> As someone who is very interested in AI and taught myself how to code (I don’t know any other real estate people who know anything about code), I think it’s going to be incredible hard to uproot the brokerage industry.
Has learning to code generated any benefit to you? What areas do you think have space for code to help buyers?
I think there are many, and - 'AI' or not - a clever tool can generate real value for buyers. Things like mis-zoned units, unreported square footage, signs of buyer motivation, estimates of homeowner equity, etc. can be estimated by good buyers agents but could be made available to everyone for a small subscription.
There is 0.5 mb of data per home in the US. Too much for a person to sift through, typically. We use LLMs to break that down into actionables for families via text message. The goal is to keep things digestable. People can understand if their unit is mis-zoned if you use it in this simple phrasing, but not what a "RM-3 District" is (common type of zoning in SF). The LLM + search abilities can explain a lot - this is how most people learn things about real estate, this process of googling phrases they don't know
> RealPage-defined submarkets identified in Appendix A are relevant markets in which the agreements between RealPage and AIRM and YieldStar users to
align pricing has harmed, or is likely to harm, competition and thus renters. In each of these markets, the penetration rate for at least (i) AIRM and YieldStar, or (ii) AIRM, YieldStar, and OneSite ranges from at or around 29% to more than 60%.10
> Journalists" now love to use a sequence of events and then present them like a causal narrative, ignoring the real reason the final event happened. I'm so freaking sick of it.
When was journalism, writing, or any sort of story-telling not done this way? Is the alternative a series of time stamped actions?
> A small truck has just as much possibility to kill as a big truck, and in fact due to the smaller application of similar forces its more likely to seriously cripple you, even if said hit was non-fatal.
> While I’m all for reducing the size of cars in USA, pretending that ”massive modern trucks” are the only ones that kill is just wrong.
The argument is that a larger truck is more likely to kill - instead of resulting in injuries that can be treated. There is evidence for this.
There is also very obvious evidence that regardless of size you are overwhelmingly likely to die when hit by a vehicle and that in fact due to the application of force via the hood even in non-fatal crashes smaller vehicles have a higher chance to cripple you for life. I really shouldn’t be surprised that saying ‘getting hit by a car, regardless of size, is deadly’ on HN.
>Calculate how much water you actually need to drain. You can use the “100 year flood” values for your locale to get an upper bound on rainfall, then multiple by drainage area
> You're just trying to ban apartments here. Corporate SFH housing basically doesn't exist but mom and pop large apartments don't either.
This definitely isn't true - there are markets where apartments aren't allowed/desirable, and in which corporate actors 'build to rent' entire subdivisions.s
That's rare though. Build to rent an entire subdivision is a risky business because there are so many unpredictable costs involved with owning that many houses, and you have to maintain them even if they're not being rented out. It's much more common to build them and then sell them off so the family owning it has to deal with it. As I said, that's why the average landlord in the US owns two properties.
Apartments have the opposite situation for weird legal reasons; there are almost no condos built in California anymore because the laws allow the original developer to get sued over quality issues, whereas for apartments the landlord has to deal with it.
It is rare in my part of the world, but it isn't (currently) rare in places where roofs don't have to withstand snow. No comment on the historical trend.
> Apartments have the opposite situation for weird legal reasons; there are almost no condos built in California anymore because the laws allow the original developer to get sued over quality issues, whereas for apartments the landlord has to deal with it.
This feels like it confuses a couple of issues - one really needs to know the recourse a homeowner has against a home builder and how long this lasts (and to compare that recourse to what is available to condo owners)
> i.e., gene editing (by whatever means) is not really a win any more than changing bits in memory is. The win comes from having an in-vitro model that replicates the in-vivo disease
I take your point, but this would have been a challenging, expensive endeavor before CRISPR-Cas9. Likely, the perturbation would not have been done in the same paper as the announcement of the locus due to the challenge of deletion work before CRISPR-Cas systems became readily available
> for example the US government wanting to encourage home ownership and skewing the mortage [sic] market and the money supply and requiring lenders to accept more default risk
Are there economists who share your view on this? I don’t see how the issue of repackaging CDS and related products by a financial rating agency has anything to go with the government.
If you’re saying that the government forced buyers to abandon due diligence… I think we will have to disagree about the facts of the GFC
> I don’t see how the issue of repackaging CDS and related products by a financial rating agency has anything to go with the government.
The financial rating agencies are creatures of government regulation.
> If you’re saying that the government forced buyers to abandon due diligence
I said no such thing. I said that government regulations forced lenders to accept more default risk--meaning they were forced to lend to people they would not otherwise have lent to because the risk of default was too high. That's what "subprime mortgages" means, and those were a huge contributor to the crash.
> said that government regulations forced lenders to accept more default risk--meaning they were forced to lend to people they would not otherwise have lent to because the risk of default was too high. That's what "subprime mortgages" means, and those were a huge contributor to the crash.
The ratings agencies are free to rate things as they wish - unless maybe you’re saying there’s a government directive to misrate things?
Also, please clarify how the government is compelling lenders to make loans that don’t pass the lender’s underwriting criteria… this is news to me
It's a half truth; the truth is somewhere in the middle of this and another theory. I'm not an economist, but the start of my professional life was marred by the experience of 2008 so I spent a lot of my time reading about it.
In reality, there were a lot more sub-prime loans but only one of those lenders was actually expected to take on sub-prime loans. That's to say, taking on more sub-prime loans was a choice reflected in an ecosystem of incentives where profits were falling because a few lenders started a campaign to lower borrowing standards and the rest of the herd followed to stay afloat. What also happened was that lenders were essentially over weighting sub-prime loans into these packages and then using their relationships with the privately controlled ratings agencies to rate them the way that would be if they were filled with primes. If you read between the lines lenders found the solution to their profit problem and were trying to justify its stability post-hoc through package ratings. The reality is that sub-primes are highly profitable when they work out because they have high interest rates. When they don't they're not that expensive because generally the property is offloaded but this only works up to a magical threshold depending on a lot of risk variables. Once you go beyond that threshold and the dominos begin to fall, they all fall spectacularly. Risk traditionally should be leveled by packaging them with less risky loans.
> please clarify how the government is compelling lenders to make loans that don’t pass the lender’s underwriting criteria
It didn't. Instead it required the lenders, by law, to change their underwriting criteria so that loans which the lenders would previously have chosen not to make, because they were not within their underwriting criteria, were now within their underwriting criteria. It also passed laws forbidding lenders from refusing loans to people who met their underwriting criteria. This was all done on the theory that encouraging home ownership was a good thing and that lenders had been arbitrarily refusing loans to people and needed to be stopped from doing that.
The book Fault Lines: How Hidden Fractures Still Threaten the World Economy makes a claim to that effect, that US government created some unfortunate incentivizes towards risky credit. It was of course not the only problem that led to the 2008 GFC, but it certainly was a part. I recommend the book, it was detailed and has a global focus, as the author was the Chief Economist at the IMF for some years.
My suggestion is much less fascist/communist oriented. By socialized medicine, I mean fully socialized as in "the only option"... where medical establishments and corporations themselves are effectively govt run, not just regulated.
A state option for "insurance" that covers govt employees, but operates without a profit motive, while having the same negotiating standpoint of any general insurance provider is far different than what one would see in a fully socialized environment.
Edit: It's also from the PoV of continuing to offer coverage at all to those already covered by Federal spending, while avoiding increased costs. Basically better utilizing spend to expand competition, as opposed to limiting it.
Many countries have single payer healthcare systems and that has nothing to do with fascism or communism.
Healthcare is rife with asymmetric information by its nature. Doctors and nurses understand the medical services they offer much better than most patients. And privatizing insurance makes this worse. Now not only is the service itself opaque to patients but even the fees that they will need to pay for it.
But I also appreciate your point of view and it is good that PPO works decently well for you.
Has learning to code generated any benefit to you? What areas do you think have space for code to help buyers?
I think there are many, and - 'AI' or not - a clever tool can generate real value for buyers. Things like mis-zoned units, unreported square footage, signs of buyer motivation, estimates of homeowner equity, etc. can be estimated by good buyers agents but could be made available to everyone for a small subscription.