthroughs, it seems like they will still be fundamentally limited. But we shouldn't use their lack of that ability to dismiss that they are being used now to do genuine novel work, and one shouldn't summarize articles in a way that suggests someone was saying something they were not.