Hacker Newsnew | past | comments | ask | show | jobs | submit | Gud's commentslogin

What makes you assume people who use LLMs don’t understand what’s going on?

I wouldn't call it "some inflation". The living standard of the western middle class has been on the decline for a long, long time.

No it hasn't.

Expectations are higher, competition is stiffer, and the gap between bottom and top end has grown, but by and large (especially in the US), the middle class quality of life has gone up.

Obviously specific regions that failed to transition out of low value-add manufacturing and agriculture have suffered, but the vast majority of Americans live in cities doing or supporting high value work.


It's not even competition anymore. It's a screaming void that deafens everyone, causing them to reach for the nearest "acceptable" thing just to quiet the endless cacophony of human struggling.

> the middle class quality of life has gone up.

As long as you don't try to buy a house.

I see kids, right out of college, making more than I ever made, at the peak of my career, unable to afford a house.


Yes this is a big problem but a large part of this is the total elimination of starter homes from the market. I.e. they would be able to afford the types of homes that earlier generations started in, but those homes simply don't exist anymore.

It's kind of a quality of life degradation, but it's a bit more complex than just "an attainable item is no longer attainable." It has never been normal to buy a 2600 sqft, 4 bedroom home at the start of a career.


It's not that starter homes were eliminated or were torn down, it's that construction stopped in cities. The downzonings of prior generations, combined with the limited ability to expand by car travel, finally hit its limit and the urban planning apparatus was in complete capture of people who didn't want the built environment to change.

Would be nice if true, but not really.

The reason construction slowed down so much is that developers fear another 2008. We have just barely gotten back onto a historically normal-ish pace of construction: https://fred.stlouisfed.org/series/HOUST

And this talk of "just build build build," while not wrong per se, overlooks the fact that of course prices will come down, which then discourages construction. The system is self-equilibrating. 2008 reset the equilibrium point very low for 15 years, and now the nature of the costs of construction (labor and land) means it is not advantageous for anyone to build starter homes, and it's hardly advantageous to build homes at all.

Restrictive zoning is a problem and would be a very tidy explanation of all the woes of residential in the US, but there really isn't much evidence for it mattering that much in the grand scheme of things.

The single most important factor in home prices is local income levels. This gets baked into both land prices and labor costs, which then makes it very difficult to profitably build much, and completely unprofitable to build entry level homes.

The K-shaped economy is itself causing housing unaffordability. https://www.nber.org/papers/w33576


That would be nice if true, but not really.

The building industry never really recovered after 2008 because the only surviving companies were extremely cautious. In order to get more builders, there needs to be more places to build, and entry into the industry needs to be easier. It's all permitting, zoning, and discretionary processes stopping housing from being built where it's wanted to be built.


Well I've shared a statistical analysis and raw data series backing my points and directly contradicting yours. On the flip side I guess we have "trust me bro."

To the extent "it's all [any individual cause]", that cause is rising incomes. The second major cause of rising housing prices is cost of inputs (labor, land, material). Zoning definitely plays a role, but again: there's just no evidence that "solving zoning" will actually solve affordability. We should do it anyway because it'll solve all sorts of other problems in our built environment, but there's not good evidence affordability is one of them.


You ar also doing "trust me bro" with a statistical analysis that at most shows that prices rise with wages when supplies are constrained. Which, yes! That's what everyone says! K shaped recoveries happen when there's unequal access to opportunity, and supple constraints in access to the geography of good incomes is exactly the sort of supply constraints. Further, in order to get their weak results they do silly things like transform "supply constraints" into an indicator variable, and on the basis of that single odd regression try to overturn a huge body of literature showing the opposite.

Yet this one strange paper keeps getting cited as if it were God's own truth, the holy grail of economics that changes everything that was known before.

Supply restrictions are not binary, though that's how your paper treats them, and they perform none of the causal analysis that would be needed to extend their analysis to the conclusions you are trying to draw.

Here's a random paper with completely different results that agrees with the rest of the field:

https://www.sciencedirect.com/science/article/pii/S009411902...

I remember the last time the "we can't change zoning" folks passed around a paper like the NBER paper you shared, and it was one about transit-oriented-development in Chicago, where allowing small upzonings close didn't change pricing much. It was contra to the vast majority of the literature, covered only a small geographic area with fully adequate housing supply, yet for a few years nobody could suggest doing the obvious zoning reforms without people claiming that Chicago proved that upzoning doesn't change pricing.


And again, supply will always be constrained below "affordability" by virtue of there being no profit available at affordable price points given the costs of inputs. So yes, if we imagine a world where supply isn't constrained first by the actual cost structure of construction, then clearly artificial constraints are the sole problem and solving them would solve the problem overall. But that's not the world we live in!

From your random paper:

> Fig. 5 shows the event study results for the change in log hedonic rents. In contrast to the housing supply, we find no statistically significant effect of upzoning on rents.

So it looks like your paper actually agrees with mine.

As I've said over and over: we should overhaul zoning for sure. However there is not good evidence that will solve affordability, and there's basically zero evidence that it is the cause of "all" the problems, as you so boldly claimed.


That's true.

From what I can see, those houses are being brought up by corporations, and turned into rentals.

Rental-only society is definitely possible (see Manhattan and Tokyo), but is a very different model from the traditional American suburban dream.


> I wouldn't call it "some inflation". The living standard of the western middle class has been on the decline for a long, long time.

IMHO the main problem nowadays, especially facing young people, is housing.

Otherwise there is probably never been a greater time to be alive, generally speaking, than right now. If you believe there is, can you outline the year(s) in question and how they were better?

As for inflation, using Bank of Canada numbers (since I'm in CA), $100 of goods/services from 1975-2000 increased by 220% to $320.93, while $100 of goods/services from 2000-2025 increased by 71% to $171.22.

In a 2014 article, CPI from 1914 to 2014:

* https://www150.statcan.gc.ca/n1/pub/62-604-x/62-604-x2015001...

From 1955 to 2021:

* https://economics.td.com/ca-inflation-new-vintage

1971-76 and 1977-83 had double the CPI of ~2021.

While unpleasant, and higher than that of what many young(er) people have experienced, it is hardly at a crazy level. The lack of people's experience of higher rates is simply more evidence as to how stable things have generally been:

* https://en.wikipedia.org/wiki/Great_Moderation

Tom Nichols argues that it is boredeom that's the problem: people want some excitement and are willing to stir the pot to get it:

* https://www.usatoday.com/story/opinion/2021/08/19/donald-tru...


No, you've just fallen victim to the hedonic treadmill.

What about self hosting?

I talked about that in this section https://simonwillison.net/2025/Dec/31/the-year-in-llms/#the-... - and touched on it a bit in the section about Chinese AI labs: https://simonwillison.net/2025/Dec/31/the-year-in-llms/#the-...

Personally I use ChatGPT a lot, it is a wonderful service.

I use it in conjunction with Claude. I’ve gotten pretty good results using both of them in tandem.

However on a principal basis I prefer to self host, I wonder if an advantage of OpenAI imploding wouldn’t generate basement level prices of useful chips? Ideally I want to run my LLM and train it on my data.


Same. I started coding before hitting puberty, and Im well into my 30s.

If you know the problem space well, you can let LLMs(I use Claude and ChatGPT) flesh it out.


> I use Claude and ChatGPT

Both for code? For me, it's Claude only for code. ChatGPT is for general questions.


Yes, I use them in tandem. Generally Claude for coding and ChatGPT when I run out of tokens in Claude.

I also use ChatGPT to summarise my project. I ask it to generate mark down and PDFs, explaining the core functionality.


Not really, how are they prevented from using the manual or the copious amounts of examples out there?

Memorising command line options beyond the absolute basics has rarely been helpful to me. And I use FreeBSD, where arcane commands are plentiful.


Nothing, but after becoming reliant on an LLM they may simply become overwhelmed and give up once they outgrow it's capabilities. I've seen this happen to several people I know.

“It’s really not that hard”, well a lot of people have better things to do than remember parameters to commands we barely use.

Just make sure it follows open standards and you’ll be fine.

You can have multiple tabs open.

Sorry but if you work for a giant advertisement agency you are part of the evil organisation. You are responsible for what they are doing.

If you are born in a country and not directly contributing to the bad things it may be doing, you are blame free.

Big difference.

I never worked for Google, I never could due to ideological reasons.


Even if what you’re doing is making open source software that in theory benefits everyone, not just google?

FWIW I agree with you. I wouldn’t and couldn’t either but I have friends who do, on stuff like security, and I still haven’t worked out how to feel about it.

& re: countries: in some sense I am contributing. my taxes pay their armies


When you work for Google, you normalize working for organizations that directly contributes to making the world a fucked up place, even if you are just writing some open source(a corporate term, by the way). You are normalizing working for Google.

And regarding countries, this is a silly argument. You are forced to pay taxes to the nation you are living in.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: