> An example from another industry: when I worked at a small chip startup, we had in-house capability to do end-to-end chip processing (with the exception of having its own fabs), which is unusual for a small chip startup. When the first wafer of a new design came off of a fab, we'd have the wafer flown to us on a commercial flight, at which point someone would use a wafer saw to cut the wafer into individual chips so we could start testing ASAP. This was often considered absurd in the same way that it would be considered absurd for a small software startup to manage its own on-prem hardware. After all, the wafer saw and the expertise necessary to go from a wafer to a working chip will be idle over 99% of the time.
This paragraph from the article reminds me of the recent HN discussion about resilience. In this particular example idling 99% of the time is cheaper, but even if idling 99% of the time is slightly expensive, it might still be a good idea not to outsource due to resilience. What if the vendor goes bankrupt? What if the vendor's vendor is unable to perform their contractual duty due to a disease outbreak? Or a geopolitical risk? You avoid these risks if everything is in-house, so these risks need to be priced in when determining whether to outsource. But more often than not, these risks are priced at exactly zero.
What if the saw breaks, what if the person who knows how to use the saw leaves?
There are good answers to these problems, but I find folks who find comfort in outsourcing assume it's easier for the downstream org solve them.
Personal experience suggests it's often a false sense of security, and you often have to have your own experts anyway.
Saw the same in software. We won't get budget to patch so let's outsource. The outsourced entity also doesn't patch, because there's no business case approval to do so. The problem is usually tech is ran by finance.
A lot of outsourcing decisions seem to be made on the premise of "what if it breaks?" It's a terrible fallacy.
Your wafer saw (or server) that is utilized 1% of the time is going to need a lot less maintenance than a shared wafer saw that is used 16 hours a day. You will actually experience a lot less downtime by investing in your own wafer saw (or your own servers) rather than outsourcing (to a foundry or a cloud provider).
I think the real draw of outsourcing is diffusion of responsibility: when it breaks, you can blame someone else.
At work, I wanted to use google sheets for our schedule because it just works: can have view only links, and it handles multiple editors at the same time gracefully. Got asked “well what if google goes down?”, so back to our problematic system of excel sheets it is :(
As much as GSheets is great software in theory, "just works" is something I'd never use to describe it. Google is the absolute worst when it comes to keeping stuff working outside of Gmail, Search and obviously Ads. They will silently drop functionality and randomly break things without any recourse.
Just a week or so ago, the IMPORTHTML/XML and related functions in GSheets just...stopped working. There are Reddit and Google Helpdesk threads about it so it's clearly a global issue. As far as I know, no fix has landed yet.
I had to spend a weekend writing a bunch of scripts to reimplement importing functionality so an international competition hosting more than 600 people from 28 countries wouldn't collapse on the spot. Everyone on the team, including me, was initially glad we were using GSheets as it made our work much easier - until it suddenly made it much, much harder.
That may have been a wise person! "What if Google takes us down" (rather than goes down itself) is a valid business risk to at least consider.
Google & co are doing their best to remove themselves from a rational person's critical path. They're too big, you're too small; they're full of AI and politics, you have no recourse.
if someone had told me that "full of AI" was going to be an insult one day, I would have said they were too cynical about the glorious AI future, but in hindsight perhaps you are right
I experienced the opposite when Google actually went down and all of our email, docs and information went away for the day. Then one day our Internet went down and then the company bought a Microwave internet link and an off site server.
> You will actually experience a lot less downtime by investing in your own wafer saw (or your own servers) rather than outsourcing (to a foundry or a cloud provider).
Analogy doesn't work; "in-sourcing" of previously-IaaS resources as a cost-cutting measure, is used almost exclusively for "base load" (vs. elastic load, which is the comparative advantage of IaaSes); and so your in-house servers are usually running pinned at 100%. So the hardware components will wear out just as fast, if not faster, than the IaaS's servers will; and when they do, your ops team won't have as many spare parts on hand as the IaaS does; nor the ability to live-migrate the enforced VM to another exactly-equivalent substrate host in the fleet (which was already warm) to avoid downtime altogether.
Also, apart from that, there are economies of scale in reliability. Your in-house backups are probably just a ZFS pool or a RAID5+1 array or something. They're certainly not hosted within a 17+-copies Dynamo-ish system, the way most IaaS object storage (incl. archival storage) is, because the costs of doing that in-house are ridiculous.
Rarely using a tool risks it rusting/rotting/obsoleting on the shelf, and if it is expensive you are paying a lot of depreciation. You buy a chraler consumer version that breaks, or an expensive professional model that depreciates. The time share model makes sense.
>> There are good answers to these problems, but I find folks who find comfort in outsourcing assume it's easier for the downstream org solve them.
IMHO a lot of outsourcing decisions are based on the idea that someone else can do it better than we can. This is demoralizing and will eventually become a self-fulfilling prophecy as the most competent people get frustrated and leave. I suspect it may also be projection from managers that are in over their head or don't understand what their own people are saying. And then of course there are times when it's true that someone else can do it better.
IME the most successful outsourcing arrangements revolve around scale of work, not substituting an entire department. Keeping the amount of in-house experts lean and having them manage outsourced resources. So it is more of a tool for someone internal who has the needs organized rather than a dependency on a 3rd party to know what is best for you
> Saw the same in software. We won't get budget to patch so let's outsource. [...] The problem is usually tech is ran by finance.
Same. Where actually, the budget is right there, it's just a matter of choice and priorities: do what's right now or invest 10x in it later because you didn't take care of it when you ought to.
I've been on both sides of the argument: "budget" is a magic word for those who want to skip responsibility for their choices.
Sometimes the budget is yours sometimes you're sitting in meetings upon meetings with someone else deciding there's no immediate ROI so no budget for you.
So sure, the company is always to blame, but unless we're arguing the individual was not convincing enough, the person citing budget may not be choosing alone where to allocate it.
On the bright side, inefficient companies leave the market open for the rest of us. :)
This. For every short-sighted decision made by the finance overlords, I've seen ten failures on the tech wizards' part to plan for their own future well-being. Laziness keeps them from even making a case.
As someone employing a lot of engineers right now, one of the problems is that engineering will just keep requesting resources ad infinitum if there’s no budgetary limit. The constraints of the business are actually always financial, and without proper financial modeling you’ll cease to have a business.
This is true in process scaling no matter the process.
Also the blame. On prem outage? IT management is in the hot seat for a fix. AWS outage? Well, major us-east-1 outages make the news so the blame is more diffuse. And even outside of that, it's much harder for anyone internally to blame IT management for the outsourcing, so long as they were smart and got the CFO to take credit for the cost savings of going cloud or something.
HN often underestimates how much of enterprise IT, and large business generally, is about making sure that upsides are concentrated enough your management chain can take responsibility, and downsides are either diffuse enough or deflectable enough that your management chain can avoid responsibility.
Then you go out of house until the saw can be repaired or you can recruit a new employee to run it. If it's really to your advantage to have the capability in-house, no one will bat an eye.
> In this particular example idling 99% of the time is cheaper, but even if idling 99% of the time is slightly expensive, it might still be a good idea not to outsource due to resilience.
Or in some cases it might make sense to become an outsourcing destination. Your otherwise idle equipment & skills become a revenue stream rather than a cost when your need it low, your local equipment and skills are tested & re-enforced which reduces the risk of finding them unexpectedly wanting when you turn to them in those 1% times, and you might end up with a small controlling advantage in your market (you get priority over any competitors that outsource to that part of you). It could become a significant arm of your business & profit. Think Amazon farming out its network resources.
If you go through all this trouble, you have essentially added another business line to your company. This may or may not be where a company wants to spend their resources. Rolling your own XYZ will always be cheaper than paying a 3rd party vendor, _if_ you have the time, experience, money, and competitive advantage to do it correctly, which are pretty big iffs.
I think it's that you don't notice things that work well. House foundations, light switches, filesystems, silicon manufacturing, water delivery, grocery store logistics, etc etc etc.
These are all things that most people never notice because they just work. It doesn't even occur to people day-to-day that these things can fail.
House foundations, light switches, and water delivery, along with the professions that install them, are all regulated and licensed. There is somewhat of a trend for the quality of those things to have regional variation, e.g., lower quality in places that have historically had weak code enforcement and weak unions. And yes, the regulation probably did make those things more expensive.
The filesystems example given by OP is an interesting counterpoint - Linux filesystems are the opposite of regulated, regionally varied, and expensive.
I think "mature" is the more correct term. It's not the regulation, it's just that it's been vetted and repeated so many times that it has become rock solid.
This is why I never select cutting edge tech for our company - unless it's part of our area of expertise/innovation.
That's also true. Filesystems do have an advantage of being testable by millions of people, relatively stable from one user to another over the medium term, and at least the experts share their experiences. Also, there are no gaps in the regions that benefit from good filesystems. Regional enforcement means spotty enforcement.
As for expense, the reliability of the filesystem is free up to a certain point. There are system failure modes that have to be covered by hardware and admin expenses, such as decentralized backups (just one example off the top of my head).
I have two things in my possession that are wonders of engineering. One hair trimmer and one restaurant kitchen blender/kitchen multifunction machine. Both made in 1987 in East Germany.
The blender I bought second hand, and it came with a box of original replacement parts and specifications so detailed that I believe I could probably replace the motor if it ever fails (almost 6hp! Take that, Vitamix!). It makes nut butter in no time.
The trimmer could probably be used to cut down trees.
I love these machines, and if they ever fail I don't thing I could ever replace them with anything of similar quality. They are 35 years old. The electronics in the kitchen machine still look pristine. I keep all bearings well lubed. It runs like a Swiss watch. Only very loud.
is this not the very definition of survivorship bias though?
I have things from the 80s that, obviously, have lasted a long time. But I have owned things from that era that have failed and been forgotten about.
Incidentally I bought some boots this year that are expected to last (at a minimum) 10 years, and I suspect they would, but I can't know that until 10 years from now.
Heirloom quality is probably still a thing, but only discoverable when an item actually becomes an heirloom.
The above author mentions that these machines are from Eastern Germany. As I am from Eastern Germany myself, I can tell you that we had regulations in place, that machines had to last and had to be repairable. Those regulations came into place, because we had heavily resource problems.
Funfact: Those type of regulations are heavily discussed even these times again, when I look at EU right to be repair, or the discussions specifically around John Deere and the right to be repair.
A nice trivia fact I learned is that East Germany made the world's best selling digger. There's a fun documentary (in German) that covers this. The machine from the DDR is discussed starting at 27:53
Not necessarily. I also have an East German blender, that's like 40 years old, and looks like hell, but works perfectly. I bought another one, since the old one was kinda hard to look at. It was a highly recommended somewhat upmarket type from a supposedly reputable brand.
It broke within 2 years..
I took it apart and discovered it was full of plastic gears on load bearing components which predictably got annihilated by wear and tear. The old one had metal ones.
I echo the article's sentiment that while cheap usually means bad, expensive stuff is usually indistinguishable in quality from mass-market stuff nowadays.
Probably. I don't have ten of these machines, so it is hard to make a general statement.
I mostly meant it as a comment on what the parent said; I could pay four times the price for something and have it last 35 years (which is 15x longer than the blender I had before it) I gladly would.
The fact that I can repair most of it is also a thing I miss in the things I buy today.
> is this not the very definition of survivorship bias though?
Not necessarily; it's also sufficient to consider the overall distribution of blenders and hair trimmers from the 80s that are still working compared to the distribution of items sold, which means it's also fairly easy to spot because people use their kitchen/bathroom appliances frequently and notice when exceptionally old ones still work.
If I could buy products that just work and don’t break, even if they cost 3x as much, they would pay for themselves with longer lifespans and less wasted productivity.
I replaced my 21 year old Miele washing machine recently. The solenoid-driven water intake valve broke. It wasn't a part of the machine itself, it was on the water intake tube. But the replacement part was nearly half the price of the brand new replacement Miele washing machine I ended up buying instead. The new one is slightly larger and has higher capacity, and doesn't suffer from a little design problem the original had, so I was ok with getting a new one. But yes, the original washing machine worked exactly as when it was brand new - it was just that intake valve on the tube.
Just like a quality automobile, to bring the discussion closer to the article. For example, a Tesla Model 3 might cost more than a Hyundai I10, but after three years of ownership you've not had to change the oil, fill it with gas, get the tailpipe emissions inspected, possibly see the catalytic converter stolen, etc. And the difference in maintenance only grows from there, when the plugs and timing belt and seals and transmission and other items wear.
Lots of products have more expensive buy-in buy are cheaper in the long run.
I drive another manufacturer's equivalent of the Hyundai you mention, their lowest end car that they actually quit manufacturing, and I am sure the Hyundai is in the same ballpark.
I change the oil about twice a year. Over the past eight years, that has cost me about $480.
I have changed the transmission fluid three times. That has cost me about $140.
I've changed the air filter a couple times...not very often. Let's be generous and call it $40.
Around 120k I replaced the spark plugs. They were weirdly expensive, costing about $100.
I have performed no other maintenance.
I have filled it with gas about once a week. It's a pretty small tank. I'd estimate about $6000-$8000.
Add all that up, pretend electricity is free, and I have still come out way, way, way ahead versus buying a Model 3. I've still spent less than half of what I would have on a Model 3.
I hate to sound like I'm advocating for Sunk Cost, but this helps put numbers to a feeling I've had:
The current cost-over-time of my (and apparently your) ICE car is not meaningfully high enough to want to 'upgrade' to an all-electric. Plus, when mechanical problems eventually arise, there's already a somewhat-independent parts and labor infrastructure to lean on to get it back on the road.
On a personal level, I'm also opposed to giving Elon more money, and opposed to the idea that my car may brick itself with an auto-update. I also got a promotional 0% loan on my car, have zero intention of paying that off early, and can see this car lasting until 2028 or 2030. Unless I have kids, this car should be fine for my needs until then.
Thank you for the counterpoint. I concede that my idea of what a modern motor vehicle is is outdated. I'm actually happy to see that things have improved so much.
Ehm, doesn’t Tesla have famously bad quality control in several areas, e.g. fit of body panels? EVs, as a category, need less maintenance - but Tesla’s are not an example of a trouble-free product!
Sure, the early Model 3's had body panel fit problems. I actually bought one last week, and the fit is amazing, both interior and exterior. I don't know why that point keeps coming up. For what it's worth, I used to be a tech at a Ford dealer, so I know what bad fitment looks like!
Initial news always spreads further and faster than updates. This is normal and should have been expected by the company. They took the risk of rushing and get to suffer the consequences.
Is there any car company that has not had recalls? Toyota had airbags exploding with metal fragments, killing drivers. Yet that recall is nowhere near as well-known as Tesla body panels not lining up.
How many miles are you putting on that i10?! An oil change is every 2 years or 15k km. I think it’s a bad example btw, it’s a famously reliable car (the taxi of choice in Bogota fwiw)
For some reason people in the US are obsessed with changing the oil in their vehicle.
Modern engines & oils don't need the oil changing every 3000 miles, but the folklore belief in the necessity of doing so rolls on unstoppably in the USA.
A factor is oil change shops still put a sticker on the windshield that says to come back in 3000 miles. It's of course in their interest for people to change their oil too often.
I have both, two modern cars with services all 30k km and one from the 80s with motor oil changes every 5k, gearbox oil every 10k and axles every 10k as well. Not to mention that modern gearboxes and exles tend to be greased for life.
Thanks, I updated the post. I had no idea that oil changes are now once every two years. On the wife's Subaru I still change the oil every six months, and I remember when the standard was actually every three months. She puts about 10 km on the car during those six months, we live in a hilltop village half an hour drive from a city.
Time based oil changes (as a backup to mileage) in gasoline vehicles are mostly to address oil contamination from the engine not getting hot enough to boil off fuel and combustion byproducts in the oil. It's a rough heuristic for drive type - more advanced oil life monitors know the actual drive cycles an engine is seeing and will adjust appropriately, but for a basic time-and/or-mileage schedule, it's more about picking some interval that gets the oil changed before the additive package is destroyed by combustion acids. If the car is driven only short distances and never gets a chance to fully warm up, six months is probably a reasonable interval. If it gets a monthly spin on the highway for an hour, two years is probably fine. Ideally, you'd be measuring total base and total acid levels and calculating change points based on that; this is common practice for large truck engines but for a gasoline car engine, an oil change can be cheaper than the lab tests. (I'm still in favor of having oil testing done on at least some changes, it's basically "routine bloodwork" for your car and can detect many problems early.)
Another upshot of this is that with rarely driven vehicles, you might as well use the cheapest oil which meets appropriate specs, because your oil changes are driven by additive depletion and oil contamination rather than breakdown of the oil itself.
> Another upshot of this is that with rarely driven vehicles, you might as well use
> the cheapest oil which meets appropriate specs, because your oil changes are driven
> by additive depletion and oil contamination rather than breakdown of the oil itself.
The Ferrari gets the cheap stuff, the Kia gets the good stuff. I love how some things are so unintuitive.
I just ordered an oil testing kit from Blackstone. Thank you for the idea.
What? Oil change every three month? WTF is wrong with the US industry?
The very same cars in Europe will have an oil change every 2 years or so. My Volvo S60 Model 2000 has an oil change every 3 years.
There must be something that you get sold on by marketing that is so wrong. BTW. Also a Tesla has oil in their gearbox, even if it is just a single gear integrated into the single unit combined with motor and inverter. For a Model S it is recommend to replace it every 12.5k miles, which is exactly the very same recommendation as with other car model.
Without meaning to argue that it is necessary, it persists because it just isn't that big a cost. Gas isn't the largest cost of ownership and 5,000 miles of cheap US gas is still ~10x the cost of an oil change.
> For a Model S it is recommend to replace it every 12.5k miles
There was a recommendation to change the gear oil previously, but newer versions of the S removed those. Even then I seem to recall it was way longer than 12,500 miles.
Your vehicle should generally be serviced on an as-needed basis. However, Tesla recommends the following maintenance items and intervals, as applicable to your vehicle, to ensure continued reliability and efficiency of your Model S.
Brake fluid health check every 2 years (replace if necessary).
A/C desiccant bag replacement every 3 years.
Cabin air filter replacement every 3 years.
Clean and lubricate brake calipers every year or 12,500 miles (20,000 km) if in an area where roads are salted during winter
Rotate tires every 6,250 miles (10,000 km) or if tread depth difference is 2/32 in (1.5 mm) or greater, whichever comes first
Seeing as how the recommended cleaning and lubrication of the brakes is 12,500mi, I'm wondering if you heard that and got confused about the lubrication needs. As mentioned in their manual, this is really only recommended for places where they salt the roads in the winter. Since the brakes don't get used as much its even worse than a regular car with the corrosion. If you're not in a high road salt area, its not a problem.
The recommended service interval on other EVs also have extremely extended oil change intervals. The service life of the Mach E's gear oil is 150,000mi and the coolant change is at 200,000mi.
I'm not in the US, but I have lived there and I did learn that frequency there. Like I mentioned in another thread, I worked as a service technician at Ford and that was the recommendation then (1990s).
As for the Model S fluid change, thank you for mentioning it, I did not know that. I believe that there is no fluid change interval defined on the Model 3, which I just recently bought.
There are many gasoline vehicles that specify no fluid changes for certain parts, infamously, transmissions. Longer lasting synthetic fluids allow them to advertise lower maintenance costs, but they still do wear out. They’d rather the transmission wears out after 150k miles anyway.
Don't get me started on transmissions! A failing transmission the month after a Model 3 test drive / joy ride is what really pushed me into making the jump to electric.
Thank you for the advice, I will research the service intervals on the Tesla.
Because everybody drives differently, mileage and time based recommendations aren't the best. Some places will offer to test your used motor oil during a change and will tell you if you waited too long or changed it to early.
The i10 is a rather low end car, but the upgrade could be worth it. Finland gas prices are 8.464 USD/gallon right now. That’s unusually high but assuming 200k miles at say 35 mpg that adds up to 48,000 USD in gas over the lifetime of the vehicle.
Not that electricity is free or that we have long term cost data, but I suspect the EV / plug in hybrid premium is a very easy choice in Finland.
The average car life in Finland seems to be 15.6 years (let's round that to 16, helping the case for EVs).
The average distance traveled by Finnish cars, based on 2018 data, is about 13600km per year (let's make that 14000km per year, also helping the case for EVs).
That works out to 224k km over the lifetime of a car. That's only 140k miles.
Regarding your 35mpg (was 30mpg before the edit), that's 6.7l per 100km (7.8 before edit), let's make that 7l per 100km (and 8 before edit), further helping the case for EVs.
But in reality a small car such as the i10 probably uses more like 5-5.5l per 100km (40-45mpg).
Then the gas price is super high due to the Covid economic bounceback coupled with the supply chain issues plus the latest conflict.
I'd say your numbers are too optimistic plus they're for the entire lifetime of the car. The average owner probably has the car for half that duration, at best. So they don't really care about the entire lifetime.
Still, things seem to be getting closer than 10 years ago, for example.
The really bad thing is the upfront price. There's no comparable EV in the i10 price range, which is a very cheap car (we're talking about a car around €12k).
Finland imports a lot of used cars but: “An average car in the fleet 12.6 years of age” that suggest the car actually lasts to ~25.2 years. It’s not that simple because again they are importing and exporting used cars.
there is now! I'm seeing the Dacia Spring appear everywhere - probably because it starts at only 13k in France including the ecobonus (I think base price is 16-17k). This is an absolute game-changer
I've got a 10 year warranty on the battery pack. And at the rate that independent battery servicing is expanding, there is a very good chance that independent pack refurbishment or replacement will be affordable by March 2032.
In any case, the wife's 1996 Hyundai was worth more to the breaking yard than it was to the second hand market when it's transmission failed in summer of 1997. So even if the battery pack fails immediately after the warranty period and that totals the car, I've still come out ahead in total cost of ownership.
Semiconductors are not regulated except in terms of large scale externalities - e.g. water quality, air emissions, etc. Otherwise they pretty much use any methods, processes or means that they want. They do have economic incentives based on the constraints of physics - you can't boost defect densities of wafers by screaming at Mother Nature so to assure a market for your products, there are constraints on quality and process that all things equal you'd skip since they are "cost centers".
B) There are regulations almost everywhere (especially, almost everywhere that has a profitable market to sell into), so it's probably easiest to just sell regulation-compliant equipment in the rest of them, too.
So it's not at all "preposterous" that the regulations actually work even in most places where they're not legally in force, and that you've just not noticed the few where they don't.
You have absolutely no idea what you're talking about. I live in Alabama where there is virtually NO CODE, and NO ENFORCEMENT, and there definitely aren't any God damned unions, either. My house foundation, light switches, and water delivery are perfectly fine, thanks. Matter of fact, I did the wiring and installed the well pump and a fair bit of plumbing myself. Didn't have to ask anyone's permission to do it, either.
Wow, imagine that radical concept--a society that functions well without government goons breathing down everyone's neck! Maybe we could even come up with a word to describe this amazing new idea. "Freedom" maybe? "Liberty"?
How profitable do you think it is to make special non-regulatory-compliant faucets and light switches just for the Alabama market? Probably not very, so manufacturers just sell the same regulations-proof equipment there that they sell to the rest of the US. In effect, you're freeloading on the regulatory work of non-Alabamians.
Enjoy your working non-lethal infrastructure, parasite.
These are all things that most people never notice because they just work.
Taking the example of grocery store logistics, the number of times products are unavailable in my local store makes me thing that's a thing that doesn't "just work". It's something that breaks down regularly, and possibly has lots of people working hard to keep it from breaking even more often.
The same is true for lots of things. Stuff like water delivery and silicon manufacturing doesn't break all the time because lots of people are fighting to make it work, and are actively maintaining it all the time.
I think it's possible that most things don't "just work", and we're just fortunate that there are teams of people out there stopping us seeing the effects of all the failures.
That's no contradiction. Logistics and manufacturing works because people are spending their professional lives maintaining them. It's the outside that doesn't see these efforts, for them it "just works". Like electricity.
> And it also pretty much sums up how most people in Tech have minimal understanding of Supply Chains and logistics works. Even distribution alone, within a single country ( ignoring the cross border logistics ) is complex enough.
Let me tell you supply chain and product availability in store ( especially grocery ) is still an unsolved problem. For a lot of different reasons and market dynamics. But mostly because grocery stores also have their own brand which compete with other products, and sharing sales data for better forecast is still a big no no. Compare to let say Smartphone, your average retail store will have zero chance completing with Apple or Samsung. So every time an iPhone is sold Apple knew instantly and can better manage their supply chain. Both domestic and international.
If we didn't had COVID and Chip Shortage, most of the world still doesn't give any credit or importance to Supply Chain management. Even though it is the basic fabric of our society. And that is speaking with experience working with Fortune Global 500.
Are those people optimizing for having every product available at every time? They have to balance against the very real cost of spoilage, so I don’t think they consider the occasional out–of–stock as the system breaking
Yes they are. Product availability is the most important factor in choosing a grocery store, as unavailability makes people need to visit additional stores or change recipes/plans on the fly. Some spoilage is factored in and is just a minor bit of negative PR. However, the supply chain disruptions since 2020 are too big for any grocery chain to "solve".
For some products like pasta and canned tomatoes, you can hold enough stock to deal with a 99th-percentile day without any wastage at all; if it doesn't sell today, it'll sell tomorrow.
But for those little packaged sushi snacks with a one-day shelf life? Any overstock is going in the trash at the end of the day.
And sushi snacks mostly sell to workers on their lunch breaks. You'll see big fluctuations in demand if a nearby office changes their work-from-home policy, or has a big all-hands meeting that gets everyone in. Even the greatest demand modelling can't predict such things, as nearby office meetings aren't available as a model input.
Some products are also easily more easily substitutable than others: If the 1kg pack of mid-priced spaghetti is out of stock, maybe I buy the low-priced brand, the premium brand, the 500g packet, the wholewheat version and so on.
>Product availability is the most important factor in choosing a grocery store
For many people, price is almost certainly at least as big a factor. Many, perhaps even most, people are willing to accept things being out of stock now and then for 10% lower prices.
Look up "Things seen this week during structural inspections!" on imgur. Some truly horrifying stuff from that person.
For some of these foundations to still be standing and building occupants not to notice anything's wrong ... I can't even imagine how much safety factor is built-in. If we built software with those margins, nothing would ever ship.
I'm planning to buy a house in the next year or two. I would 100% hire this guy as the inspector if I knew who he was. Those photos are more effective than any marketing.
As usual its just how much money you put in to it. We spend a lot of money making sure building foundations and silicon manufacturing works because failing is expensive and dangerous. I don't want to pay double/triple price for a toaster to slightly reduce its risk of failing because I'm happy to accept that on average it lasts a long time but there is some chance it fails sooner. If I'm in charge of buying a $100M building, you bet I want to pay extra to assure it will not fall over.
There’s a lot of stuff that works remarkably well even though it’s cheap. I just came from a supermarket. It’s filled with items from around the world, of which most are very inexpensive. The consistent quality of these products is astounding—a bag of potato chips or a box of crackers tastes exactly the same, anywhere in the country where I buy it, year-round. A can of Coca-Cola tastes exactly the same even though they’re bottled in different facilities with different owners.
These things did cost a lot to develop, but for the consumer it’s quite inexpensive. As GP said, we just take these things for granted and don’t notice them.
The coke bottles themselves are amazing too. I used to go to school with a reusable bottle filled with milk. Those cylindrical lunch bottles for kids were absolutely horrible. They leaked half the time, spoiling my bag and notebook. If you dropped them they would break because they were hard plastic. The rubber ring that was supposed to stop it from leaking would degrade quickly and start to smell funny. Those things cost as much as 20 bottles of coke, and an empty coke bottle is a vastly superior product in almost every way!
A bit of a woo-woo aside but I've been trying to practice more gratitude thinking in my daily life and the grocery store is an easy place to be reminded of how good we have it.
Practicing gratitude is probably the single easiest way to increase one's happiness, and yet it's so easy to forget to do it even if you know that. At least for me.
One thing we can do is keep a gratitude journal where we write down things we're grateful for. Can literally be grateful for the sun shining, or not experiencing an earthquake, for having the ability to write in a journal in the first place, etc.
It may be a miracle of logistics when ignoring negative externalities like carbon cost vs not shipping it halfway across the globe. Why can’t we eat seasonally? It usually tastes better and makes things less monotonous
I'm a believer that the largest part of what led Yeltsin to fundamentally change what it was, and thus cause the dissolution of the USSR, was his impromptu grocery store visit in the US.
Conversely there are lot of things that are expensive and work very badly. "Designer anything" as for example, designer light fixtures. I've had terrible experiences with these.
I think it's more that you just simply can get away with it: Artsy-fartsy types that care more about design care less about function and quality, so don't question and check them enough for manufacturers to have to keep high standards in those areas.
> There’s a lot of stuff that works remarkably well even though it’s cheap.
Even light switches are pretty cheap. You can get a basic single light switch for around $2. Sure there's decora switches, dimmer switches, and all kinds of other great things for $50+ but the basic $2 ones will still last decades.
If you replaced the contacts with a triac and replaced the switch mechanism with a thick bistable flexure, it would likely last centuries and have a BoM cost around $2.
I’ve had 2 switches fail over the last 2 years out of the ~40 switches installed in the house. One failed by welding itself closed and another failed by caving into the electrical box when I hit it too hard. Even though the 40 year life expectancy of a single switch sounds good, the reality is that one fails catastrophically every year. I’d love to get more reliable switches that last well over a century, but I’m not aware of anyone that measures this sort of thing.
Exactly. I can't imagine how expensive a computer chip should be if the process isn't optimized / streamlined.
And for kitchenware and dinningware, we still get decent quality for a still rather cheap price. Of course as the article stated it's not easy to determine which one with decent quality, however if customers only aim for the cheapest one of course it won't be good.
I beg to differ, many of these large household companies are shells of their former selves as they've been bought, bankrupted, and traded around. What's left is just a name with no solid product line backing it anymore. E.g. Sunbeam was a solid household appliance name, and now it's a crap. Same with Braun.
The easy industrial design exercise seems to be luxurious looking materials paired with cheap electronics. Amazon is full of this. Oddly, the thing I end up trusting these days are in-house brands because the store has some responsibility to make sure their own brand's reputation doesn't get too tarnished.
After purchasing the top two Wirecutter picks for electric kettles (Cuisinart and some gooseneck kettle) both died within a year. The gooseneck one was rusted on arrival, clearly awful build quality.
I decided to try paying much more for a Fellow Stagg EKG, and it was a great decision. It’s lasted over 3 years and has been an absolute joy to use compared to the prior mass market garbage.
I often wish for a Wirecutter-like site that prioritizes quality and especially longevity above all else. Wirecutter always focused too much on cost, and even their “upgrade picks” tend to suffer awful quality issues. For years their top blender pick was an Oster that had hundreds of angry reviews about dying within months. Wirecutter ignored the feedback for years despite so many people streaming into their own comments section to vent about it.
Honestly, I had spent 20 years in the US and we consistently bought the cheapest appliances ever.
When I bought my house I finally said "screw it, let's see what decent appliances look like".
Japanese rice cooker set me back $95 and I thought I would never hear the end of it, and after 4 years, it had already paid itself off (we were doing $14 rice cookers every 6 months). Air fryer was $70 but the previous $40 only lasted 13 months. Basic coffee maker was like $60 but made non-burnt coffee. A little combo oven/toaster is what I ended up on since we had one in the last apartment since we never used a full oven.
The ones that are honestly pretty difficult to find were dishwasher but one of our friends suggested bosch because we wanted a quiet appliance.
There's also just a... man, I don't know how to describe it. Kind of a mental benefit to using slightly nicer things.
When I was young, almost everything I owned was the cheapest possible version of that thing. Everything just kind of sucked, brutally cost-optimized to the point of being somewhat nasty to use and barely functional.
I was still very fortunate: I had food to eat, clothes, etc. A lot of kids in the world would have traded places with me.
Now that I'm older, I have no interest in "luxury" goods, but there's that subtle intangible benefit to using e.g. the $95 rice cooker vs. the $14 rice cooker. You feel like somebody who's worth more than the cheapest possible piece of disposable shit, I guess. Or at least I do.
It makes better rice, too, of course. And there's the ecological benefit of not tossing a $14 rice cooker into the landfill every couple of months. But there's also a bit of self worth involved, or something.
In the US, the $95 cooker lasts no longer, and works no better, than the $25 unit. (There is no $14 one.) You might be able to do better with a Japanese brand, but it is vanishingly unlikely you will get the same one as they would have sold in Japan, unless you actually get it shipped from there.
I make rice in a saucepan on the range top. I have to come back and turn it off when it's done. Otherwise, it is the same. If you care about how good your rice is, you are starting with short-grain rice. Or red, or black, or arborio for risotto.
I've had a Zojirushi for 20 years, and my mom has had hers for 30. Probably paid $250 when new. We use it nearly every day. My kids know how to make rice with it, and it comes out perfect every time. I wouldn't trust a ten year old to make rice on the stove, but they can do it with a rice cooker no problem.
They can do it with your exact rice cooker, no problem.
I grew up with Zojirushi rice cookers. They always worked. The last one I bought sputtered starch water all over the counter. Stuff you can buy in the US today is not the same as what we could buy even 20 years ago.
The one I bought from them 4 or 5 years ago is great. Same with the water boiler. But I made sure to purchase models that are still made in Japan —- not all of their models are.
I bought them off Amazon, do sourcing isn’t difficult, but some research might be in order if you want Japanese manufacturing. You’ll pay more for these models as well — they aren’t the cheap or maybe even middle-priced options... (I believe my rice cooker was nearly $300, 5 or so years ago...)
For kitchen stuff, America's Test Kitchen has amazing equipment reviews.
In addition to their testing process itself, they actually dogfood their advice by using their own picks in their test kitchen so they get used by tons of people way more often than any testing process could accomplish so they can get a real sense of how good a recommendation holds up over time.
Last Yule I bought a toaster for my brother. All of them felt kinda like crap, flimsy cheap, scratchy action... I was not even looking at cheapest but something I imagine to be reasonably mid-range that is around 50€ mark. After all it is a moving platform, some heating elements and case. Not at all complicated.
Commercial toasters (and, I imagine, commercial electric kettles) are 'good' in the sense of being well-built and long-lasting. I've got a 1980s Dualit toaster which is essentially bomb-proof (the clockwork timer will eventually stop working but is easily replaced; even the elements can be readily changed out if they get damaged).
Of course, the downside is that new ones start at £150 or so. So it's difficult to make a financial case (as opposed to an aesthetic, or a principled one) over a £10 special from Tesco.
The problem with ‘commercial’ kitchen equipment is that most of it is just up branded domestic equipment.
When you do buy ‘commercial’ kitchen equipment you’ll notice lots of things that are just downright worse. Energy efficiency, safety features, and noise reduction are all things that are _way_ worse than with their domestic counterparts.
"Real" commercial kitchen equipment is often totally different.
Commercial fridges will stay cool even if their door is opened 20 times an hour. Commercial glass washers take a tenth of the time a home dishwasher takes. And if they're noisy, ugly and they need to be cleaned every day without fail, that's just normal commercial equipment.
I bought a store brand 9A kettle and it's fine [1]. They're simple products and shouldn't break unless you have very hard water, in which case a round of vinegar should clean that. What are you running into?
I read somewhere, and this guy was talking about how if you want your house to sell for more, invest in everything you physically touch. High quality doorknobs, faucets, and light switches have a marked impact on our unconscious valuation of a house.
Check the deltas, the derivative. There exist long lists from personal experiences of things that worked very well and now are of comparatively terrible quality.
The issue is then not just with the item, but with societies that are increasingly accepting low quality: this is a horrible trend, and one side of decadence. You get both, flanked: low quality here for the occasion and decadence around for the trend.
The idea you say of some "distracted" ones "not realizing the failure potential" has a legitimate justification, beyond the simple inattentive, in those (inexperienced) that assume, for a number of reasons (especially including an internal healthy "mindset" of good standards), things are done properly. There is a line in a script for Scorsese that goes like: «I'm the guy doing my job, you must be the other one».
things that twenty, ten, five years ago were of high quality - same brand, update of same model - now you buy at a comparatively abysmal quality for a very similar price. It is today easy to find products which are cheap in manufacturing and expensive as a price tag.
This means that, in some way, people in some/many societies are tolerating quality degradation. And a decrease of alternatives is contributing. That, in some areas, it was once not necessary to spend time investigating which product was high or just decent quality (already the price could have been a good indicator), while now it is part of your task, shows that tolerance for low quality has increased. That is not for the 1 dollar item, but for whole range up to many figures.
And, a staggeringly increased inability to perceive degradation in general is evident today visiting some territories (and see what is tolerated now and was not before).
Just to give one example, CI pipelines seem to fail all the time. For closed source and open source project. Just like this, it worked in the last commit and in the new commit it fails despite the test suite passing. The ultimate reason is routine tasks pulling in a ton of complexity of which only a tiny fraction is being used.
At workplaces this creates a lot of absurd situations that eat up insane amounts of productivity.
Or another example, it's pretty common that water pipes don't work as expected. (Congestion, low pressure, undesired backflow, tricky to get water at body temperature...) Nobody really complains, everybody lives with it and learns to completely ignore it. I'm not saying these problems occur everywhere 100% of the time but often enough to show there's something structurally not working
There's counterexamples for each of those though; just thinking of Flint, Michigan, or the Great Toilet Paper Shortage of 2020. Also I had to replace a light switch in my shed the other day.
Many of those things mentioned have changed very little in decades. Some have also been under continuous improvement for hundreds or thousands of years.
Foundations change (tensioned slab, etc) and water inside of houses (copper vs pvc vs PEX vs lead). Electrical is even in frequent flux under those timescales. Insulation as well.
But the concept of water being fed through pipe, from a reservoir, chlorination, etc have been around a long time. Sure implementation details change, but they and the concepts they build off of have been around. You mention pvc, copper, and pex (lead is over 50 years old and many have been replaced). Those are just materials. How do they affect end user on a daily basis?
How long has electrical be 120v AC? How long has auto voltage been 12v vs 5v?
Details, materials, and implementation change (building off of prior versions), concepts and overarching system designs are slow to change.
That feels like we're stretching the claim a whole lot. The "basic concept" of a Tesla isn't that different than a Model T but I think most people would reject the claim that cars haven't really changed much.
Road design and laws have mostly stayed the same in the past 50 years. How about controls (pedals, wheel, etc) - more or less the same as well (subject to regulations).
There are similarities in your example. The fact that Tesla has autopilot and is an EV represent two of the biggest moves away from traditional car concepts. If you used an ICE car I would say the concepts haven't changed much.
As someone who writes software for a big grocery store chain in Germany I'm surprised the logistics work at all. It's a s*show inside once you know the details, but somehow, yes it kind of works well enough as for the customers to not even realize is there.
I think you put too much confidence in other engineering fields. They go wrong all the time (you might notice some when purchasing a house) and changes are extremely slow and expensive.
After the fall of the Soviet Union, UK experts flew in to help with the transition, and one of the apparatchiks asked: "We are eager to try this capitalism thing; now tell us: who is in charge of the daily delivery of bread to London?"
Looking through old Sears catalogs and seeing the prizes on older game shows, I think some of our memory is selective. Even in photos and SD video you can see the terrible build quality on many consumer products from those earlier times. This seems to be especially true for furniture and exercise equipment. When you convert into current dollars, this low quality junk often is more expensive than what is available today. There's plenty of junk being sold today too and paying more doesn't guarantee quality but the notion that it's more difficult to find quality today than in the past doesn't seem universal. We have plenty of items that survived earlier eras and are examples of durability but the landfills are full of junk that we don't see.
The idea behind the article is not that everyone is making junk these days. The idea is that we often have no way to find out which products and services are junk and which are good quality. How are you going to find a contractor to remodel your bathroom? How will you even know if they did a good job?
In the Jungle, regarding the experience of immigrants in the early 1900s, Upton Sinclair writes, "If they paid higher prices, they might get frills and fanciness, or be cheated; but genuine quality they could not obtain for love nor money."
In other words, 90% of everything is crap, both then and now. And thanks to Lindy's Law, you're usually better off buying something old than something new. But that's not how most people think :(
In personal experience things were more crappy then than now. The explaining factor could be that production techniques have evolved as a whole, and less manual steps harmonizes the quality of cheap goods.
To take IKEA as an example, their table were really bad 20 years ago, and the only option was to either buy from another flatpack company, that was often worse (assembly would need like 50 screws for a single table...wtf), or a hand made "old fashion" table that would last a lifetime but cost 10 times more. Current IKEA tables will last a lifetime for the same price as the crap from 20 years ago.
Another example is French and Italian cars. They were really bad decades ago, reliability issues were only compensated by the ease to repair and cheap local maintenance. But industrial process are so much better now that production quality and reliability is at a level way above what would be expected for cars of roughly the same production cost.
IKEA at least was, the last time I shopped there, willing to sell at different levels of build quality. Some products cut every corner possible in the name of cost, but also passed a noticeable portion of that savings to the consumer. Others used some slightly better materials in places where the results mattered with predictably good results. As long as the consumer made an informed choice the result was as desired.
IKEA always have had a range of different quality products. You could buy cheap stuff that didn't last or you could more expensive stuff that last 20 or 30 years ago as well as today.
Ikea is a bad example in IMO. I moved countries 3 times in the last 10 years and bought many of the same exact Ikea items all 3 times. The quality went down each time. Things that were previously metal became plastic. Things that were previously reenforced in 4 positions were now only 3. etc....
Also, several of the more sturdy things they had they no longer sell and what they have now in the same category are vastly less sturdy.
I agree though, they do still have some good, quality, sturdy kitchen tables and a few pretty good sturdy sofas.
> Thus, the Lindy effect proposes the longer a period something has survived to exist or be used in the present, it is also likely to have a longer remaining life expectancy.
Makes sense!
I was also thinking there's a related rule - something worth buying used has proven itself to not be crap.
Another way to look for non crap, is to check the used value of something. E.g. check swappa. If people are still paying a lot for an item after it is used, usually a good sign. If they are not and you still want it, get it used and save a ton.
Another good way, check how much insurance costs. e.g. How much is that extended warranty going to cost on one car vs another. How much is home owners insurance in one neighborhood vs another. Insurance companies that make mistakes here go out of business.
Especially for older 'thing' you might be looking for. If it is at all collectable, the price will be higher, regardless of quality.
Ice cream makers are one thing I have found that for. White Mountain made a really high quality product, but the used market is obscene, because people collect them apparently (at least in my part of the country).
People still try to sell their used crap. So how do you define if something is worth buying used ? For some people everything is worth buying used, even crap
It’s standard practice to reduce safety margins when build quality increases. If you think in terms of expected lifespan, when the company has little control over how long things last and wants a minimum of say 1 year then some fail at 1 year and some fail after 20. As their process improves the minimum stays the same, but maximum lifespan decreases.
"Value Engineering" - make the same thing, but at half the cost, eventually you get chip bags filled with air and reduced by 1.5 oz margins until the profit matches what's needed.
The problem here is actually very basic - markets do not enforce efficiency. Period.
The efficient market hypothesis is simply wrong. Strong form efficiency was already formally disproven quite a while ago. Weak form efficiency has been show to only true if P=NP.
There is not really a strong need to go into why inefficiencies can persist, etc, because at a baseline, it turns out there is no provable theory that markets should be efficient at all.
Instead, we have the sort of equivalent of Galileo - people really badly want markets to be efficient. They feel like they should be, because intuitively, it seems like it should work that way (much in the "god does not play dice" sense). People who suggest they aren't, even when math backs them up, are ridiculed. Eventually, as Max Planck said, science will advance funeral by funeral, and we'll stop pretending markets should be efficient based on "the efficient market hypothesis". Or we'll discover P=NP!
Which would be much cooler.
Sure, sure, you can encode difficult computational problems into markets in convoluted ways, but this is uninteresting unless you're into that kind of thing.
When people usually say that markets aren't efficient, they don't mean that optimal resource allocation is a computationally intractable problem. They're saying that they see, clear as day, obvious inefficiencies that aren't being corrected. They're saying that there are easily noticeable inefficiencies that aren't being corrected. And finding easily noticeable inefficiencies isn't NP-hard, by definition of easily noticeable.
A better version of the efficient market hypothesis would be that markets are inexploitable. There is no easy action I can take that corrects a market inefficiency and makes me money. This is the version that comes up in cocktail party conversations and it's also the one that Dan Luu is talking about. "If XXX is systemically undervalued by the market, why can't I start a company that specialises in using XXX?" The discussion following this question is much more relevant and interesting than encoding 3SAT into economic models.
> A better version of the efficient market hypothesis would be that markets are inexploitable.
No, that's absolutely not true and relies on what I think is the most perniciously false assumption behind market economics: that all market participants only make moves within the market itself.
If you're trying to win at chess then learning to master the rules and strategy of chess is where you focus your attention. But if you're trying to defeat your opponent (or not get defeated by them), you'll probably do better to bring a gun and/or body armor. If you focus 100% of your attention on the chessboard, you are completely opening yourself up for exploitation by an opponent who is willing to play the metagame.
And, indeed, in market economics, actors invariably do play the metagame. This is why we get cabals, trusts, rent seeking, regulatory capture, monopolies, price fixing, price dumping, lobbying, etc.
The ultimate goal of market actors is not to be maximally efficient. It's not even to win the market game. It's to make the most money. And often the best way to make the most money is to rig the game, cheat, or get the rules changed in your favor.
Then couldn't you reframe things to include the entirety of human civilization as a kind of market, and still show that there are unexploitable inefficiencies?
Sure, but then your definition of "market" has no meaning.
A book on "chess strategy" would be very different if it's definition of "chess" included the whole of human conflict and warfare. It probably wouldn't spend much time talking about rooks and pawns at all.
That's not what I meant. I mean there's no rule that you have to include only what we call "the market". I mean the "real" global market includes politics, crime, etc. Basically anything humans do to get their hands on money.
It's only uninteresting because economics refuses to be based on anything but the empirical (leading us into the situation we are in now), and so considers things like "math" to be mostly uninteresting "unless you are into that kind of thing".
I looked at markets for a while, this is what i saw. Therefore, it's true.
As I said, there is no reason those efficiencies should be easily correctable - because markets aren't efficient! That's the whole point. It's only in an efficient market that they would be correctable. This is the same as lots of computational and other problems. Lots of instances are easy and heuristics can often work very well. But you will still come up with situations where obviously broken things happen and are hard to make algorithms work.
It's sort of like having a cocktail party conversation about why you can't exceed the speed of light. Except because it's economics, you get stuck because there often isn't any real rigor behind it that you can push on. Just the empirical. Which I get why it's fun to talk about (really!) - without any meaningful rigor, anyone can participate and have fun - got a crazy empirical story? Awesome, that's all you need to prove something in economics! It makes for fun discussions where most people can participate and feel like it's not too hard.
I have no issue with that - my issue is of course that the cocktail party is not distinguishable from the field ;)
Beyond that, arguing they are inexploitable/unpredictable seems equivalent to whether they are efficient (and in most papers is considered equivalent to the efficiency hypothesis). I'm really unsure how you are trying to distinguish it.
Maybe you could formulate it for real and show how it is not equivalent to the efficiency question?
Inexploitability is not efficiency. Back in the 2000s, plenty of financiers knew the subprime market was overpricing assets for years before anyone figured out how to profitably short the market.
Hard to deflate asset-price bubbles are a fact of finance.
For strong form efficiency (which requires markets perfectly reflect all available information) -
You would have trouble finding people who still believe in strong form efficiency.
The weak form paper above cites several papers that go into why strong form efficiency is impossible. There are formal proof versions around. In practice, even empirical studies of strong form efficiency haven't supported it either - it's just very easy to find practical counterexamples.
I skimmed this paper and it seems to be exceptionally bad. The proof sketch offered is that a winning strategy can be verified quickly, but finding a winning strategy requires searching over 3^n possible strategies.
But this assumes that brute force is the only possible way to find a winning strategy, and thus proves far too much. A similar argument would prove that sorting is NP hard, if you start by the assumption that the only way to sort data is by trying every possible permutation.
I may be wrong, but I'm pretty sure any time you claim to have a proof that a problem is NP complete, but your proof doesn't include a reduction, you're doing it wrong.
(The paper does offer what it calls a reduction to 3-sat
, but it's completely hand-wavy, and I can't even understand the intuition behind it at all.)
It isn't about market efficiency at all, people are just incompetent. If people could create good products they would and they would become rich, but they can't. If Samsung suddenly started delivering the same quality as Apple they would make a lot of money long term, everyone knows this and even if it isn't true the people who manage Samsung actually believes it and they try to get there. The reason they can't isn't that they aren't trying but that they don't have competent enough people to get it done. Leaders are a part of "people" btw, so if you say it is structural or organizational bloat, well you fix structural and organizational bloat issues by having competent leaders, so those problems are caused by lack of competent people.
It's about what kind of competence is favored by the market - it favors people who can get funding, which is not necessarily related to the ability to get the job done.
I just moved to the USA from the UK (having lived in Germany before) and the quality of the services I'm experiencing here is appalling.
To get concrete, we submitted an application for an apartment let. After not hearing back for 4 days we reached out and it looks like somehow they misplaced our application. By then the unit we wanted was gone so they asked us to pay a holding fee to ensure they would keep another unit for us. The online system simply didn't work, so we reached out and they informed us it was fix. But the system just marked the unit as rented rather than reserved, so we had to reach out again...
As Dan says defendants of the efficient market hypothesis will say that it isn't possible to provide a better service, or otherwise this letting agency would have been upended by competitors. Only that our experience with other agencies has been similarly poor, and our experience abroad is generally less so.
And is not only renting, most of our interactions here have been similarly shoddy. Which makes me think that either we have been incredibly unlucky or there is some common pattern. Here are some ideas for our particular case:
* Barrier to entry: Businesses like real estate require large capital and tend to be crowded already, so it is difficult for a disruptor to appear.
* Liquidity: While in theory we could *shop around* if we don't like a let agent, in practice there is a very reduced number of units that satisfy our requirements. It's easier to put up with a poor experience that compromise on the property.
* Corner cases: As new immigrants we had to go through some processes that most people don't experience. Probably is not worth for companies to optimize for these situations.
* Cultural normalization?: This is the most intriguing point for me. Maybe this level of shoddy service has been normalized and accepted in America - or at least the area we live in. Maybe people just don't know better and hence don't ask for more, and therefore companies can carry on providing what would be considered a sub-par service elsewhere.
Same experience coming from France. Everything in the USA is just lower quality. Be it goods or services. And Americans seem ok with it.
There is a lot of overhead everywhere. Dentists have an army of assistants. Any product to order requires it's sale agent. Even if you already know what you want. You must deal with the agent. Car dealership have so many sales but good luck getting an appointment for fixing the shit they sold you. Look to by a house, and be ready to deal with insane fees and agents. All this overhead has to be paid somehow I guess.
Where in the USA are you? I haven't bought a new house in a while but most car dealerships (at least mine) gets me an online booking appointment for the same day or next day. Doctors sick appointment - mostly same day. I've never interacted with a sales agent for buying stuff.
I agree with you on housing. The agents charge commission and closing costs and closing companies do charge quite a bit.
Having lived in Germany, I must say the quality of customer service there is the worst I have ever seen.
+ A fridge I purchased was delivered with a 2 weeks delay, at 10:30PM (!!) on a Saturday (only because I kept calling every other day).
+ I've ordered a new broadband connection for an apartment I was renting. After several months (!) of waiting and calls, I found out that the former tenant had unpaid bills with the provider and they therefore refused to establish the connection (I was only informed about this after 5-6 weeks of wait during which I was told everything's ok and pending execution).
Here's a story about how my laptop was broken and I was calling a repair shop:
Me: "Good morning, my laptop doesn't seem to display anything, looks like it's a problem with the screen, how does the pricing work in this situation?"
Shop: "Ugh...", the guy is annoyed.
Shop: "This is like saying you have a headache, and asking a doctor what needs to be done!", He shouts.
Me: "No need to get angry, I was just wondering-"
Shop: "Just bring it to the store!" - hangs up.
This is a perfectly normal experience dealing with customer service in Germany.
The shop guy was right though, even if he expressed it bluntly. How could anyone come up with a useful answer to that question? It's actually worse than the headache question. And yes there's not much of beating around the bush in Germany, which can be somewhat shocking to those not used to it.
We charge N/hr, plus parts. Or, we charge N for an initial diagnosis.
My uncle runs a computer repair business and is doing very well because he works late hours, is in a college town, and has flat pricing when possible. This means he spends Sunday night reinstalling OSes on student’s laptops at $75/each.
> To get concrete, we submitted an application for an apartment let.
It's so hard to hire good help these days. And in the case of apartment building management, it sounds like they're renting out just fine despite the bugs and process fuckups. Hell, even you havent walked away for a competitor.
> Barrier to entry: Businesses like real estate require large capital and tend to be crowded already, so it is difficult for a disruptor to appear.
In a sense, there are startups in the "long term rental" space providing a better frontend to shitty landlords. Even airbnb is moving into that market. Whether they can successfully out UX Prometheus while relying on prometheus as a landlord I cannot say, but its one path out of capital requirements.
> Cultural normalization?: This is the most intriguing point for me. Maybe this level of shoddy service has been normalized and accepted in America - or at least the area we live in. Maybe people just don't know better and hence don't ask for more, and therefore companies can carry on providing what would be considered a sub-par service elsewhere.
Lots of people own housing instead of rent here. But you'll find plenty of similar angst about mortgage processing.
Your letting experience is missing one key factor; you’re not the customer to the letting agency. The building owner is the customer, not you. As long as the apartment gets rented out and maintenance gets done well enough, the owner doesn’t really care if you have a bad experience. And the letting agency doesn’t have to care about you either, since their treatment of you has no effect on their bottom line.
The fact that they don’t feel ashamed of mistreating a non-customer is however cultural.
This is on point. It is not that the service is not replaced by a service serving the customer better. The customer is already being served perfectly fine.
Yes, quality is going to hell in the US. It usually helps to book help from people who are from other countries, even places that don't have a good reputation. They left, right? And, managed to come here. However bad it is here, it is anyway hard to get and stay here.
It's pretty depressing how many of the comments are purely answering the title and not discussing the content of the article. In particular, the comments seem mostly focused on consumer product build quality whereas the article is more focused on e.g. businesses trying to outsource something (like same day delivery) and getting a worse result even when they pay more than doing it internally. And the cultural obstacles that make these problems difficult to fix or even sometimes perceive.
I will note that I often wish Dan Luu was a bit less apparently-uniformly-confident in some of the statements he writes. He does back up what he says, but he'll use the same tone for something he's seen anecdotally and something he's spent a month personally investigating.
> It's pretty depressing how many of the comments are purely answering the title and not discussing the content of the article.
I’ve said this before on other threads, and it seems to resonate with others, so I’ll say it here too:
I personally don’t care near as much about a single person’s opinion of a particular topic, as much as I care about the discussion of many people around a particular topic.
And by “topic”, in the case of HN, I mean the title of an article.
I often wish HN would just allow titles to get posted and HN folks just have a discussion around that. (Similar to Ask HN, but not necessarily in the form of a question.)
I’m a terribly slow reader, so there’s no way I could take time out of my already busy day to read this (very long, dense) article of a single person’s opinion of a single topic, especially if I don’t even know whether this person is an expert on said topic.
I guess this is particularly true when it’s a subject I’m only interested in on the surface, vs. caring more deeply about a particular subject.
I get much more ROI reading many people’s shorter discussions vs. a lengthy single opinion (again, given the time it would have taken me to read the article).
I have to strongly disagree. The lengthy article has much more thought put into it. It's almost always better researched and the author usually has relevant experience or expertise, and those are usually at least known to the reader in some way or easily discovered.
I have seen many, many cases where a big portion of the top comments are either saying something similar to what the author of the article said, or they are asking a question, or raising a rebuttal, that is already addressed in the article.
Comments are short and probably have close to zero research on average. Experience may play a part, but people don't usually offer up the resume with a comment, so you have no idea if they even have credibility. And there are mountains of witticisms, anecdotes, or emotional arguments that add little to nothing of value to the conversation.
Personally, I almost always read the article and then skim the comments to see if anybody has added any thoughts I might find interesting.
I simply stated my experience — I never even said my experience was right or that it is wrong to read the article, or that others should experience HN the way I do.
I’m a slow reader and don’t have time to read lengthy articles, and therefore am drawn more to discussion.
You are (likely) a quicker reader and are drawn more to articles.
There’s room for both of us in the world — and on HN (I hope).
You said "I often wish HN would just allow titles to get posted and HN folks just have a discussion around that". If there's anything I can wholeheartedly say I disagree with it's that, so yes there's a lot to disagree with here.
I think that was an inadvertent juxtaposition of the words "just" and "allow": the GP probably didn't mean that links to articles should be forbidden, and only subjects for discussion ("titles") should be allowed to be posted.
At least I read their comment as "I often wish HN would allow, besides links, just titles to get posted and HN folks just have a discussion around that".
Would you still wholeheartedly disagree with that, too?
A part of the problem here is distinguishing which articles are worth reading. I may know that a Dan post is probably good, but if I saw this exact same headline taking me to a long article on <random tech news site>, I'd probably skip it and skim the comments instead, since chances are the article itself is near worthless.
I very much disagree with this. In fact, this article is a perfect counterexample to what you're saying, ironically - it's a lengthy, well thought out essay that the author clearly spent a lot of time thinking about and putting together. Quite honestly, it's very unlikely that any number of pithy HN comments could match up to it, and I say that with a fair amount of respect for the HN community at large. It's just very unlikely they're going to have insights on the same level: the author of the post had at least a couple of weeks of actual concentrated effort to put together thoughts on the topic.
Plus I just have a lot of respect for the author. Dan Luu is a smart guy.
> And by “topic”, in the case of HN, I mean the title of an article.
The thing is, the topic of the actual article is generally more interesting than the topic suggested by a naive reading of the title (as is the case here).
People discussing things based on titles leads to generic discussions that recur again and again and go nowhere. Whereas the actual topic of the article's content is rather more specific and could lead to an actual new and more specific discussion that might actually go somewhere.
Possibly, but I would bet that many people are upvoting these articles based off a combination of the title + the discussions of the title, which would suggest enough people find those two things interesting (which is why it’s at #1 and why most of the comments are around the title vs. the subject).
Also, if the subject is more interesting than the title, then authors should consider being more thoughtful around a title that fits the subject, vs. making sure people click the article.
> Also, if the subject is more interesting than the title, then authors should consider being more thoughtful around a title that fits the subject, vs. making sure people click the article.
The problem with this is that, frequently -- as is the case here I'd say -- there is no easy way to properly pin down the actual topic in the space of a reasonable title. The title will necessarily be more broad than the actual topic; if you in general take titles as wholly delineating the topic, rather than suggesting the general space they're in, you're making a mistake.
Sticking only to those topics that can be wholly expressed in the space of a brief title is a big limitation on topics; as I said above, taking titles as topics just leads to the same discussions over and over.
I absolutely understand where you're coming from here, but in practice and reality, the point of the title really is to make sure people click the article.
I agree about the comments here often being the most interesting part.
But I think that the best way to start the conversation is usually with a well-written article.
If there would only be a title and no article, I'm pretty sure that I would find the comments to be less interesting.
But maybe I'm wrong, and since my argument is based on the original article being well-written, maybe it isn't true for most posts.
But I still think that I would rather try to find the posts with good articles, and just ignore the rest, and hopefully, that is what the ranking system does (at least help with).
But on Hackernews those one-off comments can come from years of studying the same topic. Sometimes even explaining why the long well researched opinion missed something so important that the whole article is wrong.
No, they dont. Which is pretty clear everytime something you do know a lot about pops up. Majority of it is just people shooting their opinions over coffee break. Which is fine, we go here to relax and slack, but does not produce super educated opinions.
I often wish HN would just allow titles to get
posted and HN folks just have a discussion around
that. (Similar to Ask HN, but not necessarily in the
form of a question.)
Sounds like a real recipe for hivemind to me. I mean, any community struggles with hivemind a bit. But, at least with discussions centered around articles that exist outside HN, we get a little fresh air.
especially if I don’t even know whether
this person is an expert on said topic.
He's a pretty respected voice, or at least popular on HN.
I find it very valuable to look into the authors of linked articles, see which authors pop up on HN frequently, etc.
Gives you an idea of who the leading voices in the industry are, helps you to know which way the winds are blowing.
Especially for somebody with a slower reading speed this may be crucial; helps to understand which article may actually be worth the time investment.
Sure, but then we’re discussing the topic as set by the post’s title, not the actual topic(s) addressed by the post’s content, which are quite a bit more complex and nuanced than a title of a few characters could ever convey.
What if we want to hold a conversation about the actual post content? Should we add a tag such as [please read]?
Why not both? There is good discussion being held on both topics in this thread.
That’s the benefit of nested comments, to allow different top-level discussions where others can discuss what is being shared by top-level commenters.
My original point was to not be “depressed” by others choosing to not take the time to read the article, and have a discussion on the topic rather than the subject.
> I will note that I often wish Dan Luu was a bit less apparently-uniformly-confident in some of the statements he writes. He does back up what he says, but he'll use the same tone for something he's seen anecdotally and something he's spent a month personally investigating.
Dan Luu is one of my favorite currently-active bloggers. He's obviously a smart person with a good amount of experience in specific areas.
However, I have to agree with your assessment. He has a tendency to present his personal anecdotes and perspective as the infallible ground truth and build elaborate essays and logic around it. I hesitate to talk about it because I think his blog is valuable, albeit if the reader can take it as one person's perspective rather than the absolute truth.
The biggest example of this effect is his Twitter thread about his tendency to fail interviews ( https://twitter.com/danluu/status/1470890494775361538 ). He spends a lot of time bragging about arguing with interviewers, dodging questions instead of trying to provide answers, and how he's never written production code that talks to a DB, performs an RPC call, or connects to an API. It's not at all surprising that anyone would be struggle to pass interviews like this, but he seems incapable of accepting that maybe he's doing something wrong or that he's simply not a good fit for the jobs he applied for. Instead, his implied analysis is that the interviewers are simply wrong, and that they're making a mistake to ask him those questions and eventually decline him.
The interview topic is especially challenging because blaming the interviewer is so very appealing to his audience of developers (who all abhor interviews and hate rejection even more). I think my disappointment comes from the fact that he's well-positioned to display some humility and teach some lessons about how someone can learn from their shortcomings, yet instead he uses his platform to further demonize his interviewers and make some worst-case assumptions about why he didn't get these jobs.
I think the best way to read this blog, like any, is to remember that the author is just another person with another set of perspectives and opinions. There's a lot of value there, as long as you take it with a grain of salt and remain open to other lines of reasoning.
I think that's kind of the point. Dan Luu can change his priorities to do a performative display of what interviewers expect and then Dan Luu can pass a job interview. In which case you are not hiring Dan Luu, but a different guy who has most likely internalised a different world-view and set of values. Even if you assume he can compartmentalize his interview performance skills, he will have at least devoted time to that, instead of to something "more useful".
I don't think Dan Luu is struggling for work.
So the question is do you want to hire someone like Dan Luu or not? If the answer is yes, then you might want to consider how your interview process might interact with such a person. If you are thinking from the perspective of someone doing the hiring (as those articles seem to be), it is nonsensical to simply respond by saying "well, the candidates just need to be more submissive and compliant to whatever our process is."
Perhaps Dan Luu is a unique snowflake and/or nobody needs to hire someone like that. Or it could be that there are quite a large number of developers with attitudes and experience that lead to similarly unproductive or inefficient interactions with tech interview processes because those processes may be fixating on having the candidate do a specific performance rather than trying to understand what individuals can offer and whether that would be useful when added to the existing team. It could be that such developers would be as good or better fit, in a lot of cases, than people who can do oscar-winning interview performances.
> I think that's kind of the point. Dan Luu can change his priorities to do a performative display of what interviewers expect and then Dan Luu can pass a job interview.
This is the problematic logic I was trying to highlight: It's written as though he's infallible. He knows the correct answers to the interview questions, but he also deduces what his interviewers are thinking and why they're wrong.
The other possibility is that maybe the interviewers know what they're doing when they decline him, even when he answers the questions correctly. Interviews are about more than just reciting the correct answers to the questions, but he only discusses them as a sort of pass/fail quiz where the candidate is supposed to guess what the interviewer wants to hear.
My perspective is likely quite different as a hiring manager. The part where he talks about Palantir walking him out the door despite correctly answering the questions as fast as they can deliver is the kind of thing that happens when the interviewers agree that someone is smart, but isn't a good fit for the team. If you do enough interviews, you eventually come across people far more brilliant and successful than yourself whom you would nevertheless not really see fitting into your company's work culture. Someone who openly boasts about dodging interview questions and debating interviewers because they think they can read the interviewer's mind (and they are wrong) fits this description. You know they'll do fantastic things somewhere, but they're not the kind of person you're looking for to fill the open position on your team.
> I don't think Dan Luu is struggling for work.
Neither do I! I never meant to imply as much. Dan posted the long Twitter thread over a period of several weeks, so it appeared in my timeline frequently. I only brought it up as an example of his writing style, not his career.
I've not gotten the job offer from some interviews in which I aced the tech but in which I did not 'fit' (how do I know, well recruiters and people from company told me but also I knew, I often either breeze through interviews or I fail miserably and these I breezed)
In one interview where I would have gotten it but didn't the reason was basically I wasn't submissive, the tech lead was in my opinion rude and so when he made some technical mistakes I pointed them out. Why did I do that? Because I did not need the job. If I had needed the job I would have been submissive.
Perhaps a lot of interviews fail to hire Dan because the interview process is geared towards hiring someone who needs the job. Why is this? No idea, especially as lots of interviews take place with someone who already has a job and the new place wants to attract. But for some reason companies want to attract flies with vinegar and not honey, contra the old adage.
on edit: regarding on why I did that, well I also did it because the lead was rude. Otherwise I wouldn't have pointed out when they were wrong.
When I started at a company, I listened to stand-ups of 'yeah 4 weeks' and whined I could do it -today-. And I could. I was a rockstar, everyone else was an old lump.
FF to today and I realize how much I didn't know. Mostly business relations, relations with product, proper testing and pipeline integration, and honestly...just thinking. I used to be prolific and just write and write. Today I think way more than I write, and accomplish more.
This would be more meaningful if we knew what you think makes a rockstar engineer.
Some people value extreme speed.
Others are impressed by somebody who will dive down a rabbit hole and follow however deep it leads, discovering the linker bug or kernel driver interaction that causes a failure, and patching it.
Or maybe you are impressed that they go away for a month, talking to nobody, and come back with a complete system all ready to ship.
Or they write libraries that everybody uses because they are so exactly what people need.
Or they are always available to explain things to junior people and get them going in the right direction.
Each of those would be unusually valuable, at some places, and would have trouble getting any recognition, at others.
> He spends a lot of time bragging about arguing with interviewers, dodging questions instead of trying to provide answers, and how he's never written production code that talks to a DB, performs an RPC call, or connects to an API.
I think describing what he's doing as "bragging" is fundamentally misreading him. He's just very honest, regardless of how people will perceive it. And he needs to work in environments where that trait will be perceived positively, because he feels it's not an option to turn it off.
I'm sure that, for people who put up a social front as easily as they breathe, Dan's style can be viewed as simply a more elaborate social front, a form of peacocking that he feels he can get away with because of his technical skill.
It's a weird world we live in, where being honest is viewed as the strangest, most exotic and elaborate conceit.
> he seems incapable of accepting that maybe he's doing something wrong or that he's simply not a good fit for the jobs he applied for. Instead, his implied analysis is that the interviewers are simply wrong, and that they're making a mistake to ask him those questions and eventually decline him.
In the thread you linked, he links another thread (https://twitter.com/danluu/status/1447268693075841024) where his interviewer was wrong. And what's fun about Dan Luu is, he actually cares about answering a question correctly more than he cares about the intense pressure to "play the game" - that is, to conform, be agreeable, and affirm the viewpoint of whoever has power in a situation so that they like him. He wants to work in a position where that tendency is valued. Sadly, it usually isn't.
Personally I don't take it to the extreme he does, but I admire him for doing so. Sometimes there is nothing wrong with you; it is the world that's wrong.
You can adapt to survive in that wrong world, sure, but that doesn't make the observation incorrect.
The most popular writers on this site (and others) have this habit of presenting their opinion as the ground truth and it annoys the heck out of me. There are VERY few things I can be confident about, I don't pretend to be that smart. Then you get people like Jordan Peterson who present a mix of truth and unsourced, unverifiable nonsense with the same confidence and charisma, who become thought leaders. Nobody is going to listen to the person saying "I'm not sure" or "I don't know".
As a counter-anecdote I'll offer my experience working for a company that built large precision machinery. Many of the parts of said machinery were built by subcontractors. The quality was extremely high. We did do certain things in house if they require special expertise and we did have the capability to do rapid prototyping but it wouldn't make any business sense for us to purchase all the equipment to make all those parts that we subcontracted.
I think this is common in many industries. (Automotive?)
Even with software this is far less than clear cut. There's plenty of times where you should not build it yourself and the quality stuff isn't always terrible. Seems like there's some cherry-picking going on in the examples re: bad software. Sure, there's lots of bad software, but there's also awesome software that works really well. Building it yourself isn't a guarantee that it'll work well. 90% of the time the people that think they can do better if they build it themselves can't, they just don't know it yet. And sure Kyle is great but there are plenty of solid databases built before he was around (before he was born?) by companies like IBM and Microsoft.
I've seen companies build in-house tooling that's much worse than what's available off the shelf and have an endless drain of resources due to that.
Build vs. buy isn't an easy call. I'll agree with that.
I felt like there was a self contradiction in the essay. You can't get great software but you can get great engineers that will build the great software in house? I don't think your chances of finding the team that can build that are any greater. You're just as likely to hire a bunch of people you think can do this and then find out they can't. 5 years later. Which sort of jives with the commentary of experts doing shoddy work...
EDIT: I'd also say that the examples of the stuff that was built in-house that was superior might be a case of survivorship bias.
EDIT2: Thinking more about machines... motors, sensors, cables, connectors, bolts, power supplies, pumps, linear rails, screws, etc. etc. all of which you buy, never make, the quality is generally very good, and it's almost unimaginable that you can be completely vertically integrated.
I've never worked in the automotive industry, but I've worked adjacent to it occasionally.
The amount of effort put into ensuring that suppliers do what they need to do and do it well is huge. Companies always have two (or more) suppliers and if you keep only to the letter of the contract, you get cut.
They can afford to put so much effort into managing suppliers because volumes are large and margins are low.
There are also cultural issues at play; he makes the point that things that are inevitable in one culture are unthinkable in another and different industries have different cultures
The difference between machinery and software in this example is that one is outsourcing manufacturing and the other design.
When you outsource manufacturing you're asking the other party to create exact replicas according to very specific specs your own engineers created.
With software the outsourcing usually includes outsourcing a large part of the design, which is inevitable, as there is no significant manufacturing you can outsource.
Writing the software is creating the detailed specs. In software there is only product design, you don't really have to put a lot of effort into creating exact copies of the original design.
Most people can’t. “Read carefully or for comprehension” would probably translate to level 4/5 prose literacy skills in the NCES NAAL (National Center for Education Statistics National Assessment of Adult Literacy). Recent survey puts this group (level 4/5) at 12% of the adult population. Levels 4 and 5 are no longer separated because of the small percentage of people that fall into group 5. Level 4 tasks require you to be able to understand something in the presence of distractors. Just to make up an example, think about the last time you sent an email to someone that had two questions in it. Did they answer both questions? (I’m actually a bit unsure that this falls in level 4.)
(Level 5 includes tasks such as “compare and contrast complex information, or to generate new information making high-level inferences or using specialized background knowledge”.)
> Just to make up an example, think about the last time you sent an email to someone that had two questions in it.
Pretty often they are answering easy question and leaving the harder one sleep. Pretty often, the goal is to get rid of you email as fast as possible so that they can go back to what they actually want to be doing.
> …he'll use the same tone for something he's seen anecdotally and something he's spent a month personally investigating.
A.k.a. “avoiding weasel words”. I don’t mean to be clever here, it’s just that it’s far more efficient to assume that reader can make their own informed decision, and give them citations & instructions for reproducibility to help them along the way, rather than shovel a bunch of weasel words into the parts of your writing that are less grounded in reproducible facts.
There are also different expectations for blog posts and heavily researched&vetted articles.
I don't think what you're calling weasel words are weasel words.
"Up to 50% faster!" is weasel wording. It says "slower or faster, but not more than 50% faster," but it sounds like it's saying faster.
What GP is asking for is calibrating for uncertainty. Phrases like "I've seen this tendency in multiple teams I've been part of," or "study[ref] after study[ref] has confirmed this" help clarify what premises and hypotheses are well-supported by prior work/experience and what is the author's speculation. IMO It serves a useful purpose!
I have to disagree! I simply do not understand what useful purpose it serves—IF we are allowed to assume that our readers have a proficient level of prose literacy. The default assumption, when someone is talking or writing, is that they are communicating their own opinions and thoughts, experiences, anecdotes.
I don’t know about you, but when I think back to taking writing classes in school, if I wrote something like “I've seen this tendency in multiple teams I've been part of” in my paper, any good writing teacher I’ve ever had would cross it out. Same as “it is my opinion that” or “I have recently started to form some beliefs in the matter at hand after reviewing some materials that came into my possession” or other… filler. It pads out the document and only tells us things that we should already know, or be able to figure out.
The ability to sort out opinions and anecdotes from factual claims and evidence is… well… something you’d expect from a proficient reader!
> The ability to sort out opinions and anecdotes from factual claims and evidence is… well… something you’d expect from a proficient reader!
How does a proficient reader separate facts from opinions if good writing doesn't disambiguate facts and opinions? Do they just assume?
e.g.: Butter is about as healthy as vegetable oil. Arugula is healthier than lettuce.
I can cite a meta-analysis for one of the above statements. The other is just my opinion, and I have done no research into the topic. How would a proficient reader tell them apart if there's no calibrating for uncertainty?
Side note: I went through the original blog post again, and there are lots of examples of Dan Luu doing exactly what I'm writing about, to the point where I'm confused why the original commenter thinks he's not using a different tone based on uncertainty.
Examples:
"I once watched, from the inside, a company undergo this cultural shift"
"I've both worked at companies that have tried to contract this kind of thing out as well as talked with many people who've done that"
> How does a proficient reader separate facts from opinions if good writing doesn't disambiguate facts and opinions? Do they just assume?
Facts and opinions are not different qualities of the same claim, but different claims to begin with. “Butter is about 18% water, by weight” is a factual claim and “butter is about as healthy as vegetable oil” is probably opinion because the word “healthy” is, well, often vague and not something people agree on, at least in this context.
> Side note: I went through the original blog post again, and there are lots of examples of Dan Luu doing exactly what I'm writing about, to the point where I'm confused why the original commenter thinks he's not using a different tone based on uncertainty.
Yes, I get a sense that there may be some more cogent complaint that the original commenter has, a complaint which provoked the comment, but I can’t figure out what that complaint is from the comment.
Just that people who complain about tone often have something else to complain about, but tone is more obvious.
Nothing in Dan Luu’s writing is ever ambiguous. Quite the opposite.
He just writes in the same voice for both researched facts/studies and personal anecdotes. But it’s always very clear which is which, and an informed reader should (IMO) not have a problem differentiating.
The commenter you’re replying to is objecting to unnecessary couching and cloaking of opinion/anecdote. And I think there’s very, very little of that in Dan Luu’s writing — and it comes off as a “matter-of-fact tone.”
I disagree. it sometimes gives me that feeling as well even though I like the guy and his blog. i don't know if it's a cultural thing. i'm no native english storage
People hold beliefs of different levels of confidence. I have some opinions I’m very confident in and others that are merely vague hunches. My level of confidence in something is the signal I’m trying to convey, not merely the fact that it is my thought at all.
It can also be useful to communicate brief info about why I have a particular confidence level. “I had a friend who…” vs “I’ve seen several times…” vs “Several studies have found…”
> when I think back to taking writing classes in school
Yeah I think most of that stuff was bullshit. Or, being more charitable, they were trying to convey a general idea like “don’t overly hedge” with a coarse rule like “never use hedging language”.
I think most writing “rules” tend to be coarse approximations at what good writing really is. Another example is “give your essays a general structure, introduce your main point, etc.” => five paragraph essays.
I hedge a lot in HN comments. Uncharitably, because I’m thinking defensively about how people respond to my comments. Charitably, because it’s more accurate that way.
Long-form content suffers from hedging more that HN comments do. As a writer, you may think highly of your confidence level, and you may want to communicate the difference between a claim that you’re confident about and a claim that you are unsure about, but the right way to do that is to provide the information necessary for other people to come to the same conclusion. If that’s not possible, or it’s not germane, or you’re just busy doing something else, then you don’t do it, and that’s okay. It’s often just irrelevant for people to understand how much you believe something.
The tradeoff I see here is between clarity and precision. If you focus too much on precision, the clarity of whatever you’re trying to say suffers.
You’re going to have to explain more of the argument, because I don’t follow.
It sounds like you are arguing that writers should spend more time spelling things out for readers which aren’t very good at reading comprehension. That doesn’t seem very efficient—it wastes most people’s time, since the writer is spending more time spelling things out explicitly and avoiding ambiguity, all of the readers with decent reading comprehension skills have to spend more time sifting through crap that doesn’t add value to the text (for them), and only the people with lower reading comprehension skills get any benefit. People with lower reading comprehension skills are less likely to be reading text-heavy blogs in the first place!
The following claim may be controversial—generally, there is a tradeoff between clarity and precision, and the right tradeoff depends on the context and what your goals are. The internet, and forums like HN, distort our perception of where the correct tradeoff is, because the people complaining about lack of precision are the loudest. You should be aware of who your audience really is… is it people capable of interpreting claims in persuasive documents? Or is it people who decide to complain on HN?
Giving a confidence level is different from using weasel words. Weasel words make statements weaker or even vacuous. Confidence levels make statements more precise. Here's an example:
statement: Alice stole the briefcase
+weasel_wording: I think it's possible Alice could have stolen the briefcase
+confidence_level: I'd bet 5:1 that Alice stole the briefcase
Also note that I didn't say I wanted Dan to write less confidently. I wanted him to convey more varied confidence.
If there is any doubt regarding the stealing, then in this example i prefer the weasel wording version. The first one is incorrect, the last one you're suggesting an accuracy that you can't really back up, in order to convince someone.
So in this case the weasel wording doesn't make the statement weaker as it shows strength to admit you may be wrong.
Is it more honest? Or is it just more self-centered, or more defensive, or more timid? Are you admitting that you aren’t sure about something, or are you brazenly assuming that people care about how confident you are about things, or are you defensively trying to say the least possible?
Framing this as a matter of “honesty” is, well reductive.
Your confidence is irrelevant to the discussion, unless there’s some particular reason which makes it relevant. A good reader should not hang much weight on your confidence anyway.
> Is it more honest? Or is it just more self-centered, or more defensive, or more timid?
Yes it is more honest. Pretending certainty where you don't have none is lying and pretty often has exact that effect. It is way more egoistic and self centered to act with certainty just so you dont look "defensive" or "timid".
Because then you are exchanging truth for feeling good from appearing strong.
> are you brazenly assuming that people care about how confident you are about things, or are you defensively trying to say the least possible?
This does not make any sense as accusation. You are expressing level of confidence. You can do that independently of whether "people care". It is also perfectly OK to defensively say the least possible. Literally nothing wrong with that.
Aggressively saying maximum, even if you know it is likely half truth at best, is much much worst.
> he'll use the same tone for something he's seen anecdotally and something he's spent a month personally investigating
At least the parts about build vs buy sound extremely familiar to me. Setting up and integrating something you’ve bought often takes as much or more time than building it yourself.
The best argument for buying is that you’d make all stakeholders equally unhappy.
Reminds me of hiring an in-house team of Salesforce engineers to build out and maintain the hugely expensive CRM you bought because you… didn’t want to spend engineering time on a CRM.
The new CTO getting drunk on sales Kool aid and deciding to move many functions to Salesforce was a strong signal of incompetence at a previous workplace.
I read it but it's a mind dump that's not well structure so it's hard to follow. If asked I couldn't give you an answer to the title. I still don't know, " Why it's hard to buy things that work well." He has a lot of citations and footnotes but if it's hard to follow then the essay becomes a bunch or random thoughts which get condensed into the title. I suspect that's why most of the comments relate only to it.
It is very common to see HN commenters only address the title not the contents of the web page, a genuine phenomenon IMO.
Experiment: No titles, just URLs. What would happen.
I have some experience with this as I freqently stream edit HTTP response bodies to make customised web pages using only simple HTML wrapped around only the data I am interested in, no cruft. For example, when I do web searches I process the response, i.e., the SERP, into simplified HTML, e.g., <li><a href=[url]>[url]</a>. I do not include titles.
Titles can be descriptive and helpful, however I find most times they are a distraction. Think about the "clickbait" tactic. It is heaviliy reliant on misleading titles.
> It's pretty depressing how many of the comments are purely answering the title and not discussing the content of the article.
This is a HN pattern as old as the hills. If the headline invites bikeshedding, most people will bikeshed. If the article is long and complex, almost no one will read the article. This one has both.
You are depressed for other reasons. People read the title and connected more with the title than the hard-to-read and dense article. If anything, call it a click-baity title.
It is hard to buy consumer products that work well also. Look at the number of things you order on Amazon, and see how satisfied you are with stuff. All the synthetic material low-quality stuff.
Consumer product build quality and outsourcing quality are the exact same problem. Somebody is making a thing and they don't make it well and the customer isn't happy. There is nothing special about the problem, it's simply hard to build things well. When you prioritize "shipping" over quality, you make crap. If you try to avoid the difficulty of quality by hiring somebody else to do it and they also prioritize "shipping" over quality, you buy crap.
Re: "getting a worse result even when they pay more than doing it internally", they either paid the wrong people, or they were trying to solve the wrong problem, or weren't good at using the product. If you're making Toyotas and you buy your parts from CheapPartCo, chances are good you end up with a crap car. If you buy your parts from Denso, at least you have the chance it will turn out well. If you take the Denso parts and assemble them terribly, you still end up with a crap car. And if you shouldn't have even used that part because your overall design was crap, you also end up with a crap car.
I'd say it's more likely the average person/company/etc will not make something well. Not only does it take more skill and hard work to make things well, they end up more expensive. If things are made well they probably took a lot more work to make them well, or the people who make things well are in higher demand. Show me a company that pays a premium for good vendors and full training for all their staff and holds back products until they pass a Steve Jobs-level of quality engineering, and I'll show you a company whose products work well, and probably charge a premium. (The only exception I know of is Toyota, because they are crazy enough to literally stop a production line just to troubleshoot a tiny issue. Their focus on quality has led to efficiency which reduces cost and increases production. But this is all Lean 101)
And this isn't a tech-specific problem because everything in the world has quality issues. The reason GM couldn't make a half-decent car while Toyotas were rock solid for decades was simply working harder on quality. You have to work hard to make something work well.
In what way is the article about that? How is Apple getting a better chip by learning to do it themselves, because cpu manufacturers are trapped by misaligned benchmarks, just a trivial statement about supply and demand?
It isn't about supply and demand at all, but about another, even more basic, economic concept: incentives, and a second relatively modern economic concept, information asymmetries.[1] Over-summarising and over-simplifying:-
"Third-party service providers are incentivised to minimise their costs. If their clients cannot assess the quality of their service and/or end-users cannot reward or penalise them based on the quality of the services, then quality is poor, because it is cheaper to not provide services and later argue about it than to provide the service. Third party deficiencies/misbehaviour get so bad sometimes that there are no gains from trade and occasional losses.
"Internally, inside companies, the incentives may be to act similarly to third-parties, or they may not. This depends on "culture"."
And here the essay would be improved with an analysis of the compnents of "culture" that matter. It talks about trustworthiness of company leadership, but that is only one factor, surely.
1. Akerlof's "Market for Lemons" paper is the seminal one here.
Perhaps the strategic decisions of one the largest and one of the most complex organisations in human history cannot be explained by simple reference to undergraduate microeconomics
"Variants of this idea that I frequently hear engineers and VCs repeat involve companies being efficient and/or products being basically as good as possible because, if it were possible for them to be better, someone would've outcompeted them and done it already"
Reminds me of the old joke -- an Economist walks past a $100 bill on the sidewalk, but doesn't bother to pick it up. He turns to his friend and declares smugly: "If it were real, someone would have taken it already."
Because we have unreasonable expectations. Making things work well is hard, and thus expensive.
Take a chair, for example. Anyone can make a chair: just get a bucket, turn it over, and sit on it. Boom; backless chair. Want a backrest, arms? Get a 2x4, a saw, a drill, and some bolts. An hour later: Boom; chair.
You want a chair that works well? How do you even define works well for a chair? Do you need cushioning? Do you need to swivel? Do you need ergonomics? Do you need to sit in it for 15 hours? Do you need it to cost less than $1,000?
To the featured article's point, it's gotten so hard to be able to judge which products will be any good at all that buying expensive things expecting lasting quality becomes a huge gamble.
There's also the huge mental barrier of getting over the sunk cost fallacy. If you buy an expensive thing and a part of it breaks in such a way that it still remains mostly functional, it's really, really hard to admit that it should be replaced. You end up living with a broken thing for a long time, just because you're mentally depreciating the thing.
I've been through this with a lot of furniture. It doesn't seem to matter whether I pay $100 or $500 for a bookshelf, it's a crapshoot whether or not the listing lied about it being made of solid wood and not particle board, or whether the screw holes will align correctly, or been tapped correctly for the screws to actually engage without crossthreading, or even use proper screws and not those damn "lock"-bolts (that never actually lock, so given enough time they always work themselves loose).
My wife and I built our children's bunk bed. It was an expensive, time-intensive, physically painful endeavor, partly because this was the first time we were building something so large. It's not great. But it easily matches or exceeds the quality of any of the other furniture we've bought. I'm very proud of the work we did, but at the same time, I'm furious that I can't count on just buying things.
It's also gotten hard to trust people's recommendations on things, which goes back to the featured article's comments on cultural expectations. Most of the people I know just live with broken furniture. They think IKEA is great stuff. It's not. It's just that it's so impossible to find anything better that you might as well buy IKEA.
I might switch to only buying antique furniture, i.e. use survivorship bias to my advantage. Find the stuff that has survived taking a beating already.
an aeron chair costs about $2000 dollars (give or take). The construction is sturdy, so i expect it to last at least 10 years, if not 20 (you take care of it etc).
So the amortized cost is some 100-200 dollars a year for an excellent chair. If you buy a crappy chair, i bet that the cushion starts failing after a year or so of constant sitting. So you'd probably replace it yearly, or suffer a bad chair for a few years before replacing.
I bet that most people would choose an excellent chair, if they could guarantee themselves the 10 yrs of good operation and comfort.
But there’s a corollary - if you do not need the thing to last, you may be better off with the cheapest one you can find (which may be used - and which may be better than new).
This is the “but it from harbor freight, if it breaks now you know you use it enough to make it worth getting a good one”.
I consciously adopted the "poor young constable Vimes" strategy for footwear last year. I like, and would probably prefer, quality shoes -- but it's such a crapshoot what those are nowadays that I just can't be bothered to buy a pig in a poke yet again. I mean, brands that used to be OK when I was younger... Mostly aren't any more.
So last year when Lidl had trainers for sale at 30 € (or was it 20? 25? 15?) I bought five pairs. Still on the first ones, so definitely financially ahead -- "better" brands haven't lasted any longer than that in my recent experience.
> if you do not need the thing to last, you may be better off with the cheapest one
Well, it's not even sure if you take resell value into account. If you need your good chair for one year, just buy it $2000 and sell it used for $1900.
As a random anecdote, I own a steelcase gesture that runs about $1,100. I also own a 70$ chair from costco and I can "barely" feel the difference. I have all my settings configured correctly. I really don't buy into the the expensive officechair hype. I think it's a level of diminishing returns.
I think a $200 office chairs is about what anyone needs. That'll get you 90%+ of the way.
Where would you say you sit on the physical-dimensions bell curve?
I'd hope that the more expensive chair has better adjustability to suit a wider range of critical dimensions than the cheaper chair. But that may not matter for a lot of people.
I have a more expensive herman miller embody. It... doesn't really have much to adjust. Just the shape of the back. That's kind of an irritating aspect of it to me.
Same for my $70 IKEA chair (which also survived 4 moves). I only stopped using it because my job let us take home the office chairs once work from home started a few years ago.
Well, personally, I have a lot of durable chairs, but I love my Aeron because the adjustability got me out of RSI. I don't mind replacing the chair every year if it would help me avoid RSI. I haven't had to, but I wouldn't mind. Durability in chairs seems to be easy.
I've been out of action due to RSI a couple of times in the past 7 or 8 years. I haven't worked out what the cost was but, sure, it was multiple times the price of a premium chair.
An Aeron is nice but plenty of other chairs last way longer than a year and are maybe a fifth of the price of an Aeron. You also have to consider for a lot of people there are other products that will give them a much better pay off.
When I interview, I like to look around and see how many short people work there. A place that picks up the short people other companies don't must be recognizing their value.
I would like to favor places with more women in tech jobs, but generally just don't find any at all, anywhere. Everywhere I have worked they were always trying to hire more women but never got any applicants, or couldn't hire the ones they found.
Huh, interesting. I would second MtF transgender. I've worked at places that will interview so they can appear what you might now call "woke", but then not hire based on their dislike, not skills or personality.
And there are so many MtFs in tech. I'd love to see statistics, but it feels skewed.
This is one of those things I Can't Say in most places, but (unless you transitioned in elementary school) the MtF experience in tech and the cis female experience are totally different, and they (the MtFs) know it, because so much of our work is done online where you can be whatever gender/sex you want. I know how people reacted to me when they thought I was male vs. female.
Age plays into it, too. Everything was fine before I went through puberty, but a lot of young techie men have a really hard time with their sexuality and take that out on their female colleagues in various ways. (I've also seen some gross reversals in female-dominated professions re: issues with men, so...) Now that I'm in my 30s and the men in that insecure/figuring things out age range aren't interested, it's better again.
I don't know what it's like now as I've not worked in an office for 15 years, but back in the day being a woman in a tech company must have been horrible. Most tech guys I knew had zero experience with interacting with women and therefore would unconsciously act in the most creepy manner towards any and all women. Also, the women would get no respect as they were automatically assumed to have inferior technical skills.
My situation is interesting because I haven't worked in tech because of my experiences pre-22/leaving college. I was raised by geeks and carted off to computer shows on my dad's hip, started programming in elementary school, etc. So I spent a lot of time in tech spaces but not professional environments (I mean, this was also the 90s and early 00s so there were just fewer professional tech spaces).
Everybody was super chill until I was about 12. Then things got very ugly, very quickly. Anytime it was known that I was a female teenager, it was terrible, especially as one that was (at the time) better than some of the males. I'm also gay, so at the time I was stuck in this horrible situation of often being surrounded by dudes ~5 years older than I was who were a.) very creepy and b.) might be homophobic (including violent because again, 15-20 years ago) if I came out. And of course all the ones who are nice until they learn you'll never date them, or the ones that viewed me as a challenge/trophy of some sort.
Very weird to have aged both in and out of creepy male fuckability: Most of the creeps don't go for women over 25 (and I finally look like I'm in my mid-20s) unless they're elderly, and I can beat up most 75-year old men, so that's less scary.
I have some sympathy for the boys as an adult, because a lot of them were bullied a lot, but...dudes, you think I had it better? I had to deal with their bullying and yours. Thanks for that.
I don't have statistics at hand, but folks on the autism spectrum are more likely to be queer or otherwise gender non-conforming. So if there are more autistic people in tech this certainly makes sense.
Shhh men and women are exactly the same, have the same desires and strong points, men that decided at 25 that they are not men can magically become women, there should be a 50/50 split in prestigious jobs (but not construction or underwater welding), and if you disagree with any of the above you don't believe in equality and are practically a Nazi.
"Meanwhile, in some Asian countries, like Taiwan and Vietnam, people mostly complied with lockdowns when they were instituted, which means that they were able to squash covid in the country when outbreaks happened".
And there's a huge difference between quarantining a small island and quarantining a large country whose leaders purposely don't enforce the borders and actually allowed in people who tested positive for COVID with no enforced quarantining: https://news.yahoo.com/dhs-dropped-40-000-covid-190800213.ht...
The people who said lockdowns would be pointless were right. History has proven it. It's time for people like Dan Luu to move on. There was never any way the world at large was going to be able to organize a meaningful lockdown policy. The lockdowns were a waste and likely caused more deaths than they saved lives.
I think these sorts of failures of 'common sense' are common among people who have studied economic theory. Economic theory often does a decent job at describing steady states: where the asymptote will eventually land, but does a very poor job of describing the chaotic middle steps, IMO.
If you chop a 10 huge trees down in a forest, an economist would tell you "There should be 10 300-year old trees here". That may be true in 300 years, but the procession of species in the 300-year interim will be much more interesting, and understanding that tells you much more about ecology than knowing the eventual result. And the forest may not last 300 years before conditions change so significantly that it invalidates the prediction.
Similarly, the efficient markets hypothesis really breaks down in the modern era, because market conditions don't ever experience an unperturbed steady state. Technology, politics, availability of expertise, and shifting social norms rearrange the foundations more frequently so equilibrium is never reached.
Any long-term theoretical predictions on quality or efficiency of CRT monitors or horseshoes or fondue pots converging to an efficient ideal state were nonsensical and doomed from the start: the market conditions around them changed faster than the market is able to adapt.
Maybe I missed it, but I don’t think Dan even attempts to answer the titular question! The article, while quite good, is just a very long list of things that suck, and anecdotes about people who tried to make things suck less.
I’ve thought about this question very often, and I think the answer does come back to the market efficiency argument that Dan ridicules at the beginning. I think in many cases, what the market provides is actually about as good as what the market can sustain.
Sure, most products have obvious flaws, and maybe you could fix them. But could you build a successful company solely on the basis of addressing those flaws? Sometimes, yes. Some products do get displaced by better alternatives. But can you expect that process to happen reliably, for all types of products, at all times? When you put it that way, I think it seems silly to expect the answer would be yes.
I went out of it with the conclusion "The world is broken, deal with it"
Just to get caught by that line in the culture part:
> If you read these kinds of discussions, you'll often see people claiming "that's just how the world is" and going further and saying that there is no other way the world could be, so anyone who isn't prepared for that is an idiot.
I'll defend my own claim as not trying to put the blame on the victim, it's nobody's fault most of the time, and we get to see some progress, sometimes. People should enjoy the stuff that actually work and praise those who make real efforts to make it less broken.
I've seen lots of products and services that are materially similar, if not identical, with vastly different prices, and the companies selling them are equally successful. Sometimes it's the company charging more that seems to be doing much better. I don't think there's any particular efficiency advantage. I think it's just a matter of marketing.
Before getting into software I did fair amount of construction. Once you know how a job should be done you never look at the resulting product the same way. From drywall to sidewalks, everywhere I look I see that someone rushed or didn't care. I wish it wasn't hard to find quality craftsmanship, but it is. Ignorance is bliss, but I lost that long ago.
> Amazon eventually solved this problem by having their own delivery people (and Apple has done this as well for same-day delivery). At scale, there's no commercial service you can pay for that will reliably attempt to deliver packages.
It's funny, I have actually observed the phenomenon he described in the same company. Safeway in San Francisco either uses their own delivery drivers, or they subcontract to DoorDash depending on the time of day of delivery. If I order with Safeway delivery drivers, they will happily deliver all the way to my apartment door inside the building. If I get a DoorDash driver, the intercom will magically be broken 100% of the time. It's a perfect comparison.
> [It] involve companies being efficient and/or products being basically as good as possible because, if it were possible for them to be better, someone would've outcompeted them and done it already.
Must be a generational thing but I don't think anyone i know (ie under 30) genuinely believes this anymore.
I think the more correct statement is that products are as good as profitably possible and no better. If the margins on a better product aren't large enough to justify spinning up a new business just to sell it, the existing product will remain as it is indefinitely. To top it all off, consumer expectations can be surprisingly low. It's often the case that no one even wants or knows to want a better product, let alone being willing to pay for one.
I think you're right--particularly w/ regards to consumer expectations being low. I'm a part of this problem as well--when making purchases, price is almost always the deciding factor.
It's also worth noting that "as good as possible" and "as good as profitably possible" aren't just two flavors of the same principle w/ regards to quality--they're directly at odds. I'll prove this (with scientific finality) via an anecdote about toys.
My kiddo was gifted a little Fisher-Price school bus that was made in 2005.[0] It's pretty neat! As you roll the bus forward, some internal mechanism causes the driver to turn left and right (making it look like they're turning the steering wheel back and forth), and the passenger seats to move up and down (a dope allusion to the classic kid's song, 'The Wheels on the Bus', in which, as the wheels of the bus go 'round and 'round, the people on the bus go up and down). It's clear that this product was lovingly designed, with playful curves and playful details, like the face of the bus. The wheels not only have hub caps, but the shape of the tires even mimic a pneumatic bulge! The space inside of the bus is even big enough to accommodate human hands!
More recently, she was gifted a newer version of the same toy.[1] It's smaller in every dimension, and flimsier. The slick design is gone, replaced by simple, boxy one. The little things that I loved about the old bus are gone--the wheels are now a single piece of plain molded plastic. While the wheels of the bus still go round and round, the people on the bus, alas, do not go up and down.
It's fine, as far as toys go--the kid probably doesn't notice a difference. But it's clear that the guiding principle in the evolution of this product is not only not quality, but something that is directly in conflict with quality.
And then the operative problem there is that there are entire industries devoted to creating knock-offs that are slightly cheaper, much better marketed, and don't work, which end up replacing the good-as-profitably-possible product because they're more profitable and marketing subsumes consumer feedback.
Great article! As a consumer that cares a lot about buying long-lasting quality products, here are a few resources:
Project Farm (youtube.com/projectfarm) - Great independent reviews of tools
Mcmaster.com - Higher quality products than you can get at your local hardware store and excellent customer service. When I've gotten the rare incorrect or damaged item, 1 email gets me a refund and a replacement overnighted.
Wirecutter - Not as go-to as it used to be in my mind, but great for background
ConsumerReports - Check if you can get free access through your local library website
ReviewMeta (reviewmeta.com) - Analyzes Amazon reviews for authenticity
All of these have problems but they are still good resources. Any others I'm missing?
McMaster also has the best ecommerce tech stack I've ever seen. It should be the case study. It is incredibly fast. The checkout flow is different from anything I've ever used before, it's so fast, you don't even feel like you should be done. And, if you type plain text in the search box, it will translate that to a parametric search (for example, try '1/4"-20 screws', and you'll see that it takes you to the "screws" page with 1/4"-20 selected as the thread size).
If there's a such thing as rockstar software engineers, the people that made McMaster's site are it.
At my workplace, we have what I call "McMaster syndrome" - employees design things around parts available on McMaster because that's the easiest website to order from.
As a frequent business user of McMaster I find it serves a very particular purpose - everything on there works and does not break. However, their markup is massive. I once placed an order large enough that they redirected me to their supplier and found that they had over 100% markup on their orders. You're paying this markup so that McMaster verifies the quality of what you're buying. This makes a lot of sense for a business - you're not wasting employee time on reading Amazon reviews. I doubt it makes much sense for most consumers.
I'm continually disappointed with the quality of fasteners bought from Amazon. What you get from McMaster is much better. The problem, I guess is that Amazon is too low quality and McMaster is too high quality.
I also don't think they mark everything up 100%. I have built a lot of stuff out of T-slot extrusion ("80/20") and their price seems exactly the same as buying from 80/20 directly. I guess there is probably some Aliexpress vendor cheaper than the brand name, though.
Seems like it, yes. I've seen similar arguments made regarding foreign labor. "I value people of all countries the same - it's just that only <some nationality> seem to be willing to work for that value".
You can paint it negatively but hiring them when others don't also improves their salaries - if everyone else followed their lead there'll be no inefficiency and they won't be underpaid.
Not because they are women though, but because they had trouble finding work somewhere else. So the company had a bigger negotiation power. That's just capitalism, supply and demand...
Originally yes, of course. But that's not the fault of the person that actually hires them. If you are the only company that hires everyone without any discrimination, you just have a bigger supply of people.
> Some commonly repeated advice is that firms should focus on their "core competencies" and outsource everything else, but if we look mid-sized tech companies, we can see that they often need to have in-house expertise that's far outside what anyone would consider their core competency unless, e.g., every social media company has kernel expertise as a core competency.
I wouldn't call Twitter or any other unicorn a "mid-sized tech company".
I believe the larger a company is the more it makes sense to do stuff in-house. You need a certain amount of people to focus on your core competencies, but once you have that and still have more money than you can ever spend, it makes sense to spend it on doing stuff in-house, as that might lead to better quality and/or lower costs and removes the risk that an external supplier might not be there next month anymore.
Twitter is a mid sized tech company though, with a mere 5,500 employees. Facebook and Google are an order of magnitude bigger, and half the size of Uber and Netflix. It’s also a tiny bit smaller than Doordash and AirBnB by comparison.
> Twitter is a mid sized tech company though, with a mere 5,500 employees.
I guess that depends on the definition of mid sized company. I don't know if there is a fundamentally different/better definition, but Gartner defines a mid sized company to have up to 999 employees and $1 billion of annual revenue [1]. Twitter has way more employees and revenue than that.
Modern western society has not built the social capital necessary to cope with its material complexity. It's built on an ethic of grifting, in the large or in the small. See the part of the piece where people think you deserve what's coming to you if you aren't assuming everybody's out to f* you.
The appendix is interesting because I have noticed that a lot. I think it might be because I'm on mostly American sites but it seems like any sort of "why can't we do this in America?" always has some sort of explanation why not. No one ever considers that the answer is "we could, actually, if someone chooses to do it". For instance, gigabit fibre in San Francisco was this impossible task if you ever asked someone. It would cost too much, SF is too old a city to trench, America is too big, etc.
In practice, all it took was a dude called Dane Jasper deciding that the barriers were mostly regulatory and he could beat it and boom! Sonic covers so much of the city in $60/mo symmetric fiber. I think people will still explain why it's impossible in SF.
Occasionally, and I am saddened that I haven't bookmarked these for later amusement, someone will ask "Why can't we do X here?" and people will come up with an explanation for why X would never work in America, etc. while I'll be sitting experiencing X here.
Perhaps that's the thing. Human beings have a very natural status quo bias and we have a very quick explain-why-this-is bias. So if you tell someone that Y is the case, they can come up with a posteriori explanations. But that's why good science is hard: you have to make a priori claims and then subject to hypothesis testing. And in normal conversation that is really hard.
I don't claim to be immune to this. But knowing it is a flaw is a better position than not knowing, I hope.
> In practice, all it took was a dude called Dane Jasper deciding that the barriers were mostly regulatory and he could beat it and boom! Sonic covers so much of the city in $60/mo symmetric fiber. I think people will still explain why it's impossible in SF.
It often is impossible to get past the regulatory burdens in SF, although sometimes it's just because the supervisors ask you for bribes.
He talks a lot about how other companies are incompetent or inefficient and says that this is why it is often better to build in-house. I think the reason this doesn't happen to the degree that would be most efficient is people assume the counterparty business is acting in long-term self-interest as a single unit (i.e., a business with a reputation to maintain). In-house, there's almost always someone you can get in touch with has the ability to enforce this behavior on your counterparty - when you're buying something externally, that's not the case. There are so many examples of large, well-known brands that made a name in quality and then sacrificed their reputation to cash in.
This is a really good essay, but kind of depressing.
Re: professional services ... I think there is some kind of sweet spot where professionals stop focusing on delivering a good service and start focusing on their brand. Also, some people realize they are not actually a very good accountant/developer/lawyer and focus almost entirely on their brand. Trying to discern if your potential professional is one of those two types of professionals is the hardest part.
> Often, people will say things like "I would never get into that situation in the first place", which, in the circumstance where someone is driving past a parked car, results in absurd statements like "I would never pass a vehicle at more than 10mph", as if the person making the comment slows down to 10mph on every street that has parked or stopped cars on it.
One could write an entire series of posts on the topic of situations you would not have avoided.
Just picking a passage from the top of the article:
> There's a vague plausibility to that kind of statement, which is why it's a debate I've often heard come up in casual conversation, where one person will point out some obvious company inefficiency or product error and someone else will respond that, if it's so obvious, someone at the company would have fixed the issue or another company would've come along and won based on being more efficient or better.
---
This sounds like the question whether the market mechanism that ought to remove grave inefficiencies is actually working because there appear to be many examples where it looks like it's not.
My explanation for why the market actually works and we can _still_ see those inefficiencies:
Complexity.
The reason why - even though the mechanism of the market making efficiency a necessity (to not be replaced) is working correctly - we often see inefficiency, is that a complex world enables niches for inefficient unnecessary complexities to exist, where they can consume resources without being easy to remove.
The more complex the environment, the higher the number of possible/existing additional overhead and inefficiencies.
I think complexity is definitely part of it but it is also mis-aligned incentives. The person running an assembly line has a vested interest in changing nothing. Everyone knows how it works and consistency sometimes is more important than raw efficiency. If they change something and if breaks the machine or they have a month of lower productivity, they don't get a promotion and/or bonus. Why would they risk it? Plus, people generally hate change.
You actually have to have someone whose job it is to look for inefficiencies. Someone who goes and talks to everyone on the line because they are the ones that really know. Someone who also recognizes that a 5% improvement in one area may not be worth the risk but a 1% improvement somewhere else may be. Just like staff scheduling: do you schedule the bare minimum of staff? What happens when 3 people next to each other all get the flu? Now where are you? Are you better off normally over-staffed and constantly cross-training? Most companies make money despite their operations. This is where Amazon has taken over the world, they have that as a core goal and corporate focus.
> I think complexity is definitely part of it but it is also mis-aligned incentives. The person running an assembly line has a vested interest in changing nothing.
Is this really a case of mis-aligned incentives though?
The person running the assembly line has an interest in the business model of the company not being destroyed, because their salary depends upon it working. So they'll try to prevent the worst (to the business model) from happening.
Risking to break some important production process for a lousy low percentage potential gain in efficiency sure doesn't seem to be in the interest of neither the individual person, nor the company, nor the customer.
As you write: the risks and benefits need to be carefully weighed perpetually .. simply an ongoing optimisation.
Reading this I get the feeling that Dan is missing a major selection bias: he has extremely high standards and is very competent, and tends to be friends with people who also have very high standards and who are competent.
It stands to reason that he and the people he interacts with are going to build things of higher quality than whatever is commonly available.
> For example, in my social circles, there have been two waves of people migrating from iPhones to Android phones over the past few years. Both waves happened due to Apple PR snafus which caused a lot of people to think that iPhones were terrible at something when, in fact, they were better at that thing than Android phones.
I wonder what Apple PR snafus is that?
>Amazon eventually solved this problem by having their own delivery people (and Apple has done this as well for same-day delivery)
Apple employs they own people for delivery ? That is news to me. And I cant seems to find anything on Google to confirm this.
Otherwise the article sums up, The world is largely run by Pepla, people who cant tell the difference between Pepsi and Coca Cola. And the Market can stay irrational longer than you can stay solvent.
What Dan is describing is the triumph of politics over everything else. Deng Xiao Peng diagnosed Chinese society during the Mao era and said the very real inability to contradict the party line was what ailed China. People were not allowed to believe in reality but had to swallow every line from the central committee. This is the Seek Truth from Facts speech that marked the opening of China to the modern world. Corporate America is its own Potemkin village where political lobbyists and legal protections keep them alive despite massive corruption and the harm done to customers.
> You might think that, if a single person can create or maintain a tool that's worth millions of dollars a year to the company, our competitors would do the same thing, just like you might think that if you can ship faster and at a lower cost by hiring a person who knows how to crack a wafer open, our competitors would do that, but they mostly didn't.
He mentions a small chip startup he worked at for these examples, but I couldn't find it anywhere. Did those companies beat out their competitors and win their market?
Seems like his point is that "We did this optimally and competitors didn't even though you'd expect them to", but if that company didn't end up being the best one it's not really proving the overall point
You miss the point though, as the problem is two-sided. The consumer struggles to find a quality product because they simply don't know how. The producer also struggles to produce a quality product, because it's hard to know what the consumer really wants. The cheat is to sell brand instead, to market quality regardless of if it's actually there.
I would argue Apple is an example of a company that started with quality, established brand, and has been coasting off brand for a long while.
If buying and selling were a chess game, the ELO of the consumers would be very low, but the producers ELO is not much better, as they just learned a tricky opener and nothing else. Of course the real game is not two player and open information, but this only makes it worse for the consumer, and easier to fall into marketing schemes for the producer.
I use chrome and you have to enable the special flag for that and I'm not doing that for a single site.
I think it's only needs a really small amount of css i.e. some padding/margins a slight change in the background colour and some better spacing that's all probably. Five minutes worth of work and would add virtually nothing to the page weight.
I suspect whomever Dan Luu is they don't really give a f about my opinion so I'll stop now.
Lack of professionalism and following through with promises at scale sound like massive reasons for this. I say think of ant colonies, especially in the work from home era (I'll have some grace to beg the sensible people lets ignore covid rather than the authors ranting).
Ant colonies are hugely successful yet estimates are only about 1/3 of workers actually work. There is no 'ant-police' there is no 'don't work no food', this is just the way that it ends up.
In human situations I can count from experience being one of that 1/3 that work a lot even out of hours or when left un-monitored I know I'm in a minority because I'll throw up an experimental full stack deployment in a week with new toys/tooling/features and run into meetings to listen to someone winge about having difficulty in deploying a new ssl cert on an nginx box entirely because the instructions are only 3-4 months out of date.
This used to drive me nuts but I've come to accept humans at scale are mote like ants. This even applies to the good ones. They're normally only working 1/3 of their time, hence why project management needs to more accurately account for this rather than thinking they can encourage the average person to be the super-geek who has fun turning on the light using https-over-dns-over-shoe-string one weekend because they simply wondered if they can.
People unfortunately then take this as an insult that I'm comparing to them as ants.
My response is, if you understand this correctly you've identified you're a normal rounded human being, congrats, enjoy this and go contribute to society in the way that makes you special.
If you are insulted that I've identified you're working 1/3 of your time, either a) you're striving to be a super-nerd when you're not, please consider changing, this isn't healthy for you, or b) you are actually one of these super-worker people and get annoyed at the rest of the system like me, try moving into a career where you enjoy your work rather than trying to be the next millionaire, unless you love working in the markets, in which case, go for it and have fun :)
survivorship bias is real when people say "they dont make things like they used to". Yeah your armchair from 1960 was built well becuase... it was built well. There were hundreds others that were complete crap
My theory is that it's because the end-user is less familiar with how household goods and machines work than in prior generations. With chips in everything, it's hard to tell at a glance what's good and what's rubbish; you can't kick the tires. Even if you could, there have been a couple generations where buying stuff new has been cheaper than repairing it, so knowing how to sew or fix a dishwasher isn't the value proposition that it once was--and that stuff has become as cheaply made as possible as a result.
They can only get away with it because they know most of their customers wouldn't do their own repairs anyway. In 1950, they'd say "Screw you then, I'll make the part in my garage."
Because more than a few people buy something because of it's (low) price. When something is inexpensive, it often has a shorter lifespan and you tend to replace it more frequently than you would a more expensive item.
Admittedly 1) not everyone can afford something pricier, and 2) higher cost doesn't always mean better quality. On top of that, if manufacturers made things that lasted forever, they'd go out of business quickly. Who'd buy a replacement, or something newer, if item X is still going strong after 10 years or more?
Like politics, marketing thrives on "noise" and the loudest voice tends to win, not the best candidate
If you have a lesser product, marketing rolls out FUD to combat competitors and then it doesn't matter if you have a better mousetrap, no-one will believe it
For example most 5-star products on Amazon got that way because they often secretly give out tons of free product (or even cash) for positive reviews. Amazon and bot analysis can't tell a real human written review was secretly paid for. Noise "wins".
It comes down to information having a cost. Noise is everywhere in the market. For instance the lawyer who spends the most money on advertising on the radio might not be the best guy to do your case.
Information also decays. Yesterday's best burger in town may not be today's.
EMH is basically wrong. Like the pre-relativity concept of absolute time and space, it sounds like it should be right under certain conditions, but unlike the physical analogy we live in the inefficient information world mostly.
I always enjoy danluu's blog; one of the few I have bookmarked and go randomly read from time to time. Some random comments on out-of-context snippets:
> For example, we tried "buy" instead of "build" for a product that syncs data from Postgres to Snowflake. ... Despite being widely recommended and the leading product in the space, the product has a number of major design flaws that mean that it literally cannot work.
This is nearly always my experience with buying software, especially software I know I can write myself. I always run into massive pushback if I propose writing such software. It comes across as too risky -- what if I get bored, what if it's actually too hard and I can't do it, etc. What people never think about are the benefits; full knowledge and control over the process.
I have bought so much software and it's made me sad every time. You're always promised the world before you sign the contract, and then you sign it and none of the things that were supposed to work work. Support tickets take days to be resolved; passed around like hot potatoes. In the time it took to get someone to reply to my email, I could have just written it myself.
Many years ago I was looking at Istio. I wanted some enhanced Kubernetes ingress support, and Istio checked the checkboxes; what I needed now, and room to grow. But it sure had a lot of detractors on the Internet, and a ton of alarming-sounding open bugs. I thought about rewriting the parts I needed from scratch, and went to see what parts of their code I could steal.
I looked at the core XDS implementation, and it was completely wrong. It didn't implement the protocol correctly at all, conveniently swallowing crucial errors that probably explain most of the filed bugs.
So, I decided to write my own version and a few days + 2000 lines of code later, I had the parts I needed and they worked perfectly. I didn't have to scrutinize the docs, I didn't have to file 30 bugs. I just typed in some code, deployed it to production, and ... it's never bothered me since. It was risky, but it worked out better; I can explain the entire system to anyone with a question and I can add whatever features I want. It was way better than buying. (I'm equally happy with the two version of a single signon system I wrote. It works exactly how I want it to work, and costs $0 month for an Enterprise contract. Savings!)
There are counterexamples, of course. Once upon a time, I needed to learn CAD. Instead of writing my own software, I just bought Fusion 360. I didn't know anything about CAD at the time, so there is no way I could have built something better. (I probably could have implemented the "don't crash for no reason" functionality that eludes their developers, but wouldn't have even known to implement sketches or geometric constraints. They just know the domain so much better than me, there is no way in the world I could ever have made something better.)
> Since then, many people have changed their opinion to "having ever locked down was stupid, we were always going to end up with endemic covid, all of this economic damage was pointless".
This happens all the time, not just with pandemic restrictions. The pattern is that people want to cure X, and suggest remedy Y. People try remedy Y for a while, get bored, and then declare remedy Y a sham. It's universal across fields; "I tried running but didn't lose weight", "We tried writing more tests but it didn't increase reliability", "We tried sheltering in place but the disease didn't die", "I took antibiotics for a few days but the strep throat came back". The problem is not that remedy Y is ineffective (though it certainly can be, that's where this heuristic comes from), but that it wasn't used for long enough. I think changing habits is one of the hardest things humans can do, so we are quick to declare defeat when success can only be measured over a long period of time.
> You could imagine services would, like Amazon, request a photo along with "proof of delivery" or perhaps use GPS to check that the driver was plausibly at least in the same neighborhood as the building at the time of delivery, but they generally don't seem to do that?
Amazon does this sporadically. Sometimes I get photos, sometimes I don't.
> having ever locked down was stupid, we were always going to end up with endemic covid
COVID lock down/shelter in place orders were never intended to avoid endemic disease altogether. They were about slowing down its spread just enough that the healthcare system would remain functional while the disease became endemic. And they mostly worked OK for that.
One of the issues with how it was handled was the handwavy and contradictory information that many governments and public health czars circulated: that handwashing is most important, that masks are effective or not effective or only some kinds are, that the lockdown will be "two weeks to slow the spread" and so forth.
It was the public health equivalent of "take two asprin and call in the morning". While there were successes for sure, clear and consistent communication was not among them.
A lot of that came right from the top. In the UK we had a prime minister who said he was shaking peoples hands in hospital and attending his mothers birthday party to nearly dying within a couple of weeks.
There’s a difference between “plumbing” software and “user” software. If photoshop or autocad become crap something will replace them (the pressure Dan talks about).
But most enterprise software is plumbing - that you figure out how to use “well enough” and mostly leave it alone to do its thing. That software has little pressure and often ends up being basically bypassed.
The users of enterprise software are captive, they can't choose another solution or supplier without changing jobs. Mostly they hate it and IT in general which I think is at least partially the reason for the success of SaaS.
Dunno which it's more of, tragic or hilarious -- probably both, about fifty-fifty -- how they usually don't realise that they're just as stuck with their SaaS supplier.
One effect that might also be at play here is, assuming that excellence follows a Pareto distribution:
If your team is very good, i.e. on one end of the distribution, most products will be made by teams that are average, i.e. not that good. In that case it is easy to notice flaws.
If you are smart and knowledgeable enough so that you could design/create/implement the products you use easily, you will see room for improvements.
Because there is a power-relationship between the manufacturer and the consumer. Initially, a small amount of power is with the consumer, but at the moment of purchase the power shifts to the manufacturer. Since the power concentrates at the manufacturer, the latter will end up with massive amounts of power. This results in products that are in many ways shitty for the consumer.
One explanation would be that if you are an "extreme" user, you are bound to find fault with all the offerings on the market. The need to tweak/customize/adapt naturally arises when you have reached a certain level of sophistication and when it does, you can only be satisfied by a product/service appropriately tweaked/customized/adapted.
I foolishly bought one off the top of the best-seller list at Amazon a couple years ago. It was like $35 and came with a little card that said "Hey, if you give us a 5 star review and send proof to SHADY_EMAIL_ADDRESS@gmail we'll give you a $25 Amazon gift card!"
It did work reasonably well, but if you failed to clean it and fully dry it out within 6 hours of using it, it would get an INSANE mildew smell.
Amazon of course didn't bother to do anything about the company when I tried to report them, and they seem to have disappeared/rebranded since then
My teakettle is moist for hours, some of my dishes are moist for hours, they don't smell like mildew. I'm in the bone dry south bay in a warm house
My guess is water is getting trapped somewhere in the device's Chinesium crevices, or maybe they found a way to make plastic cheaper by making it susceptible to mold.
Ours just runs until dry. Pretty sure it’s ultrasonic but it’s cheap and there’s never been a smell. Those are the joys of living someplace you really need a humidifier. Water just disappears.
You have to add bleach or humidistat to evaporative ones or keep them running full time. You also have to replace the media once a month or so. The steam ones are better imho.
One of the major issues with humidifier design is that whatever you put into the humidifier gets into the air, which sounds obvious but is an issue. We don't want to use chemicals to keep the paper wick from getting moldy because then we are breathing that. Or with ultrasonic they throw up mineral dust into the air. We end up just buying a lot of paper wicks.
I bought the one he didn't like (the electrode boiler) after watching that video (and his tear-down video [1]). (I have hard water and have had trouble with ultrasonic humidifiers.) It works great! You do have to descale it regularly to keep it going.
Or just get a hot mist humidifier, which does this except:
1. Electronically (ex: you can do this in your bedroom directly, or anywhere you need the warm-mist)
2. With safety measures (its still dangerous, but they minimize the amount of boiling water to minimize the possible harm. Better than an entire pot of boiling water, but still half-a-cup of boiling water can severely burn you still)
3. Probably as cheap as the pot you were using to boil water anyway. Warm mist humidifiers are like $35.
Unless you have a gas stove, in which case prolonged boiling could lead to indoor air quality issues (unless you turn your fan on - but then you're drawing in cold air.
If you have an electric stove, boiling water will be less efficient than evaporating water and letting your heating system make up for the loss in sensible heat unless you have an electric furnace.
1.1 Lack of buyers' expertise (tfa goes into this a lot)
1.2 Long timescale to assess quality:
that is, a lot of types of 'poor quality' aren't apparent until long after the sale; so even if you polled every buyer once a week, if sales were growing, the unhappy poll results due to poor quality would be overshadowed by the happy poll results from new buyers
1.3 Dishonest reviews:
even if buyers are able to assess quality after the fact, it's hard for old buyers to communicate this to potential buyers due to honest reviews being buried in dishonest reviews
2. Peter principal (experts are promoted, so at any one time most of the people doing a job haven't achieved mastery):
This is especially a problem for services, where someone who has achieved mastery at providing the service can make more money by becoming a manager. For example, in business consulting or in construction, often there is a person with a lot of experience at the top but the people actually doing the work don't have a lot of experience.
3. Market won't pay for quality:
Even if buyers can tell that one product/service is of higher quality than another, often they would choose the cheaper one over the high quality one, either because the product/service at issue just isn't that important to them relative to the money they could save, or because they have a hard budgetary constraint and they are forced to compromise somewhere.
4. Cultural factors:
Some people report that, for example, quality in the U.S. seems to tend to be worse than quality in Germany or Japan. If this is real, i don't know what it is specifically about US/German/Japanese culture (or workplace culture/structure) that causes this.
My grandma had old stuff that was built like a tank.. AC that worked for 40 years, bulbs that still work, Fridge that has been working for 50 years, sewing machine also 50+, TV (huge bulky) still working, fans, etc..... I bought a coffee grinder from Amazon few days ago, it broke on the first try, like completely obliterated.
From an energy perspective, a 50 year old fridge is likely not great and probably a new one at some point would have paid for itself (and it's carbon footprint).
There was a story[1] about a guy who kept the same heating boiler for nearly 50 years. Can you imagine how much more he spent on fuel in that thing for 50 years compared to changing to a combi at some point? Condensing boilers are 90+% efficient, who knows what that is.
I would like to see the essay revisited in a more cogent way. There are massive differences between local business disappointments and global business disappointments. The essay is too unfocused to draw actionable conclusions and would benefit greatly from concentrating on more specific issues. I would recommend focusing on one specific point of advice and tailoring everything to support that one point.
> Amazon knew that the courier service they were using didn't really even try to deliver packages4 promptly and the only short-term mitigation available to them was to tell support to tell people that they shouldn't expect that packages have arrived when they've been marked as delivered.
Well, that's partially the fault of Amazon for paying absolute shit on parcel contracts that literally cannot support a parcel business. When the incentive for the parcel delivery service is to mark a package as delivered to get money from Amazon and at the same time hold the packages for one to three days to group them by recipient, of course this is what will result!
> There's a cocktail party version of the efficient markets hypothesis I frequently hear that's basically, "markets enforce efficiency, so it's not possible that a company can have some major inefficiency and survive".
That's not the efficient market hypothesis but a complete misunderstanding. The hypothesis is that the market (not products as you assume) are efficient, leading to the optimization of share prices to reflect reality. [1]
>But, when people are mostly making decisions off of marketing and PR and don't have access to good information, there's no particular reason to think that a product being generally better or even strictly superior will result in that winning and the worse product losing.
Absolutely hit the nail on the head. Advertising distorts the market, meaning what would be a good idea in theory becomes a completely different thing in practice.
> It's considered normal to have unattended property stolen in public spaces and not in private spaces, but that's more of a cultural distinction than a technical distinction.
Surely it's also a legal distinction? In Australia at least, I believe there can be significant jail time given for breaking and entering.
Sparing the contents of the article and the great discussion around it, as an aside: who in this day and age has website text that spans 100% width of the viewport?
On my 32 inch monitor it's completely unreadable. I get the low/no style aesthetic, but the wall of text is quite unreadable to me.
It's like asking why consumers go for low prices rather than good quality. Similar answer: Because it's easy for companies to hide poor quality, but impossible to hide high prices. (Although companies do their best to hide prices anyway, like airlines charging everything as an addon).
In many domains, higher prices fail to deliver better quality. You just pay more.
People learn to optimize for less anger over being cheated: it doesn't work well, but you didn't pay much. If you had paid more, you would (probably) be equally dissatisfied, but also feel like you got suckered.
Consumers have by and large come to believe that price matters more than anything, so the free market economy of goods has naturally optimized for that priority.
Or through the erosion of real wages over the past 50 years, consumers have been more or less forced into buying lower priced items. And was this not the promise of free trade and globalization - cheap stuff?
Cheap stuff that used to be wood or metal and last for 100 years has turned into plastic cheap stuff that you write a nyt column about if it manages to survive the demands of regular expected usage for 1 year.
They'd pay some people like my immigrant grand parents from europe like a few dollars a day and give them no benefits to stamp out ten thousand can openers out of some sheet metal that was probably formed into a roll locally from the steelworks up the river using regionally sourced coal and iron. How can that not be cheap?
I don't have any plastic stuff that breaks. My supposedly bad Ikea furniture has all survived 3 moves, none of my kitchen tools break, car's about to last 10+ years. I think my worst problem with something I own is my PC fans are too loud.
Really? That's surprising. I look at something wrong and the plastic part snaps in half. It's even worse if you have some plastic piece that sees regular usage in direct sunlight. The UV will make it brittle.
It turns out anytime you update the Gigabyte BIOS it loses all its settings including the custom fan curves/undervolting I built from the advice on random reddit posts.
I think there are plenty of good things made out of plastic. Legos never break, for example.
Particle board furniture is annoying. You scuff off the veneer, and the item is essentially ruined. Wood is "self healing" in the sense that scratching it exposes more wood, so it doesn't look as terrible.
Personally, I've found software to be the worst thing about modern devices. A long time ago, I bought a terrible Tiertime 3D printer. It doesn't accept gcode from the computer, so you have to slice models using their proprietary app, and it's absolutely horrible. Requires registration before use, doesn't work at all. I used it once, was disgusted with the software (being unable to slice a model more complicated than a cube), and put the thing in the box with the goal to return it. For various reasons that never happened. Three years later, I decided I was tired of looking at the box and can always use another 3D printer, so I did a "brain transplant". I replaced their proprietary logic board with an open source one (Duet 3 Mini 5+). Now the printer works great. (Before the project was done and I was getting my bearings on the internals, I was appalled at the shortcuts they took. Sheet metal instead of extrusions. Heated bed connected to the power supply with flatflex cable, hotend connected with a ribbon cable, 19V power supply instead of 24V just for a tiny bit of extra savings on inductors. But honestly, the design is fine, it was just their software that made it unusable. I learned a lot about cost reduction by taking that thing apart, and I'm impressed how good of a job they did making it cheap without actually making it work badly. The ABS enclosure is also top notch, some of the best injection molding I've ever seen on a $300 product. No way you could make something that good yourself for the price of the whole machine.)
I've also had some good repair experience on modern, cheap, made-in-China electronics. I have a Siglent oscilloscope, and one day, one of the knobs locked up. I resigned myself to just never using that channel again, but on a 2 channel oscilloscope there aren't really channels to spare. Knowing I wasn't going to find $10,000 laying around for a proper instrument that would have good encoders, I begrudgingly opened it up. Everything was held together with screws, and I had the front I/O board out in a half hour. Desoldered the broken encoder, soldered on a random encoder from my parts bin, put everything back together, and ... perfectly working oscilloscope. It wasn't unrepairable, it was merely uneconomical to repair. If I was writing software instead of repairing the 'scope, I could have just bought a new one. But it was some Saturday night at 3AM when I was too tired to do anything except watch crappy YouTube videos, which nobody will pay me hundreds of dollars to do. So it ended up being quite economical.
I forgot where I was going with this, but basically if you can open something up, today's manufacturing is as good as any manufacturing in the past. Someone that wants to repair or mod, can. (Until you run into glue. Oh how I hate glue.)
"erosion of real wages" isn't actually calculable since the price of your house doesn't really have anything to do with how many electronics you can own, but it's most likely not true - houses are much bigger and higher quality than 1970, cars are incredibly safer, you can buy infinitely more computing power, most importantly you don't have lead poisoning now.
Economists (imperfectly) take this into account when calculating inflation. In cars, for example, economists estimate the price of individual features (e.g. a power steering wheel, ABS, a cup holder) so that they can accurately account for increases in quality.
Those calculations are the ones that make it look like American factory workers were automated out of their jobs, because a 2010 computer is a zillion times faster than a 1970 computer, therefore each factory worker is now producing a zillion times as many computers.
The jobs were mostly shipped over seas. The ones that were left had to compete with the lower prices that were the product of third world cost of living/labor. Management took advantage of the fact that most of their workers don't understand/pay attention to inflation.
Liz Warren and Andrew Yang both used these numbers to run campaigns about specifically automation taking away everyone’s jobs. I do think they know how outsourcing works, but the automation thing turned out to not be real.
The author uses a lot of examples of industries they seem to have only passing knowledge, e.g. geely's involvement in Volvo's daily management. I can tell you, it's non-existant. Same for Tata in JLR.
Been there and done that. There aren't any silver bullets. Just some random advice.
First, pick your battles. You can't fight every one at the same time.
Second, DOCs and Email are your friend. Don't bother with Slack/Teams. If you came from an important meeting, reply to everyone who attended the meeting and CC and BCC people, and write your version of meeting notes and focus on what was agreed upon and what questions remained unresolved. After several emails like, this copy/paste them into a Word document that you save. I have shut down the worst offenders by simply re-sending the same email I sent six months ago.
Third, be prepared for the blowback. I got screamed/yelled at/for days by executives. Eventually the truth will come out. But it might take years.
I think people build custom homes for this reason. Customize it to work the way you want it, and usually use better quality materials and finishes than you get from the typical builder.
New-build housing in the UK is complete shit[1]: flimsy, shaky construction, no acoustic insulation, little storage, tiny windows, every dimension slimmed down to the minimum, inside and out and slapped together with only the most cursory quality checks. I'd never buy a new build unless it was custom (not that I could afford that compared to the same older house).
But, considering the developers aren't going to charge their ways, I'll remain grateful that there are people who do want to buy the "new" houses (they won't stay new for long, they'll be wrecked after a few years of being lived in) at a premium rather than compete for livable housing.
It isn't hard, just more expensive. Things that meet MIL-SPEC work very well (your taxes already paid someone to verify it) and can be easily purchased while sitting on a toilet.
I tend to write stuff that works well. I'm working on the last two screens of a project that has been underway for a couple of years. It's still a ways off from release, as we have lots of fender-polishing to do, but I'm pretty chuffed with how it works.
One of the reasons that it has taken this long, is because I am totally anal about Quality. Constant stopping to fix bugs, or even refactor out bad design decisions.
I've found that doesn't always make me popular.
[EDITED TO ADD] See what I mean? Advocating for high Quality work is not received well. I've learned to keep it totally to my own work, and even that is often perceived as an attack.
I didn't get very far because Silicon Valley is very discriminatory and I can't stomach the denials. There was a long, sordid history of preventing Blacks and other people of color from buying. There were few opportunities to participate in the wealth engine compared to the nominal chances in a place like NYC where the most impoverished tracts in America (Bronx) were just a subway ride to Wall Street.
So yeah, Silicon Valley undercuts itself on talent, quality, ethics, equity and the like and always has.
Because most consumers don't know how tech products work, so they're unable to make the best choice for them. They end up forcing "tech-literate" people to use crappy products, by the force of the market.
That's why marketing beats product quality. Liars can win as long as there are enough gullible people, and as long the government doesn't enact laws to protect the public's interest.
There is no way to improve product quality other by regulations. The "invisible hand of the market" was misinterpreted.
You can win in any competition if the rules don't prevent bad faith players.
Libertarians and other silicon valley start up bros will downvote me.
I think it does mention that. At least it talks about "lemon markets" and how the term "lemon" comes from the car market, and IIRC it links to the Wikipedia article on lemon market theory.
This article is just so rich it could become a book. It reminds me of
Applied Systemantics by Gall.
It's really a list of psychopathologies, misunderstandings, erroneous
thinking and gotchas that lead to poor quality. Among them are
economic fallacies, biases, abusive power relations, cultural malaise,
and FOMO.
In the end I think what the author is saying boils down to Sturgeons
Law, that 99% of everything is rubbish and that it's our fault. We get
the quality we deserve, because we are lazy, entitled and much more
stupid than we think we are. Such is the nature of mass culture and
everything else is an outlier.
I think there's a reason why that's a comedy film instead of a serious take on modern consumerism.
We have plenty of "indestructible" tools today. Cast Iron skillets, hammers, screwdrivers, wrenches, ladders. Most furniture qualifies as well (though MDF-board crap exists, they still have a use. Its more efficient for a college-student to get disposable MDF-based crap for their dorms with a lifespan of ~4 years or less, rather than buy actually quality furniture).
These are called the "durable goods market", and plenty of them can be bought and work for years, decades even.
------------
Everything wears out given enough time. But my roof is expected to last 50 years and has a wind rating far in excess of any 50-year storm in my area (Florida roofs are famously shingled with the cheapest crap, because hurricanes hit them so often it doesn't make sense to invest into the roof... no roof can possibly survive a Cat3, Cat4, or Cat5 storm).
Because my area doesn't have tornadoes or hurricanes (at least, no notable ones in the past century), it makes sense for me to invest into my roof.
I'm amazed at how they keep trying to reinvent hammers. I have two. One good old one with a wooden handle and a steel head, and one with a super-ergonomic rubber/metal handle and a smaller, treated steel head. They both bang nails and both feel alright. If I need to wield one for hours, I can see the rubber handle being better, but I'll wear gloves if it's an issue with the other. Rubber handle will likely last < 20 years.
The hammer thing mainly comes from Home Depot and friends selling to “prosumers” - people who don’t know enough to know why the more expensive hammer might be better for a particular use but buy it anyway.
When doing roofing the experienced guys would have three or four hammers of various types and switch between them as necessary. Me? I couldn’t tell the difference.
Take it from a former pro, get a wood handle. Steel lasts forever, but your arm won't. Fiberglass is a lot better than steel and will last a almost forever, but wood is cheap and the best for your joints.
Maintenance is also something worth taking into consideration. If you get a nice thick wooden benchtop, you can keep sanding it down as needed while the years go by. If you opt for a cheaper particle board top or laminate top, you will likely just be replacing it instead of resurfacing it.
Its not profitable in current iteration of laws. If we had laws that forced manufacturers to create stuff that lasts two decades -> everything would look differently.
I want to give a shout out to the Roomba designers. That is my favorite gadget. I have a very old one. The quality is awesome. The ability to clean & replace parts is easy.
The bit about being unable assess expertise reminds me of Dunning-Kruger, and also The Prince:
> A prince who is not wise can never get good counsel, unless he puts himself completely in the hands of a wise man; but such a man will soon take over his state.
What do you do? Become expert at everything? Form a coalition of experts, who share their assessments? Or... somehow... acquire the ability to assess "expertise" in general, regardless of specific field?
I think Elon Musk's big advantage over his competitors is that he has a high enough level of engineering skill to identify talent. Elon himself isn't the best engineer ever born, but he's good enough to know when someone's ideas or designs or plans make plausible sense. Most CEO's of large corporations, even engineering companies, can't do this, and it shows. So they run the company on the basis of making numbers bigger, and making sure that they cut costs enough to turn a profit each quarter, which eventually runs the company into the ground. They don't pay attention to the ground reality of what the company is actually doing, and only govern on the basis of what the stock price is doing, because the ground level reality is inscrutable and meaningless to them.
one thing you quickly learn from pursuing an academic topic more deeply than surface level is that you often can't trust your perfect-model-based intuition to be correct. I think this is where the Dunning-Kruger effect comes from: people who haven't learned about how their intuitive reasoning is wrong in a specific field, taking a look at it and just saying "oh, it's obvious". For some reason (not sure if it's systemic or just my perception) this seems especially common among STEM-oriented people, and especially directed towards soft sciences and the humanities.
The efficient market hypothesis seems like a perfect example of this over-reliance on assuming paper-thin idealised models actually reflect reality.
This article is amazing. I'll probably read it again twice. Its ambition is sprawling (is it a takedown of capitalism, or a takedown of American culture, no it's both!) and it's somehow successful. I get this is a superfluous comment, but goddamn.
The website has basically no CSS, so it's up to your user agent to decide on its default presentation.
Just so happens that modern user agents don't really care about bare HTML, just CSS and JavaScript, so your default presentation is optimized to 1990s hardware at 800x600.
I guess "reader mode" is slang for "what the user-agent should be doing to make documents reasonable for modern displays", so press that button in your user-agent.
I'm usually just in reading mode, but occasionally I turn it off to see a/the blogs design because I feel it gives just a little bit more context to any post of text (for example, the author has a playful design and it gives me a small idea of their character). Unfortunate it's hardly readable, but it still says something about the author. HN is fun :)
Haha, yeah, but actually it reminded me of old Web pages. Just text, format it however you want (make the product good yourself!), nothing in the way, just the content.
I'm surprised how much I've gotten used to fancy websites, colors and formatting.
They've put quite a bit of effort into it. That much text is a lot of work.
For what it's worth, their Patreon nets over 3000 / month, so quite a few people think they put in just the right amount of effort, or even above and beyond :)
With Chrome Desktop, and a fairly wide browser window (1475px in my case, required by many web applications), the text is small with minimal margins, and practically illegible. If I shrink the browser width considerably (to a minimum width) then the text does look good, like an ebook.
Because china.
Things used to be good 20 or 30 years ago. Then china started in with all that cheap crap.
Same shit going on with India in the software space. Just look at oracle. Tot shit.
Not every customer can assess the quality of a product. I can judge the quality of a violin, but I can't tell you how long the compressor in a refrigerator will last.
This is why the only signal that customers used to have got removed - a refrigerator that had the same design and components for twenty years is likely well made (if it weren’t people would have found out or changed the parts).
But these are rare (one of the few I know of is Speed Queen but I suspect there are others in other industries, mainly commercial/industrial products).
Amusingly enough these work best when innovation is basically dead, so that a new one and a twenty year old one is basically the same.
Except being made well isn't the only criteria. From what I understand, speed queen uses so much water compared to a more modern machine that you could buy a new machine on the cost savings.
There are a multitude of reasons, companies no longer make products to last. Watch videos where they dissect products, you'll find machines are designed to fail after a certain amount of time. Motors that only contain a certain amount of oil and cannot be serviced.
This paragraph from the article reminds me of the recent HN discussion about resilience. In this particular example idling 99% of the time is cheaper, but even if idling 99% of the time is slightly expensive, it might still be a good idea not to outsource due to resilience. What if the vendor goes bankrupt? What if the vendor's vendor is unable to perform their contractual duty due to a disease outbreak? Or a geopolitical risk? You avoid these risks if everything is in-house, so these risks need to be priced in when determining whether to outsource. But more often than not, these risks are priced at exactly zero.