Until just now, I've been trying to figure out why people think that JSON is necessary in the database? Yes, lots of data is hierarchical, and you just normalize it into tables and move on. The fact that some people don't work this way, and would like to put this data as it stands into a JSON tree hadn't occurred to me.
What problem does normalization solve? You don't have to parse and run through a tree every time you're looking for data. You would, however, need to rebuild the tree through self joins or other references in other cases, I suppose. It depends how far you break down your data. I understand that we all see data structures a bit differently, however.
You gave an IOCCC snippet as an example of a C99 coding trick you know? I mean, the code looks visually cool, but it's funny to explain a code concept using code shaped like an anime character. (At least that's what I think it is.)
Surely that's not what the JAG is referring to since the article covered it
"While the Posse Comitatus Act refers only to the Army and Air Force, a different statute extends the same rule to the Navy and Marine Corps. The Coast Guard, though part of the federal armed forces, has express statutory authority to perform law enforcement and is not bound by the Posse Comitatus Act."
Ahh, I wasn't sure if it did or not, I told the model the file directly in my example "Edit file ABC.php to do..." so I wasn't sure if had that context or just found the file (I also had that file open).
I really do hope they improve this further. Junie (Jetbrains Agent) has much nicer UI. I'd love Claude code with more native UI.
in addition to the diff viewing, the plugin also allows claude code to pick up more context from the IDE!
it should automatically pick up your current selection, as well as any highlight/diagnostic errors (squiggly errors and warnings) that the IDE finds. there's also a shortcut key (Cmd+Option+K) to send your current selection as a @File reference to the prompt.
It's a $20k, street-legal, EV modding platform. Sounds like you can mount your own infotainment system. Just an electric motor, battery, and chassis, and the rest is up to you. Isn't this what we've been asking for?
Yea, it's pretty exciting. I'd like to see how much more they could strip out to reduce the price and still have a viable commercial product. I guess I'm living firmly in the past, but $20K still seems to be a high price for a car. Then again, I haven't bought a car new since the 90s, so I'm probably just an old fart who hasn't grokked what things cost today. I still remember the day when the base-model Corolla started costing more than $9999 and I thought the world was coming to an end.
EDIT: Yep, I'm just old. Another commenter linked to a "10 cheapest new cars" list and there seems to be a price floor of around $20K. No major manufacturer seems capable of making one cheaper!
According to the US Bureau of Labor Statistics[1], $9999 in 1995 is equivalent to $21,275.25 today, so it's a pretty spot on price for a barebones car.
Except, with advances in computational design and engineering, manufacturing automation, and moving to plastic for the body I would expect a reduction in price, in real terms. Not impressed.
>Except, with advances in computational design and engineering, manufacturing automation, and moving to plastic for the body I would expect a reduction in price, in real terms.
Except with all the safety equipment, crumple zones, airbags, sensors, etc. I would expect an increase in price.
And this back-and-forth here is why the folks at the BLS
have a hard job. Both options—
a car in 1990 is a car in 2025 and real value/utility is unchanged and price should be compared 1-1 ignores that cars are actually better now. But at the same time you literally can't buy a new car at 1990's quality so the additional value/utility might not actually be wanted by some and so this in effect makes real price go up.
Some of those $10k cars in the 90s had more plastic in the bodies than cars today, e.g. Saturn S-series, where all body panels below the belt-line were plastic.
It isn't necessarily the cost savings one might expect though, because steel panels can also be load bearing and part of the crash structure, which is not really practical with plastic panels.
Cost savings wasn't the reason for the Saturn plastic panels, IIRC -- they were intended to make the car more durable; they were hard to dent. Some Saturn salespeople would kick the side of the car, hard, to demonstrate their resilience.
With plastic panels, that means they're replaceable. Possibly even swappable (custom 3D printing?). This just adds to the "modding platform" they could be marketing to.
In fact, on modern cars many times these panels are replaced.
If you get a big enough dent in a door, a good body shop will offer to replace the outer skin instead of filling with bondo. They cut the weld on the inside of the door all the way around, take off the shell, and epoxy a new one on. The body shop owner told me that the epoxy is actually stronger than the factory weld.
Yes, bodywork is quite a mature discipline. I was presuming the parent commenter meant user-replaceable, i.e. bolted on.
> The body shop owner told me that the epoxy is actually stronger than the factory weld.
Often this is because the special high strength steels used in vehicles today depend on proper heat treating to attain their strength, and welding can compromise this. Many OEMs even specify panel bonding for repairing particular crash-critical parts of vehicles now because of this.
It's mostly because the factory welds are the result of someone running numbers until they find the bare minimum whereas the autobody guy would rather not risk it.
The OEMs have proper repair procedures that are the correct way to fix the vehicle, and if the autobody shop is reputable, they follow them. And the stated reason OEMs specify panel bonding instead of welding is:
1. because UHSS is sensitive to heat, and robots are much more accurate in how they heat than Jimmy with a tig torch, and they were programmed by a process engineer, where as Jimmy welds until 'it looks good'.
2. welding may compromise anti-corrosive treatments on the inside of inaccessible cavities, which can lead to corrosion issues
In nearly all cases they're faster. 10+ second 0-60 times used to be pretty normal for "regular" cars. Now days, people will complain that a car is slow if they can't put down 7 second 0-60 times. And "quick" boring cars of today are as fast as sports cars of the past.
The 1996 Ferrari F355 Spider and the 2025 Hyundai Elantra N both have a 0-60 time of 4.8 seconds.
The average price of new cars sold in the US last year was nearly $50k. The manufacturers make more money from expensive cars than cheap cars, and people keep buying them, so that's what they sell. Before they canceled the Fit, Honda was selling almost 10 times as many of the larger CR-V each year.
You can find numerous new cars for sale in Mexico for under $15k USD.[0] Even Europe has several new cars under €20k.[1] These are the same manufacturers we have here, but lower cost models that are only sold in lower-income countries.
Most of those models are not real. They exist, but they don't make many of the base model and if you attempt to purchase one, it is a 4-18 month lead time.
I was in the market in Mexico last year, just looking for a cheap city car with a warranty for when I am there. I test drove half the cars on that list, the other half I immediately eliminated after sitting in the driver's seat for under a minute.
You can think of those base models like MSRP GPUs right now. They exist on paper, but good luck actually getting one for MSRP.
In the end, I didn't purchase any of those and got a two year old certified preowned vehicle with the top trim and comfortable seats from a dealer with a warranty for about $3000USD more than the cheapest actually available model of those linked in your post.
I guess I'm living firmly in the past, but $20K still seems to be a high price for a car.
You're not even living in the past. Our 20 year old Scion xB cost us $20K out the door new (granted, that's with most of the paltry list of options added, $15K base). And that was a cheap car at the time, Toyota marketing to "the kids".
The last time $20K was "a high price" for a new car was probably before most HN folk were born.
For those price-comparing, it is $20K after the federal incentives. So, its real cost is around $27K which makes it way more expensive than what the article claims.
One would be wiser to based on annual depreciation in real $ plus time value of purchase price. I suspect out of new trucks a tacoma would be the cheapest since the depreciation is low to negative (IIRC recently a Tacoma was worth more 1 year old than new).
All new car brands/models will not have comps for several years. Even folks buying Rivians, etc have no idea how the resale value will play out so you’re always going to have to take a gamble
That's with federal incentive and likely before they factored in the tariffs. Those 500 parts aren't all coming from US. I wouldn't expect any usable version of it to be below 30k once it's actually available.
This is an $27k car, with $7.5k rebate, so much for the unfairly competing Chinese.
The MG4 is £22500 in the UK before tax, which is about $30k in USD, and that's a full-featured 5 door car, with double the range.
I do applaud the philosphy of cheap and barebones, and easily moddable, but my two cents is that trim is not the thing that's making cars expensive to make.
Is it? They show speakers mounted in the front as a "soundbar". Will people figure out there is a reason cars with good sound systems have them mounted all around the vehicle?
I just want some power ports and good mounting points, then I can put whatever I want there, and upgrade it. I'd imagine that people will come up with a mountable radio kit, like the DIN format radios of old, but with less restrictions.
DIN format radios are still around. My recent-ish corolla's infotainment display is just a well integrated double-DIN. I'm surprised this car doesn't as far as I can tell, have a DIN slot for one.
Yes, they have a nice storage bin right behind where you put the optional tablet mount, but the only option I've seen for that bin is a speaker kit. I don't want a tablet mount or speakers in the bin. I want the left side of the bin (above the controls) to be a double DIN mount.
> Will people figure out there is a reason cars with good sound systems have them mounted all around the vehicle?
No, because they knew what they were getting into when they bought this truck. And I'm sure there will be a dozen DIY ways to add a more traditional sound system.
The interesting modern tools for passengers wanting to DJ are shared Spotify playlists and Apple SharePlay.
A lot of Bluetooth speakers today can fill a car with a sound wall better rear speakers used to. Apple says you just need two of their Bluetooth speakers to fill a room in a house with great stereo and reasonably good surround sound. The square footage of a car is generally smaller than the supported room size.
Samsung Galaxy A Series (A15 / A15 5G, A14 / A14 5G, A25) all have headphone jacks. These are their "lower-end" models, though. The higher-end ones don't. And of course, iPhones haven't for a long time, too. Alternatively, for other phones with just a USB-C port, you could get a USB-C to headphone jack adapter.
edit to add: if Slate is successful, I wouldn't be surprised if a decently sized ecosystem pops up around easily installed custom sound systems and the tablets (possibly with headphone jacks!) to control them.
Thank you! I'm not a huge Samsung fan, their Android flavor doesn't appeal to me much, but I'm glad that someone is still making phones with headphone jacks. USB-C adapters are mostly fine, but I like being able to charge and listen to music at the same time without a clunky adapter, it's yet another thing I'll lose.
Good find, I currently have an aging Pixel 6a (with, regrettably, no headphone jack) but I like Motorola's hardware and their fairly vanilla Android. I may consider a Motorola device, if they provide updates for at least 5 years.
> a comprehensive active safety system that includes everything from automatic emergency braking with pedestrian detection to automatic high beams
No stereo, but luckily they still found space for a few DNN accelerators that will slam on the brakes randomly when getting false detections. Likely still has a 4G uplink and all the modern car cancer to make sure they can datamine their clients as much as possible and offset the subsidized purchase cost.
This made me think: what if my little utility assistant program that I have, similar to your Stevens, had access to a mailbox?
I've got a little utility program that I can tell to get the weather or run common commands unique to my system. It's handy, and I can even cron it to run things regularly, if I'd like.
If it had its own email box, I can send it information, it could use AI to parse that info, and possibly send email back, or a new message. Now, I've got something really useful. It would parse the email, add it to whatever internal store it has, and delete the message, without screwing up my own email box.
I’ve been thinking lately that email is a good interface for certain modes of AI assistant interaction, namely “research” tasks that are asynchronous and take a relatively long time. Email is universal, asynchronous, uses open standards, supports structured metadata, etc.
If you want to get ahead of the curve, look into the Agent-to-Agent protocol Google just introduced. I'm currently using my own custom AI agent assistant to perform life tasks. If I could integrate a better tooling/agents into my own assistant system like your's that'd be awesome.
It's kind of like sure, I could manage my own emails, or I could offset this to someone who does it better. If you do it better and it's affordable, I'm in.
We are on that starship to the future right now and I love it.
I've build adaptive agent swarms using email, mailing lists and ftp servers.
If you don't need to have the lowest possible latency for your work and you're happy to have threads die then it's better than any bespoke solution you can build without an army of engineers to keep it chugging along.
What's even better is that you can see all the context, and use the same command plane as the agents to tell them what they are doing wrong.
I am still very open to this one. An email-based, artificial coworker is so obviously the right way to penetrate virtually every B2B market in existence.
I don't even really want to touch the technology aspects. Writing code that integrates with an LLM provider and a mailbox in E365 or Gmail is boring. The schema is a grand total of ten tables if we're being pedantic about things.
Working with prospects and turning them into customers is a way more interesting problem. I hunger for tangible use cases that are actually compatible with this shiny new LLM tooling. We all know they're out there, and email is probably the lowest friction way to get them applied to most businesses.
I have a couple companies that force me to send them data via email. They have an email template that you have to conform to, and they can parse it. Mainly just very rudimentary line breaks and 'LineItem: content' format. But json in the body should be fine as well. Given the way email programs strip or modify html at times, I would be leery of xml.
Email is decent for intermural communication. If it's intramural and you control both the sender and receiver, MQTT or ntfy are likely better communication channels since they increase flexibility and lower complexity, IMO.
Not if I want it able to have conversations with people, they don't.
I could see installing or implementing a custom client if there were some functionality that'd enable, but "support a conversation among two speakers" is something computers have done since well before I was born. If the wheel fits, why reinvent it?
If you're having conversations with people, then you don't control both ends and email is fine for that. Email is suboptimal for communicating between services/applications under your full control.
Consider the use case from the article: this is a family management support or "AI butler" application. So I control the end with the LLM on it, which I administer - but not necessarily the other, which is anyone in my family, not just me. So unless I want to try to make everyone use my weird custom AI messaging app like I aspire to Bay Area thought-cult leadership, I'm going to meet people where they are and SMTP's cheaper than SMS.
If I'm building myself a toy, then sure, I can implement whatever I want for a client, if that's where I get my jollies. React Native isn't hard but it is often annoying, and the fun for me in this project would be all in the conversation with the agent per se. Whatever doesn't get me to that as fast as possible is just getting in my way, you know?
And too, if this does turn out to be something that actually works well for me, then I'm going to want to integrate it with my phone's voice assistant, and at that point an app is required anyway - but if I start with a protocol and an app that that assistant already knows how to interact with, then again I have an essentially free if admittedly very imperfect prototype.
Under the hoods, is your AI butter one service or many? It would be not-great for your weather or family-event-calendar-management components to communicate with each other or the orchestrator via email.
Receiving an email from the AI-butler rescheduling or relocating a planned outdoors family event because rain is expected would be excellent, using IMAP to wire-up the subcomponents together would not.
Who suggested using email in the service layer? I mean, you're not wrong, but this feels like you handed me a banana and then said I should have picked a better hammer.
We're talking about a conversation that has a human on at least one end, so email makes sense. For conversations involving no humans, of course there are much better stores and protocols if something like an asynchronous world-writable queue is what we want.
"Number of humans in the conversation" wasn't the distinction you initially established, I believe, but I wonder if it's closer to the one you had in mind.
This was the attack vector of a AI CTF hosted by Microsoft last year. I built an agent to assess, structure, and perform the attacks autonomously and found that even with some common guardrails in place the system was vulnerable to data exfiltration. My agent was able to successfully complete 18 of the challenges... Here is the write up after the finals.
For gmail, there's also an amazing thing where you can hook it with pubsub. So now it's push not pull. Any server will get pubsub little webhooks for any change within milliseconds (you can filter server side or client side for specific filters)
This is amazing, you can do all sorts of automations. You can feed it to an llm and have it immediately tag it (or archive it). For important emails (I have a specific label I add, where if the person responds, it's very important and I want to know immediately) you can hook into twilio and it calls me. Costs like 20 cents a month
Mailgun (and I'm sure many other services like it) can accept emails and POST their content to an url of your choice.
I use that for journaling: I made a little system that sends me an email every day; I respond to it and the response is then sent to a page that stores it into a db.
+1 for Mailgun. My only gripe with it is that they detect and block bot activity on their frontend. So if you have end to end (e2e) integration tests built with something like Puppeteer, you can't have them log into Mailgun and check the inbox table's HTML to see that an email was sent. So you have to write some sort of plugin manually - perhaps as a testing endpoint on your website that only appears in debug mode - that interacts with their API.
This might not seem like much of a big deal. But as we transition to more of these #nocode automated tools, the idea of having to know how programming works in order to interact with an API will start to seem archaic. I'd compare it to how esoteric the terminal looked after someone saw a GUI like the one used by Apple's Macintosh back in the 1980s.
I looked forward to this day back in the early 2000s when APIs started arriving, but felt even then that something was fishy. I would have preferred that sites had a style-free request format that returned XML or even JSON generated from HTML, rather than having to use a separate API. I have this sense that the way we do it today with a split backend/frontend, distributed state, duplicated validation, etc has been a monumental waste of time.
> I use that for journaling: I made a little system that sends me an email every day; I respond to it and the response is then sent to a page that stores it into a db.
Yes. I know note taking and journaling posts are frequent on HN, but I've thought that this is the best way to go, is universal from any client, and very expandable. It's just not generically scaleable for all users, but for the HN reader-types, it'd be perfect.
is there a reason you went with telegram and not slack or discord? i was thinking that it could open up a broader channel for communicating with your assistant. i understand you're also just building more of a poc, but curious if you'd thought about that. great work btw :)
I made an AI assistant telegram bot running on my Mac that runs commands for me. I'll tell it "Run ncdu in the root dir and tell me what's taking up all my disk space" or something and it converts that bash and runs it via os.system. It shows me the command it created, plus the output.
Extremely insecure, but kinda fun.
I turned it off because I'm not that crazy but I'm sure I could make a safer version of it.
*Update*: I tried writing a little Python code to read and write from a mailbox, reading worked great, but writing an email had the email disappear to some filter or spam or something somewhere. I've got to figure out where it went, but this is the warning that some people had about not trusting a messaging protocol (email in this case) when you can't control the servers. Messages can disappear.
Other alternatives for messages that I haven't tried. My requirement is to be able to send messages and send/receive on my mobile device. I do not want to write a mobile app.
I'm building something similar and related to the other comments below! It's not production ready but it will hopefully be in a couple of weeks. You guys can sign up for free and I will upgrade you to the premium tier manually (premium cannot be bought yet anyway) in exchange for some feedback:
My one concern there would be edits: a CMS needs to support easily making edits to content (fixing typos etc) - editing existing posts via email sounds like it would be pretty fiddly.
I built up an AI Agent using n8n and email doing exactly this. Works great and was surprised I'd not seen any other place kicking the idea around.
Probably my favorite use case is I can shoot it shopping receipts and it'll roughly parse them and dump the line item and cost into a spreadsheet before uploading it to paperless-ngx.
"I can shoot it shopping receipts and it'll roughly parse them and dump the line item and cost into a spreadsheet" - very difficult to do that without using a vision LLM.
As somebody who lived in Northern California in the communities which first started spiking in the early 2000's, I've always had a pet theory from the time.
Housing Appraisers realized they could tack on $10k per house when appraising, and they realized that nobody would stop them. Realtors loved that because they got more commission, buyers realized that houses were going up fast so they'd better get in while they could, appraisers got called back more in a tight market and everybody made money. There were *no* controls on the appraising.
Still remember how when I bought my house, there was a clause for 'if the house appraises under X, buyer will pay up to Y out of pocket to match the difference' (This was to help make the offer 'stronger' according to realtor.)
Lo and behold, my house magically appraised exactly for X, despite it likely being more like X-5,000 based on realities and other sales in the area...
At least in my state there is no incentive for appraisers to do this. 99% of appraisals get assigned to random appraisers. There is no such thing as “repeat business”, at least within the realm of your typical home sale where the buyer is financing through a bank.
In California, at least at the time, Realtors called the Appraisers they knew could be a little generous with the numbers to make the clients happy. Speed-dial. Inflated numbers were easy to make plausible, and as time went on, the cycle became self fulfilling.
Appraisers knew this, they got lots of easy business for an afternoon's worth of work, and grew their businesses. Realtors would shrug and say "That's what it appraised at." Banks were happy. Sellers were happy. Lots of money. There was no natural regulation or push-back stopping any of this.
Source: friend made lots of money doing this at the time. They probably made out better than the Realtors.
Yea I believe that. In my state (Georgia) the randomization rules only came about after the '08 crash. I come from a family of appraisers, oddly enough, so I have an unusual amount of insight into the industry for a simple software engineer. Prior to the new rules my family's company had a set of clients (banks) that they would get business from, and they had to reach out and do marketing/sales/shmoozing to get new clients, like any other service business.
After the new rules, banks just bid for an appraisal into a black box and it gets fulfilled ~randomly. The family rolodex became pretty useless. So the playing field was leveled, and it's certainly a fairer process with better overall results for homeowners, but it also kind of neutered the whole appraisal industry since there's not really a good way to compete anymore.
Kind of going on a tangent here, but the appraisal industry is one of those "silver haired" industries that is not able to replace it's older workers who are retiring. It's unclear what the future holds for appraisals, but it seems inevitable that there will be some sort of pivotal change in the industry in the next decade or so.
> Appraisers they knew could be a little generous with the numbers to make the clients happy
This can only work if only very few appraisers in the area are overvaluing. But if that works, why wouldn't all of them get in on the game? And if most of them do, it falls apart because the actual sale prices will be out of line with the appraisals. So, it doesn't really work outside of edge cases here and there.
There are definitely laws and lender requirements that should make appraiser selection somewhat random and their findings somewhat neutral, but based on my house buying experience I don’t believe for a second that they’re effective. We made a bunch of offers and somehow the appraisal was always right at the offer price or ~5-7k above. I’m convinced that the appraisers somehow have knowledge they shouldn’t and covert/indirect mechanisms exist to motivate them to play ball to make deals go through.
Besides small sample size giving me a skewed sample, the other explanation I can think of for this is that appraisal is a fairly exact science and realtors have mastered pricing based on it. Considering the volatility of my market, lack of comparable sales, and the IQ of the realtors I’ve met that seems laughable.
> somehow the appraisal was always right at the offer price or ~5-7k above.
Yes, this is pretty normal. Contrary to popular belief, appraisers don't have any sort of special data or processes that allow them to determine the exact value of a house in any given market (because such a value does not exist). An appraiser is working for the bank and simply serves as a risk mitigation officer. Their job is not to answer "what is this house worth?", it's to answer "is the deal you're lending money on within reasonable bounds?". So when the appraisal value comes in at or around the sale price - it's just a simple "Yes". And when it comes in somewhere else, it's a "No".
There is generally zero incentive for an appraiser to inflate prices (today, this was not always true in the past).
The housing market is vast and complex, without question. And still, the reason that prices go up is overwhelmingly the simple fact that buyers are willing and able to pay those prices.
The banks got greedy, realized their mistake, and then promptly blew up the market on purpose. The secondary effects were not the drivers of the crisis. This is one of the most documented and unpunished crimes of the century.
The best part: it may well not have been illegal. The economy is a constant cycle of upswings caused by new forms of crime, followed by a crash and then new laws. They then set out to research new forms of crime.
There was lots of legal activity that added to the size of the crisis. You'd be naive to believe that everything else was perfectly above board.
> economy is a constant cycle of upswings caused by new forms of crime
"The economy" as a concept is not. The current US economy certainly does seem to be dominated by criminal activity and has been for 30 years or so now. The connection of the Internet and financial markets were not a strictly great idea.
> followed by a crash and then new laws.
Laws also get repealed. Like Glass-Steagall was.
> They then set out to research new forms of crime.
I posit that they don't. They enjoy a monopolized economy in an incredibly deregulated market. They take whatever they can take until they get caught. Our justice department happily negotiates a settlement and sweeps the whole thing under the rug.
This isn't just a clever scheme. This is nearly complete and total government and business corruption.
Taxes are suppose to keep this from going out of control. but the only way to appeal a tax appraisal increase (here in Washington) is to use OTHER home values in the area that are similar, but less.
Mortgage appraisals are far higher than the local tax appraisals which seem to run 20-25% lower than say Zillow says. Local tax appraisals are also typically only updated once every 10 years or so (depending on the jurisdiction).
Besides, I think these days the appraisers don't actually do appraisals, they just look at Zillow.
The people that do appraisals for property tax assessment reasons may be the most disliked people in municipality. My opinion is that it makes no sense for people to be doing it anymore. It should be strictly algorithmic with very few input variables.
I kind of wonder if there could be some kind of antagonistic pricing checks, by finding parties opposed to inaccuracies.
For example, actuaries have to be accurate about the costs of using a vehicle. If their estimate is too low, the insurance company will lose money. If they are too high, they will lose business to competitors.
But probably insurance companies are not the opposed party, because they will sell more insurance, and losses may be lower than inflated costs.
On a different note, california prop 13 has made people keep their house longer or forever.
Also you don't need a proposition in your constitution to affect the same outcome. Where I live they almost never do reassessments, so we effectively have the same result.
The great thing about having states is that we can experiment with lots of different models of how to run society. If you don't like your particular, society then you can move to a different one.
Residential zoning laws in most California cities (and frankly, across the country) do no favors either. They're tuned tightly to encourage home values to go up not to stabilize the cost of housing and maximize housing accessibility.
This strictness is the backbone of what allows everything else to happen. There's no market dynamic - because cities (or more specifically, city residents) don't seemingly actually want a functional real estate market for a variety of reasons - therefore there is no way for enterprising developer to come in and build and sell at a price target they look to set.
Instead, it all has to go through this machinery as you describe. It has always been ripe for exploitation as a result.
Laws of geometry have nothing to do with it when zoning laws and other regulatory hurdles are the real barriers. Real estate is an in incredibly artificial market. It has none of the characteristics of a regular marketplace
Appraisers don’t set prices, though. In my purchases and sales, they’ve only been involved after an offer is accepted and the lender wants to be sure that their collateral is worth what it needs to be. Even if some people are using appraisers to set their asking price, that doesn’t mean buyers will offer it.
The network effect of the prices going up seems to permit the appraisers to add their bumps up as well. I don't think it's as influenced by appraisers as some think, and it might vary by region a bit. If prices in a region are rising (for a number of factors) I don't think many appraisers are going to round down in their judgements.
Sure. There’s no objective value of a house, so they have to go off of recent sales of similar properties. If prices go up, appraisals go up. But I don’t see how the reverse would be true, as suggested above.
Appraisals also act as a gate for prices due to lender requirements. A bit round about, but I can see how looser appraisals can enable inflated prices. Imagining the opposite extreme is interesting: What if appraisals never returned with higher prices than the last sale of that house? Some markets would see increases from people paying the difference out of pocket, but I’d guess the rate of price increase would be much lower. (plus other effects, of course).
The value of a property is whatever it can sell for. Models and assessments attempt to figure out that number without actually performing a transaction. An accurate model will account for the same factors a buyer will use to decide what to offer, but ultimately it’s down to what people will pay. If that number goes up 2x, there’s no model that can tell you that increase is somehow incorrect, or that the true value is lower.
I've explained this before but the players involved don't really care. The bank just wants to make sure they aren't on the hook for a worthless shack. If the appraisal comes back inflated by 10-20% that's fine. The goal is to prevent loaning $1.4m against a house worth $700k.
I mean this is true, but I think its more realistically like the appraiser doesn't want to create a situation where it tanks a sale if they can avoid it. When I bought my house in 2022 there was a bidding war (like every house sale in Seattle), I paid 150k over asking price which was relatively speaking pretty reasonable as the other houses I bid on went for 250-500 over.
That said it was on the very high side of valuation. My agent told me if the first appraisal wasn't enough to cover the purchase price, we would just get another appraisal. The first appraisal went fine, compared to other properties it was within range of reasonable so they approved it and the mortgage went through without issue.
I didn’t realize how close LLMs are to the old magazines. Let it give you a seed, then use that springboard to learn everything else.
reply