The point isn't to make it easy to register warranty for a product.
The point isn't to make it easy to talk with customer service.
The point isn't to make it easy to communicate with the company.
The point isn't to make it easy to obtain service or replacement parts.
The point is to "technically", and therefore legally, offer those things and minimize the cost of offering those things.
All of these things are working precisely as intended. The company is not optimizing for customer experience or product quality, they're optimizing for profit.
When support needs are infrequent enough on the customer end, this effectively becomes a market for lemons--a customer can't know how good your support is until they've bought your product, and by then it's too late for them. People can advertise world-class customer support for one-time purchases because the few customers that encounter the awful support won't move the needle that much compared to shifting money from long term support teams to sales teams.
That being said, I realize this dynamic is likely much different for frequent/long-term buyers such as B2B solutions where quality support does translate to better retention and word-of-mouth advertising.
> All you have to do is hire customer support who is empowered, knowledgeable and actually cares
Those people you have to pay, because they can do well in other positions. But paying them contradicts the goal of minimizing cost.
I'm not saying there isn't an opportunity to invest in top notch customer service as a product differentiator, but it's probably a safer bet to have barely-existing support and lower the sticker price.
> it's probably a safer bet to have barely-existing support and lower the sticker price.
Depends on the product and the size of the company selling. Sure, in certain circumstances it makes sense to skimp on customer support. In others, it absolutely does not.
For instance, one of the reasons Amazon was so successful when introduced in Europe was its great customer support. Years earlier, IBM had similar success stories.
I've been at two different companies who also had support as a competitive differentiator. Both of them were bought, both had support stripped and outsourced and both are more profitable than they were.
Sadly, as much as we want to believe it, support does not seem to make enough market difference to justify the outlay.
It's not that support doesn't make a market difference, it's that customer loyalty is a liquidatable asset, and almost everyone wants to retire someday, so good companies almost inevitably will eventually become bad companies and cash out all the good will they've accrued over years and decades.
The only possible solution is some sort of regulation of transparency in ownership structured similarly to banking regulations (routinely moving amounts of money just under reporting limits or in any other way seeming to attempt to dodge scrutiny is considered malfeasance), so customers are promptly and continuously informed when Mom and Pop Shop is bought out by Established PE Vultures™ or Scrappy Small Business gets acquired by S&P #342, Enshitiffication Incorporated or a majority of Public Infrastructure Corporation is owned by Slacktivist Investors Group.
I want it to be obvious, and I want ownership to necessarily be as prominent as branding. "A Unilever Brand since 2003" needs to be just as big as "Natural Small Local Foods Coop Inc, since 1913," so that we can start to associate the corrosion of service and product quality with ownership and acquisition and corporate mandates, and any attempts to overly convolute the source of a clear business direction with scattered minority ownership just happening to band together needs to be treated as fraud.
And most of industry today—but especially the tech industry—is laser-focused on making sure they never actually have to compete on a level playing field with anyone.
This absolutely works, and was the key selling point for Rackspace when I worked there, "Fanatical Support". Unfortunately it got purchased by private equity and subsequently ruined, eliminated almost all US staff, and offshored everything to India, and this included a rebrand to remove Fanatiguy from the logo and the "Fanatical Support" promise.
While you /can/ make a premium offering with competent support as a competitive advantage, the most short-term profit to be made is then to buy that company after it's established a brand reputation and burn that brand reputation as fuel to produce extra profit until it becomes worthless and anyone who held your stock previously is left holding the bag.
That's a real shame to hear. Many years ago I took over the running of a website that was hosted on Rackspace, and their support really was great. We rarely had problems, but when things did go wrong they always resolved it really quickly.
that can work out in large scale b2b or expensive consumer goods... but in mass market goods? forget it. people vote with their wallet and only their wallet, as evidenced by the shiploads of (sometimes life-threateningly defective) crap Temu and Alibaba have flooded most Western countries with.
YES and I feel like "Planned obsolescence" needs to be taught to every engineer and web developer and tech people. They will understand why or how some corp. decisions are made and what is the aim.
I agree with you, but is awareness among that cohort really the problem? “Planned obsolescence” is engineered by tech people, after all. It certainly wasn’t divine intervention that came up with internal Li-Ion batteries (IMHO the most brazen example, where manufacturers effectively stick a 3-year timer inside your gadgets).
It wouldn't matter so much if they were standardized. With all the billions of devices sold, there could easily be a standard set of a hundred sizes. Someone petition the EU...
Its amazing to me that people still dont get this.
A business is in the business of making money. Full stop. I dont care what the service or the product is. The goal in a capitalist society is to make profits. Some are more socially conscious than others, but the goal is the same.
I had to buy one a couple of years ago. Snarkily I asked the floor salesman if I could get the washer “without all the smart features”. He said “let me check”, which had me puzzled. He came back to inform me that they still had last year’s model which was before the “smart” features were rolled out. He said they can sell it on the same warranty, & since it was older I would get a significant discount. I cherish that machine for its dumbness.
It was a pain to do, but I have a Sony "Google TV" that I fixed to remove or hide the Sony bloatware and Google adware. I hated how the home screen would display full screen loud ads and the TV would constantly want to update so that it could display an even more obnoxious screen layout with more ads, so I loaded a simple launcher that always displays a static list of hand-picked apps. It required (or successfully dark patterned) me to sign in with a Google account, so I created a dedicated burner account. Google really didn't want to let me create one "please provide a working phone number to verify" but I managed to create one.
Of course I'd prefer a plain dumb TV but there weren't any cheaply and conveniently available at the time. Second best thing is a de-Googled TV. Now if only I could figure out a way to disable the Google buttons on the remote so that kids don't accidentally get into the app store (ads!) or activate the voice control.
Open the remote, put insulator over the contacts. Works on Roku/Samsung - voids warranty but, it looks like you're already past that (or moving in that (correct) direction)
The annoying part is if you disable all the ads and recommendations, they disable voice search. Feels like they are punishing me for opting out of ads...
On powering up, my TV often switches to its default "smart features, prepackaged video streaming service integrations" mode from the HDMI input (the only source I ever want, given my AV receiver manages everything). If it weren't a "smart" tv, I doubt it would keep trying to switch from the configured settings. /anecdote
My smart TV is not on wifi and physically unplugged from the internet except for when I deign to upgrade its firmware. Which I've done once since I bought it 5 years ago. I use 0 (zero) smart features and it is unable to report whatever it may have collected. This seems like a good way to handle them to me.
Some tv's bother you with popups and notifications and all the other crap related to its "smart features".
And yes, even when not connected to the internet, then they show you popups to connect it to the internet, updates may be waiting, new features may be in a new update, you software has been last updated 726 days ago, click here to troubleshoot the internet connection, etc.
To be honest, looking at the PiHole logs the TV itself hasn't tried to do much of anything. The apps on my streaming device (namely Netflix before I uninstalled it) tried, but the PiHole always caught it.
And DoT/DoH don't even add new capabilities, it's been possible to use a VPN to a popular shared hosting provider (e.g. AWS) to hide traffic for a long time before DoT & DoH became standards.
Yes. Put your TV behind a second router, manually assign IP address and route to your local network, and don't give the router an upstream gateway. Then any packets the TV might send even to a plain IP address will be dropped at its router before reaching your main router.
Someday I'm going to download one of HN data dumps and find the earliest mention of this particular scare story. It must be ten or even fifteen years old by now, ever since Amazon started offering 3G equipped Kindles with an always-on connection.
Sneaky cellular access hasn't happened so far, not because the vendors wouldn't like the capability, but (IMO) because it would introduce enough of its own costs and complications to be unprofitable. It's easier to piggyback on customers' internet and disregard the small fraction of privacy-conscious buyers.
This is the actual reason there was all that hysteria around the "race to 5G" with the dire warnings that we'd be buried by China if we didn't roll out 5G ASAP. The actual reason is that 5G works better with congested cells and large numbers of clients than 4G, so it's a lot easier to put cellular modems in every device and bypass those pesky users.
If I have to take defensive maneuvers against an appliance, I don't want the makers of that appliance getting my money.
Sceptre dumb TVs from Walmart's web site. That's the cheat code. If they run out of those, I'll use a large computer monitor. Attach an outboard HTPC or Apple TV and you're set.
I personally like the ad-subsidized TVs. The money I save easily covers a used Apple TV on eBay, and I never connect the TV itself. Regardless of TV brand my interface stays the same. Bring back modularity!
My LG washer and dryer send a push notification to my phone when the load is done, or I can check on the minutes remaining from anywhere. The other stuff I don't care about, but that feature is pretty useful for me. Having an open API would be better, but I know 99% of their customers wouldn't know what to do with that.
$10 smart home vibration sensor works great for this. I have it all fancy in HomeAssistant so I can toggle if I want to be notified or not. You can also cue off the light if it has an led to indicate it's running.
The parent post saying they “cost a bit higher” is, AFAICT, understating it a wee bit. LG’s consumer 65-inch 4K OLED TVs run between $1200 and $2500 at retail, from a quick search; the 65-inch 4K OLED “professional monitor” linked on that page—e.g., the one that’s closest to an actual TV—retails for around $8000 at B&H. The cheapest OLED commercial TV I could find at b&H was a 55” LG for $3K…and as far as I can tell it’s really just a Smart TV, complete with all internet connectivity, for twice the price because it has the word “commercial” in its name.
I got a Spectre TV (pretty cheap, they sell it at Walmart). It's a little more expensive than other TVs in the same class, but has no smart features: you plug HDMI into it and it displays it. You hit "Source" and are greeted with a no-nonsense menu that lets you choose the HDMI port, composite input, etc. It's great.
Couldn't you just print the product number as a barcode/qrcode and let a "dumb code scanner" read it, instead of having to download a multifunctional LLM?
Less than 100 lines of code to give the overwhelming majority of consumers "LanguageModel is not available." and a free snipe hunt for 0.5% of them to try to figure out how to enable it via chrome flags or snapshot browsers.
Yeah maybe it is overkill but this would be a semi hard computer vision task not too long ago, it is pretty amazing you can get it that easily nowadays
So now marketing is working through the disgruntled-at-stupid-tech HN bait (as someone who works on the built-in AI team)? Or wait, how long have they been doing that?
Why even have to register the product in the first place? When I buy a Lenovo laptop from any store, it just has a warranty associated with the serial number, and if I need support or repair, I can just contact Lenovo, and get service under the warranty.
The actual reason they tell you to register for warranty via a phone number is for a salesperson to pick up, and upsell you on "enhanced warranty" or "insurance". It's proven that people feel awkward saying no to a salesperson, and agree to pay extra much more often than they would do online (they'd just tick "no" on the enhanced warranty paid product).
That's also why the author went on a queue: the call-center is not for the washing-machine company, it's an insurance-selling center that works with multiple companies.
Why not just print a $.0001 sticker and stick it the machine listing all of the info needed to fix the machine... instead of building a $86million datacenter and burning through $100,000 of clean electricity every month that could have been used to power homes, but was instead used doing this?
As someone who has a side job doing repairs, I charge more when customers try to repair things themselves.
There is a whole class of people who are smart enough to fix simple things, but not smart enough to recognize their limited ability. They will strip out all the screws on the machine and then claim warranty after replacing random electronic components on the control board. In reality the problem will be a dirty contact that is a 5 minute fix.
On the one hand what you're saying seems somewhat fair, but on the other hand, I prefer encouraging the attempt. A portion of the people who do this will never have to call you, having managed to solve the issue, and any experience gained with the attempt will go into doing it better next time.
In a world where people know less and less about how to solve problems themselves, I think repair skills are incredibly important for people to have.
Personally, I would happily pay more if you show me what you're doing and/or explain what I did wrong :)
Yeah, I had the pump on my washing machine go out. I called the repair guy, he said there was nothing to fix, it just needed a new pump. He told me he could do it for $X (on top of the call-out fee), but it wasn't that hard and I could find the replacement pump online for ~1/3X and do it myself. I did that and it took 20 minutes. We will definitely be using that repair company in the future for anything that isn't immediately obviously trivial
You're only gonna learn you made a dumb mistake when you go poking around $10k equipment with no documentation, no schematic, and no community support.
Speak for yourself. I can make dumb mistakes with $5 equipment and all of the docs, schematics, and the spectral presences of the implementors there to counsel me.
And how did you build the skillset to do this as a side job if it wasn't by making the attempt yourself using available information and learning? Isn't your position just ladder-pulling that creates a population of less informed and less capable people?
As long as, and some suppliers seem to find this onerous, the sticker can be peeled off cleanly.
(I once bought a grass rake with 2 or 3 dozen metal tines, and it arrived with a huge sticker across all the tines. Which when I attempted to peel off, left a scattered layers of paper hard glued to all the tines. Not happy.
Same with a rolling hot dog cooker. Glued “temporary” sales sticker covering half the transparent hinged top. Not happy.)
Fender did this with my bass guitar. Trying to get the sticker off without scratching it was impossible, and even after using alcohol there's still some residue D-:
The sticker will get lost, peel off, get scratched until it's unreadable and then the user has no way to get that information. Maybe if they are really on top of things they will take a picture of the sticker when they first get the machine and save it somewhere that they will remember 5 years later when they need it, but most people will not do that. And if you aren't the first owner of the machine, well then, you're just sol.
I’ve got a tube based oscilloscope with a hand written sticker saying service is due in the ‘40s. The important info is stamped metal (model, serial number), or internal paper stickers (field manual).
It still worked last time I plugged it in. I’m not sure if I’ll be able to find someone to calibrate it when the 40’s roll around again. :-)
Do you really think this cloud service will be up and functioning for 20+ years? How much stuff like that was set up in 2005 and is still working?
But even without that calculus, you can put a bunch of stickers on, all over the machine. They cost nothing and can be applied automatically. Better yet, punch it into some central metal part of the washing machine like the VIN on a car.
Absolute garbage. I have node.js code < 6 months old that won't work anymore because some ding dong removed backwards compatibility or the infrastructure needed to run it.
Not criticising the article or the decision to buy a new washing machine (20 years is a long time), but just noting that the old machine was likely fixable. If spare parts are even still available, that is. Whether it's deemed worth fixing is another matter.
This must be the case for so many discarded appliances these days, especially underengineered ones with common issues.
Also, not using the QR code protocols properly is a pet peeve of mine. I recently scanned one that was just a URL in plaintext (no web link protocol). If I was on an iPhone or using a simpler QR scanner, it would not work at all.
> This must be the case for so many discarded appliances these days, especially underengineered ones with common issues.
While it's true that lots of those old appliances are easily fixable, depending on how old they are it's better to replace due to other factors.
I just recently replaced my 10 years old washing machine instead of fixing it. I was absolutely surprised by the difference. The newer one uses less electricity, less water, washes and dries in half the time, and is absolutely silent.
As someone who just replaced the bearings on a washing machine motor (carbon-commutator), I uhh… agree?
The old machines are absolute workhorse beasts and they can work indefinitely as brand new with some maintenance here and there.
However my expectation of people doing this are basically zero. So this is an anomalous post. By the time you write a blog post complaining about how a machine has a required IoT thing, you could have fixed a handful of issues short of soldering in new relays or triacs on the control board.
Conterpoint: We had a new samsung washer dryer pair with a 90 day warranty; they advertise something like 10 years, but it is a lie. The electrolux in the article is probably similar.
Anyway, it said it lost communication between two boards. I opened it up, checked the wiring harness, and found zero visible problems. I replaced both boards. Same error code. There are 3-4 other computers in that model, so I guess the next step was to replace all of them.
The first two were already a substantial fraction of the price of a new washer, so the entire setup went to the dump (or, hopefully a parts reuse company, but I doubt it). Most technicians refuse to touch Samsung appliances because they are impossible to debug.
Anyway, we replaced the pair with a brand that’s supposedly repairable. Fingers crossed.
I wonder if they ever made front loaders that were affordable, energy efficient and reliable/repairable. There’s no reason these things shouldn’t last more than 20-30 years on average. Maybe there’s a market for such hypothetical old machines.
>There’s no reason these things shouldn’t last more than 20-30 years on average.
This is actually a difficult problem I feel. Misaligned incentives aside. How do you keep a company running if it is so good it captures the whole market in 15 years, but it takes 30 years before it's products need replacing? (This is a simplistic presentation of the issue, but I feel it's understandable enough to start to formulate ideas on how? You could think of lightbulbs that last a century if you like {would lack of competition inhibit progress??}.)
I would love to buy Samsung's washer division, say, and work to make the machines invincible and completely repairable. Then use the profits to bootstrap other such projects. Eventually make the company cooperative, etc, maintain the longevity and work on reducing running costs, improving cleaning, etc.
- support model. You pay 1/10th the manufacturing cost per year. They immediately give you a new one if it fails. Profits are dictated by the difference between the real mean time to failure and ten years. “10” is set by law.
- the price of the machine includes the cost of supplying the above service contract for 30 years, by law. The price of the machine therefore drops as the reliability increases.
- all machines must be 100% recycled by the manufacturer, who also pays for environmental externalities. They pay a prorated multiple equal to the number of years under 10 that a machine is in service before replacement.
- warranties must be 10 years and renewable, and must cover parts, labor, and installation, including things like modifications to cabinets and and legally required code improvements
Not everyone would buy a new fancy machine the same year, so in steady state, they should be able to sell machines, just fewer per capita than today.
Hmmm, even if plans of that nature are not mandated to exist, I wonder if having a clear legal definition of them could help create a market anyway.
> including things like modifications to cabinets
Or perhaps require that the "plan" being purchased includes specs on min/max dimension and max weight.
> the price of the machine includes the cost of supplying the above service contract for 30 years
This seems like a recipe for bankruptcy though. For example, suppose Acme Appliances has a lot of customers, and then that country elects a crook as President who abruptly declared an (illegal?) import-tax (tariff) of one bajillion percent on all the spare parts.
> How do you keep a company running if it is so good it captures the whole market in 15 years, but it takes 30 years before it's products need replacing?
I think that you don't. Things don't have to be infinite money-spinners to deserve doing. Use the googobs of money you made to invest in something else, and let the parts and repair for the old product take over the majority of the old business. Fire up the lines once a year to make new ones, and keep a few engineers working the rest of the year to look for improvements. Or just keep a tiny shop assembling new ones all year, whatever's cost efficient; if you focused on durability and repairability, your best customers probably could put their own unit together from your parts catalog, so you certainly can at any scale.
That might be a good metric for repairability. Could one of your customers build one of your products if they purchased everything from your parts catalog (assuming prequisite knowledge)? If so, you can scale down indefinitely.
>Things don't have to be infinite money-spinners to deserve doing.
The point here is to optimise societal benefit of the production.
I like your ideas about repairability and such. One thing in the back of my mind here is the issue in the UK where we used to lead, somewhat, in nuclear power. We stopped building new power stations and lost the ability as a nation to do so. One would want not just the 30 year old washing machine to be manufacturable, but also for tech improvements to be added in over time. One would also want to extend the manufacture from washing machines to other goods in order to extend the societal benefit. If it can't sustain personnel, then that impacts societal benefit too; as well as impacting ability to attract talent. People being able to insert a replacement control board is different to the company making new control boards that include necessary modifications for changes in the power supply system or improvements to the motors, say.
that's because you bought samsung products. If you don't buy samsung products, you won't have this issue.
Source: boycotting samsung because all of their products are like this except their cellphones, those just explode. Boycotting for 12 years, this year.
Also front loading machines are notorious for having a multitude of issues, from mold, to leaking seals, to bearing issues. There's a reason laundromats don't use them.
In fact, get a speedqueen if you want reliability. I bet they even have HE2 washers, now.
I've read somewhere, that washing machines need to be run on hotter programs (at least 60°C) from time. This dissolves detergent accumulation on the bearing, slowing it's wear down. Apparently bearing failure is more common with modern washing machines as eco 30°C programs are used more often.
By the time detergent and water reaches the bearing, you've already lost. There is a seal that prevents that from happening and when that seal fails, the bearings will be not far behind.
It is a grease-lubricated bearing; 60°C detergent-laden water would be slightly worse than 30°C detergent-laden water, but neither is helpful.
No need to go anonymous. I fondly remember my achievement of taking apart my washing machine to replace the drum bearing. Thanks to the interweb gods for making blogs and videos that helped me.
I'm pretty sure that's the fault of terrible tooling being available to most people. No devices have built-in easy-to-find QR generating abilities, so to create a QR code most people end up searching the Web which is overrun with trashy URL-shortening-and-analytics services, freemium or paid, that wrap your link in their crap and make the URL expire or die with their fly-by-night website. Hackers know that it's possible and free to just make QR codes of the right type, and are able to find proper software to do it, but most people are throwing darts with the assistance of Google so they end up with crap usually.
Don't know about washing machines, let alone this particular one, but my reasoning for replacing a twenty year old machine is that a new one will hopefully use less energy and might be better for the environment thus?!
[Author here] The quote from the repair guy was 270€ for the motor replacement. The new machine was €450. We ended up paying the repair guy €59 for coming, detecting the error, and telling us to let it go. We had the same repair guy exchange the carbon brushes earlier in the year. We really did not want a new machine, but, hey…
Do you know where I can read more about QR code protocols? I was under the impression that a simple URL (with http/s) is common, and I've never had it not on any device.
chrome://flags/#prompt-api-for-gemini-nano to enable this on current Chrome (at least on Linux). It worked after I enabled that flag and restarted Chrome, though it required two tries (apparently it needed to download a model).
I tried to figure out if this stuff was available in other browsers but unfortunately came up short.
The Prompt API [0] works in Edge, though it uses the Phi-4-mini model instead of Gemini Nano. [1]
Currently it doesn't work in Brave (at least on my machine), and I can't find anything online suggesting whether they plan on supporting the Prompt API. You can go to brave://flags/ and it shows "Prompt API for Gemini Nano" and "Prompt API for Gemini Nano with Multimodal Input", but it doesn't seem to actually work.
Article: "Listen to this overcomplicated warranty registration process with some jabs at overcomplicated IoT bullshit. They print the custom unit serial number on the sticker anyways, why not just also customize the QR code to also embed the serial number?
"Anyways, here's an overcomplicated way that uses AI ML to parse that custom unit serial number from a picture.
"Oops, it doesn't work for complicated browser jerry-rigging reasons."
I believe it's the nightly build on windows currently. Actually just looked and came back I'm mistaken it's on mac looks like too. Not sure why it's not working.
For such an operation like an ownership registration for a washing machine, this is totally true. But for some things (e.g., asking details about an invoice) I need to speak to someone, not to fill a form I receive an answer to in 3 days (and possibly not what I need). I say this because some companies are actually removing the option to call them in the first point, or hide it so it's very hard to find the number to call.
I’ve dealt with multiple companies that hang up after hearing what your issue is, before and after AI became common.
Since it is standard industry practice, I think draconian regulations are appropriate.
Off the top of my head, if you are caught having a policy to do this, you pay all customers $100/hour retroactively for the total time they spent on hold (clawing back executive comp if necessary).
An amount equal to the total automatically goes to the whistleblowers that reported the behavior (even if they engaged in it).
> For all that's holy, can you just leverage the Web, please? Don't make me talk to people!
Twenty years ago I would have agreed with this sentiment.
Today... definitely not.
Leveraging the web is an unfortunately-implicit invitation to all kinds of tracking about me, my computer, my stuff. I already get too much spam, no I don't want your f***ing newsletter that I will not read when you send it to me after I already declined receiving it.
> They could still offer to register the machine by telephone as an alternative
They "do". You had the experience of that call. They literally do not want you to actually call. That's a tragedy because even in 2025 there are plenty of people who don't have smartphones, don't have computers, and many of them don't even want one. I certainly don't want to have to install some f***ing app on a smartphone just to register a washing machine. Do you?
Quite ironic for someone at Google to be complaining about a lack of long-term support for customers, when Google seems to have basically no customer support at all.
[Author here] The AI feature is their implementation on the server. I was just trying to see if the same could be implemented client-side as a good use case for the, as you say, experimental, Prompt API. (It's in origin trial, so you can use it already on production sites.) Downloading a model for just this use case is obviously stretching it. You'd use it if it's there, but not download it on purpose just for this convenience.
There is something I can't seem to understand in the relationship between technology and non-tech people / companies.
As a tech-savvy person and someone who relocated recently, I rely a lot on tech solutions for daily problems. When I was choosing my medical insurance company, I concentrated on one that has a good app and English interface. When I need a barber, I search Google Maps and look for barbers who have a website where I can book an appointment. Same for doctors, workshops, or leisure activities. And most of the websites are crap. It's some obscure reservation system, or a generic website with poor design built on Wix or something.
And from one side, I feel like there is an untapped potential here. Any good developer who can build and design a website, can just offer this to their barber or dentist. But on the other hand, I also understand that I'm not the only one who came up with this idea, so there must be a different reason why non-tech people / companies seem to neglect their digital presence. Could it be because they still think that digital presence is not important (and only tech-savvy people like me search for a barber on Google Maps)? Or maybe it's because most small businesses can't / don't want to pay a few thousands for a website and a Google Maps presence (but rather pay $39/mo for Wix), meaning that despite the fact that we are basically surrounded by tech 24/7, the general population still expects tech to be free, or nearly free.
Some thoughts based on talking with small business friends:
- most business comes from local word of mouth, so digital presence is not a priority
- existing solutions are good enough; contact info and maybe a scheduling flow is all the website needs, regardless of UX polish
- changing tech solutions (e.g. website migration) is painful; they have more important things to do with their time and energy
- will investing a few thousand dollars in a website pay off? Probably not, and hard to convince them even if true; they have more important things to do with their tight budget
- I think there's a lot of mediocre devs and agencies offering mediocre work; the talented devs delivering great, user-friendly work are probably not pursuing these projects
[Author here] Hah, our tech writers make me replace this word every single time with "using" in our company blog posts. But this is my blog, my rules ;-)
I was thinking the samething about subway. Ideally everyone would just order a sandwich online and by the time they arrive its ready to go. You scan a QR code(or enter a code) and your sandwhich pops out of hole. Having the option to make or in front of you slows down production and is awkward talking over fans and such.
Inside the heart of every customer support department there are two wolves. One is "do whatever we can to prevent them from phoning us: use every resource, every trick and every stalling measure to try to make sure they don't call". The other is "meh, just direct them to the call centre".
> The other is "meh, just direct them to the call centre".
I worked at a large insurance company and this was definitely the approach. There was a website, but you had to call to realistically get almost anything done.
One product manager's big innovation was to completely remove passwords. Every time you wanted to log in, you had reset the password and be sent a link via email. Of course the didn't announce this, so you would be probably spend 20 minutes frantically looking for your password that didn't exist.
Or, since they're already printing something unique when they print out the product number with the qr code, just print a qr code that goes to a website with the product number already encoded. No AI necessary.
edit: I see a number of people have already suggested this. Clearly a very obvious answer.
I'd say this goes for a lot of developers too. The amount of push back I see online when Chrome implements something that was previously only available in app land is weirdly high. Like do you seriously want to write separate ios app for everything?
I have a better proposal. Let's scrap the web. It's corrupted and broken, it hasn't been fun since MySpace. We serve native apps like VNC over kitty graphics protocol + mosh/ssh + ios/android/windows/macos mosh/ssh client. Even if we need to distribute a customized terminal to make it work. Get rid of the browser and the app stores at the same time.
Would this guys experience meet your criteria? What are the consequences for failing?
Is it 10 years of normal use or still 10 years if you basically run a laundrette. What if you never used it at all. What if (and where I live this is a serious use for a washing machine) you only used the washing machine for making a special rum based cocktail?
This stuff really annoys me but the problem is really well solved elsewhere.
When you buy an Apple product it has a number on it Axxxx and a serial number somewhere. That's all you need to identify your product to anyone. That includes service manuals and spare parts.
And as far as warranty registration goes, they register it at the point of sale/activation as the warranty starting. Job done. No humans / lookups / anything required. It just happens.
I mean, the thing is, the warranty starts automatically, there's no need to register or activate anything. As per EU consumer protection laws (the author lives in Spain).
As for finding service manuals and user manuals - maybe what they need isn't the web but FTP. I mean, if it were still supported by browsers. I remember when some vendors just used to have folders with PDFs you could browse.
The problem with "the web" is that this is no longer a website, but a content management system, or worse, a "customer engagement platform" that is hostile to creating a folder full of PDFs that have stable links. They probably still have that FTP site in a webified form somewhere for service partners, just not for Joe Public.
Author chose a brand based on loyalty instead of doing research. Leverage the web indeed - Google Search would have quickly revealed Electrolux is mediocre.
Plus, the manufacturer can't run a simple website to look up serial numbers, so author prefers their AI-based solution? And then builds one himself? This post reads like satire.
I hate the word "leverage" as a substitute for "use".
I remember this being absolutely endemic in the Microsoft world in the 00's but the first time I heard it was in the 1980s when the current commander-in-thief was a young upstart punk who leveraged the word "leverage" every chance he could get. For him the real use of leverage in business (e.g. "debt") rapidly turned a moderate-sized real estate empire into a hole in the ground and he recovered by becoming a television performer.
If I had to guess, they don't want to leverage the web because calling the phone number to get the 10 year warranty is a dark pattern they hope most people give up on. And they know the current version of the Electrolux washer is hot garbage that won't make it 10 years, compared to the previous one that made it 20 years. Now this is not great long term for the brand, but Yannick Fierling, the new CEO, just started in January 2025 after stepping down from being CEO of Haier Europe in March 2024, a job he kept for 9 years. The previous CEO, Jonas Samuelson, started on February 1, 2016, so why would Yannick care? Yannick will probably be out with a nice golden parachute and have a shorter life than the washer. Yannick is 54 years old and I guarantee the only thing he thinks about every day is retiring soon with as much money as possible.
This enshittification cycle runs every day in the world and I don't know how we can stop it.
They may also try to use the warranty registration as an opportunity to sell insurance products, such as "protection" for other white goods. I had this with a new boiler (American English: "furnace"). They took the registration but wanted to sell me a further warranty on related heating system components not directly part of the boiler. Interestingly, it was a different company that specialises in such products. I think they probably have a business arrangement with the manufacturer to handle the warranty registration admin and possibly the warranty servicing too, presumably at a discounted price compared to the raw cost of merely servicing the warranty itself, in return for the opportunity for upselling their own products.
My guess would be more that they don't have a support org in house. Were likely running very lean on a manufacturing pipeline and are adding a lot of other services on through partnerships.
But they don't have in-house capabilities precisely because of the above-mentioned enshittification cycle: CEO-monkey does not care about long-term. CEO-monkey lowers costs today. CEO-monkey moves on to new job soon.
Electrolux has existed for almost 100 years, and on its way it acquired a lot of other companies. The chances that none of those companies ever developed any sort of in-house assistance, over that time, are close to zero.
That is precisely a path that would lead to you not having a coherent support network, though? My assertion is that they did mergers to consolidate manufacturing. Any support orgs that they acquired along the way were not the aim of the acquisition. Quite the contrary, they complicated things in ways that are not as easy to work with as getting new manufacturing expertise.
The point isn't to make it easy to register warranty for a product. The point isn't to make it easy to talk with customer service. The point isn't to make it easy to communicate with the company. The point isn't to make it easy to obtain service or replacement parts.
The point is to "technically", and therefore legally, offer those things and minimize the cost of offering those things.
All of these things are working precisely as intended. The company is not optimizing for customer experience or product quality, they're optimizing for profit.
100%
Customer support representatives, with very few exceptions, are just going to tell you whatever they think will get you off the phone.
In I.T., this is generally accomplished by blaming whatever adjacent equipment/services they can plausibly pin the issue on.
Huge competitive opportunity here.
All you have to do is hire customer support who is empowered, knowledgeable and actually cares.
Source: support is a competitive differentiator for $CURJOB.
When support needs are infrequent enough on the customer end, this effectively becomes a market for lemons--a customer can't know how good your support is until they've bought your product, and by then it's too late for them. People can advertise world-class customer support for one-time purchases because the few customers that encounter the awful support won't move the needle that much compared to shifting money from long term support teams to sales teams.
That being said, I realize this dynamic is likely much different for frequent/long-term buyers such as B2B solutions where quality support does translate to better retention and word-of-mouth advertising.
> All you have to do is hire customer support who is empowered, knowledgeable and actually cares
Those people you have to pay, because they can do well in other positions. But paying them contradicts the goal of minimizing cost.
I'm not saying there isn't an opportunity to invest in top notch customer service as a product differentiator, but it's probably a safer bet to have barely-existing support and lower the sticker price.
> it's probably a safer bet to have barely-existing support and lower the sticker price.
Depends on the product and the size of the company selling. Sure, in certain circumstances it makes sense to skimp on customer support. In others, it absolutely does not.
For instance, one of the reasons Amazon was so successful when introduced in Europe was its great customer support. Years earlier, IBM had similar success stories.
I've been at two different companies who also had support as a competitive differentiator. Both of them were bought, both had support stripped and outsourced and both are more profitable than they were.
Sadly, as much as we want to believe it, support does not seem to make enough market difference to justify the outlay.
It's not that support doesn't make a market difference, it's that customer loyalty is a liquidatable asset, and almost everyone wants to retire someday, so good companies almost inevitably will eventually become bad companies and cash out all the good will they've accrued over years and decades.
The only possible solution is some sort of regulation of transparency in ownership structured similarly to banking regulations (routinely moving amounts of money just under reporting limits or in any other way seeming to attempt to dodge scrutiny is considered malfeasance), so customers are promptly and continuously informed when Mom and Pop Shop is bought out by Established PE Vultures™ or Scrappy Small Business gets acquired by S&P #342, Enshitiffication Incorporated or a majority of Public Infrastructure Corporation is owned by Slacktivist Investors Group.
I want it to be obvious, and I want ownership to necessarily be as prominent as branding. "A Unilever Brand since 2003" needs to be just as big as "Natural Small Local Foods Coop Inc, since 1913," so that we can start to associate the corrosion of service and product quality with ownership and acquisition and corporate mandates, and any attempts to overly convolute the source of a clear business direction with scattered minority ownership just happening to band together needs to be treated as fraud.
This only works when there's actual competition.
And most of industry today—but especially the tech industry—is laser-focused on making sure they never actually have to compete on a level playing field with anyone.
This absolutely works, and was the key selling point for Rackspace when I worked there, "Fanatical Support". Unfortunately it got purchased by private equity and subsequently ruined, eliminated almost all US staff, and offshored everything to India, and this included a rebrand to remove Fanatiguy from the logo and the "Fanatical Support" promise.
While you /can/ make a premium offering with competent support as a competitive advantage, the most short-term profit to be made is then to buy that company after it's established a brand reputation and burn that brand reputation as fuel to produce extra profit until it becomes worthless and anyone who held your stock previously is left holding the bag.
That's a real shame to hear. Many years ago I took over the running of a website that was hosted on Rackspace, and their support really was great. We rarely had problems, but when things did go wrong they always resolved it really quickly.
Private equity asset stripping is a cancer.
that can work out in large scale b2b or expensive consumer goods... but in mass market goods? forget it. people vote with their wallet and only their wallet, as evidenced by the shiploads of (sometimes life-threateningly defective) crap Temu and Alibaba have flooded most Western countries with.
Zappos got a huge exit by having exceptional customer service. Enough people notice.
Yeah. You can win both ways, by racing to the top with service or racing to the bottom with price. I know which strategy I'd rather be part of.
Honestly, I'd rather have a product that works so I don't have to call support when it breaks ...
YES and I feel like "Planned obsolescence" needs to be taught to every engineer and web developer and tech people. They will understand why or how some corp. decisions are made and what is the aim.
I agree with you, but is awareness among that cohort really the problem? “Planned obsolescence” is engineered by tech people, after all. It certainly wasn’t divine intervention that came up with internal Li-Ion batteries (IMHO the most brazen example, where manufacturers effectively stick a 3-year timer inside your gadgets).
It wouldn't matter so much if they were standardized. With all the billions of devices sold, there could easily be a standard set of a hundred sizes. Someone petition the EU...
Its amazing to me that people still dont get this.
A business is in the business of making money. Full stop. I dont care what the service or the product is. The goal in a capitalist society is to make profits. Some are more socially conscious than others, but the goal is the same.
All decisions derive from this goal.
> Electrolux washing machine
I had to buy one a couple of years ago. Snarkily I asked the floor salesman if I could get the washer “without all the smart features”. He said “let me check”, which had me puzzled. He came back to inform me that they still had last year’s model which was before the “smart” features were rolled out. He said they can sell it on the same warranty, & since it was older I would get a significant discount. I cherish that machine for its dumbness.
…No such luck for TVs.
It was a pain to do, but I have a Sony "Google TV" that I fixed to remove or hide the Sony bloatware and Google adware. I hated how the home screen would display full screen loud ads and the TV would constantly want to update so that it could display an even more obnoxious screen layout with more ads, so I loaded a simple launcher that always displays a static list of hand-picked apps. It required (or successfully dark patterned) me to sign in with a Google account, so I created a dedicated burner account. Google really didn't want to let me create one "please provide a working phone number to verify" but I managed to create one.
Of course I'd prefer a plain dumb TV but there weren't any cheaply and conveniently available at the time. Second best thing is a de-Googled TV. Now if only I could figure out a way to disable the Google buttons on the remote so that kids don't accidentally get into the app store (ads!) or activate the voice control.
Open the remote, put insulator over the contacts. Works on Roku/Samsung - voids warranty but, it looks like you're already past that (or moving in that (correct) direction)
A bit of kapton tape under the button works and is reversible if you change your mind. Also works well for childrens toys that produce music.
The annoying part is if you disable all the ads and recommendations, they disable voice search. Feels like they are punishing me for opting out of ads...
super glue the buttons in place
I don't get why people care about about whether a TV is "Smart". I have a "Smart" TV and never use the features, and it doesn't bother me.
I have a PiHole in case it tries to do anything funny, and that's good enough for me.
On powering up, my TV often switches to its default "smart features, prepackaged video streaming service integrations" mode from the HDMI input (the only source I ever want, given my AV receiver manages everything). If it weren't a "smart" tv, I doubt it would keep trying to switch from the configured settings. /anecdote
My smart TV is not on wifi and physically unplugged from the internet except for when I deign to upgrade its firmware. Which I've done once since I bought it 5 years ago. I use 0 (zero) smart features and it is unable to report whatever it may have collected. This seems like a good way to handle them to me.
5G is in large part designed to fix this. Eventually these devices will only ask for wifi as a formality.
You don't use the features, but those features use you. Gobble up all your data to sell to the highest bidder.
Some tv's bother you with popups and notifications and all the other crap related to its "smart features".
And yes, even when not connected to the internet, then they show you popups to connect it to the internet, updates may be waiting, new features may be in a new update, you software has been last updated 726 days ago, click here to troubleshoot the internet connection, etc.
mine used to strobe the light underneath when it lost internet connectivity. used to.
What counts as funny? Have you connected it to the internet? If so it's probably still spying on you and that is why people care.
To be honest, looking at the PiHole logs the TV itself hasn't tried to do much of anything. The apps on my streaming device (namely Netflix before I uninstalled it) tried, but the PiHole always caught it.
At this point I wouldn’t rely on PiHole DNS logs alone. You’d need to check network traffic from the device in general, it could use DoT/DoH.
And I suppose there’s the even-creeper Amazon Sidewalk: https://www.amazon.com/gp/help/customer/display.html?nodeId=...
I hate sounding like I’m wearing a foil hat but there are a lot of easy ways to get around trying to neuter smart devices now.
And DoT/DoH don't even add new capabilities, it's been possible to use a VPN to a popular shared hosting provider (e.g. AWS) to hide traffic for a long time before DoT & DoH became standards.
Don’t forget about DNS over HTTPS which bypasses PiHole
Is there a way to set up the modem, router, pihole, or any other part of the network to close that loophole?
Yes. Put your TV behind a second router, manually assign IP address and route to your local network, and don't give the router an upstream gateway. Then any packets the TV might send even to a plain IP address will be dropped at its router before reaching your main router.
Wait until TVs come with their own cellular modems so they can give the corporate middle finger to people like us.
Someday I'm going to download one of HN data dumps and find the earliest mention of this particular scare story. It must be ten or even fifteen years old by now, ever since Amazon started offering 3G equipped Kindles with an always-on connection.
Sneaky cellular access hasn't happened so far, not because the vendors wouldn't like the capability, but (IMO) because it would introduce enough of its own costs and complications to be unprofitable. It's easier to piggyback on customers' internet and disregard the small fraction of privacy-conscious buyers.
This is the actual reason there was all that hysteria around the "race to 5G" with the dire warnings that we'd be buried by China if we didn't roll out 5G ASAP. The actual reason is that 5G works better with congested cells and large numbers of clients than 4G, so it's a lot easier to put cellular modems in every device and bypass those pesky users.
Except many TV models are already sold at a loss. Throwing a $20/month cellular modem in there isn't going to help any.
The number of users who opt out of connecting their smart TV to Wi-Fi is so damn tiny that it is a rounding error for TV makers. They just don't care.
If I have to take defensive maneuvers against an appliance, I don't want the makers of that appliance getting my money.
Sceptre dumb TVs from Walmart's web site. That's the cheat code. If they run out of those, I'll use a large computer monitor. Attach an outboard HTPC or Apple TV and you're set.
I personally like the ad-subsidized TVs. The money I save easily covers a used Apple TV on eBay, and I never connect the TV itself. Regardless of TV brand my interface stays the same. Bring back modularity!
My LG washer and dryer send a push notification to my phone when the load is done, or I can check on the minutes remaining from anywhere. The other stuff I don't care about, but that feature is pretty useful for me. Having an open API would be better, but I know 99% of their customers wouldn't know what to do with that.
$10 smart home vibration sensor works great for this. I have it all fancy in HomeAssistant so I can toggle if I want to be notified or not. You can also cue off the light if it has an led to indicate it's running.
I bought a projector mostly to avoid the smart stuff. Large monitors are also an option to some extent.
Try the Spectre TVs. They are normal oled TVS, not smart. Been using one for a few years now, am very glad they exist.
They are called "Commercial TVs", selection is more limited and the cost a bit higher, but no smart-TV BS.
Do they come in OLED now?
Yup, at least from LG: https://www.lg.com/global/business/oled-signage
The parent post saying they “cost a bit higher” is, AFAICT, understating it a wee bit. LG’s consumer 65-inch 4K OLED TVs run between $1200 and $2500 at retail, from a quick search; the 65-inch 4K OLED “professional monitor” linked on that page—e.g., the one that’s closest to an actual TV—retails for around $8000 at B&H. The cheapest OLED commercial TV I could find at b&H was a 55” LG for $3K…and as far as I can tell it’s really just a Smart TV, complete with all internet connectivity, for twice the price because it has the word “commercial” in its name.
The other specs are often different. Things like 100% duty cycle, uv resistant coatings, ipx ratings, serial ports for control etc etc
I got a Spectre TV (pretty cheap, they sell it at Walmart). It's a little more expensive than other TVs in the same class, but has no smart features: you plug HDMI into it and it displays it. You hit "Source" and are greeted with a no-nonsense menu that lets you choose the HDMI port, composite input, etc. It's great.
I've got one too! I hope it lasts.
I fail to see a useful usage for AI in this case.
Couldn't you just print the product number as a barcode/qrcode and let a "dumb code scanner" read it, instead of having to download a multifunctional LLM?
The author works in Google developer relations, and while devrel aren't quite marketing they will use the latest and greatest Google hammer.
His demo is pretty slick, though. Less than 100 lines of code to get “the box I want on literally every customer service site.”
Less than 100 lines of code to give the overwhelming majority of consumers "LanguageModel is not available." and a free snipe hunt for 0.5% of them to try to figure out how to enable it via chrome flags or snapshot browsers.
Provided one is using ChromeOS "standards".
Yeah maybe it is overkill but this would be a semi hard computer vision task not too long ago, it is pretty amazing you can get it that easily nowadays
So now marketing is working through the disgruntled-at-stupid-tech HN bait (as someone who works on the built-in AI team)? Or wait, how long have they been doing that?
They could, that very qr code is on there and contains the product number and serial number inside a URL query that links to qr.electrolux.com.
The page loads but doesn't offer to register the product, which is probably for the best.
I was thinking just have it included in the QR code, so the URL the QR points to is prod-reg.example.com/pid/91234567 or whatever.
Why even have to register the product in the first place? When I buy a Lenovo laptop from any store, it just has a warranty associated with the serial number, and if I need support or repair, I can just contact Lenovo, and get service under the warranty.
Yep.
When the only tool you know is the hammer you're going to hit inappropriate things with it.
The author's own AI demo falls on its face with "Identify product number LanguageModel is not available" too.
The actual reason they tell you to register for warranty via a phone number is for a salesperson to pick up, and upsell you on "enhanced warranty" or "insurance". It's proven that people feel awkward saying no to a salesperson, and agree to pay extra much more often than they would do online (they'd just tick "no" on the enhanced warranty paid product).
That's also why the author went on a queue: the call-center is not for the washing-machine company, it's an insurance-selling center that works with multiple companies.
Why not just print a $.0001 sticker and stick it the machine listing all of the info needed to fix the machine... instead of building a $86million datacenter and burning through $100,000 of clean electricity every month that could have been used to power homes, but was instead used doing this?
As someone who has a side job doing repairs, I charge more when customers try to repair things themselves.
There is a whole class of people who are smart enough to fix simple things, but not smart enough to recognize their limited ability. They will strip out all the screws on the machine and then claim warranty after replacing random electronic components on the control board. In reality the problem will be a dirty contact that is a 5 minute fix.
On the one hand what you're saying seems somewhat fair, but on the other hand, I prefer encouraging the attempt. A portion of the people who do this will never have to call you, having managed to solve the issue, and any experience gained with the attempt will go into doing it better next time.
In a world where people know less and less about how to solve problems themselves, I think repair skills are incredibly important for people to have.
Personally, I would happily pay more if you show me what you're doing and/or explain what I did wrong :)
Yeah, I had the pump on my washing machine go out. I called the repair guy, he said there was nothing to fix, it just needed a new pump. He told me he could do it for $X (on top of the call-out fee), but it wasn't that hard and I could find the replacement pump online for ~1/3X and do it myself. I did that and it took 20 minutes. We will definitely be using that repair company in the future for anything that isn't immediately obviously trivial
> but not smart enough to recognize their limited ability.
Exploring the boundaries of one's own ability is not being "not smart enough." It's learning.
You're only gonna learn you made a dumb mistake when you go poking around $10k equipment with no documentation, no schematic, and no community support.
Speak for yourself. I can make dumb mistakes with $5 equipment and all of the docs, schematics, and the spectral presences of the implementors there to counsel me.
You have an obvious financial incentive to prevent people from trying to fix their own things...
> As someone who has a side job doing repairs
And how did you build the skillset to do this as a side job if it wasn't by making the attempt yourself using available information and learning? Isn't your position just ladder-pulling that creates a population of less informed and less capable people?
As long as, and some suppliers seem to find this onerous, the sticker can be peeled off cleanly.
(I once bought a grass rake with 2 or 3 dozen metal tines, and it arrived with a huge sticker across all the tines. Which when I attempted to peel off, left a scattered layers of paper hard glued to all the tines. Not happy.
Same with a rolling hot dog cooker. Glued “temporary” sales sticker covering half the transparent hinged top. Not happy.)
Fender did this with my bass guitar. Trying to get the sticker off without scratching it was impossible, and even after using alcohol there's still some residue D-:
Goo Gone is your friend.
My usual flow chart: Nothing, water, alcohol, goo gone.
I’ve never ended up with a marred surface.
Tried that on the transparent cooker cover and it took off just enough of the cover surface with the sticker to look horrible.
Otherwise, yes!
+hair dryer heat, between nothing and water, can make a difference in a difficult peel.
For the same reason the rep hung up on him after he sat on hold.
The sticker will get lost, peel off, get scratched until it's unreadable and then the user has no way to get that information. Maybe if they are really on top of things they will take a picture of the sticker when they first get the machine and save it somewhere that they will remember 5 years later when they need it, but most people will not do that. And if you aren't the first owner of the machine, well then, you're just sol.
Rubbish. The sticker is still on my 14 year old washing machine.
Hell I've got a 55 year old piece of electronic equipment here with the serial number sticker still on it.
I've got a 65 year old FM radio with manufacturer-supplied circuit diagram still inside the case.
I’ve got a tube based oscilloscope with a hand written sticker saying service is due in the ‘40s. The important info is stamped metal (model, serial number), or internal paper stickers (field manual).
It still worked last time I plugged it in. I’m not sure if I’ll be able to find someone to calibrate it when the 40’s roll around again. :-)
Do you really think this cloud service will be up and functioning for 20+ years? How much stuff like that was set up in 2005 and is still working?
But even without that calculus, you can put a bunch of stickers on, all over the machine. They cost nothing and can be applied automatically. Better yet, punch it into some central metal part of the washing machine like the VIN on a car.
Absolute garbage. I have node.js code < 6 months old that won't work anymore because some ding dong removed backwards compatibility or the infrastructure needed to run it.
instead of sticker, they can just code something like dog tag, that is thing for century or more
Not criticising the article or the decision to buy a new washing machine (20 years is a long time), but just noting that the old machine was likely fixable. If spare parts are even still available, that is. Whether it's deemed worth fixing is another matter.
This must be the case for so many discarded appliances these days, especially underengineered ones with common issues.
Also, not using the QR code protocols properly is a pet peeve of mine. I recently scanned one that was just a URL in plaintext (no web link protocol). If I was on an iPhone or using a simpler QR scanner, it would not work at all.
> This must be the case for so many discarded appliances these days, especially underengineered ones with common issues.
While it's true that lots of those old appliances are easily fixable, depending on how old they are it's better to replace due to other factors.
I just recently replaced my 10 years old washing machine instead of fixing it. I was absolutely surprised by the difference. The newer one uses less electricity, less water, washes and dries in half the time, and is absolutely silent.
I only hope it lasts as long as the old one.
As someone who just replaced the bearings on a washing machine motor (carbon-commutator), I uhh… agree?
The old machines are absolute workhorse beasts and they can work indefinitely as brand new with some maintenance here and there.
However my expectation of people doing this are basically zero. So this is an anomalous post. By the time you write a blog post complaining about how a machine has a required IoT thing, you could have fixed a handful of issues short of soldering in new relays or triacs on the control board.
Conterpoint: We had a new samsung washer dryer pair with a 90 day warranty; they advertise something like 10 years, but it is a lie. The electrolux in the article is probably similar.
Anyway, it said it lost communication between two boards. I opened it up, checked the wiring harness, and found zero visible problems. I replaced both boards. Same error code. There are 3-4 other computers in that model, so I guess the next step was to replace all of them.
The first two were already a substantial fraction of the price of a new washer, so the entire setup went to the dump (or, hopefully a parts reuse company, but I doubt it). Most technicians refuse to touch Samsung appliances because they are impossible to debug.
Anyway, we replaced the pair with a brand that’s supposedly repairable. Fingers crossed.
I wonder if they ever made front loaders that were affordable, energy efficient and reliable/repairable. There’s no reason these things shouldn’t last more than 20-30 years on average. Maybe there’s a market for such hypothetical old machines.
>There’s no reason these things shouldn’t last more than 20-30 years on average.
This is actually a difficult problem I feel. Misaligned incentives aside. How do you keep a company running if it is so good it captures the whole market in 15 years, but it takes 30 years before it's products need replacing? (This is a simplistic presentation of the issue, but I feel it's understandable enough to start to formulate ideas on how? You could think of lightbulbs that last a century if you like {would lack of competition inhibit progress??}.)
I would love to buy Samsung's washer division, say, and work to make the machines invincible and completely repairable. Then use the profits to bootstrap other such projects. Eventually make the company cooperative, etc, maintain the longevity and work on reducing running costs, improving cleaning, etc.
I can think of a few solutions:
- support model. You pay 1/10th the manufacturing cost per year. They immediately give you a new one if it fails. Profits are dictated by the difference between the real mean time to failure and ten years. “10” is set by law.
- the price of the machine includes the cost of supplying the above service contract for 30 years, by law. The price of the machine therefore drops as the reliability increases.
- all machines must be 100% recycled by the manufacturer, who also pays for environmental externalities. They pay a prorated multiple equal to the number of years under 10 that a machine is in service before replacement.
- warranties must be 10 years and renewable, and must cover parts, labor, and installation, including things like modifications to cabinets and and legally required code improvements
Not everyone would buy a new fancy machine the same year, so in steady state, they should be able to sell machines, just fewer per capita than today.
Hmmm, even if plans of that nature are not mandated to exist, I wonder if having a clear legal definition of them could help create a market anyway.
> including things like modifications to cabinets
Or perhaps require that the "plan" being purchased includes specs on min/max dimension and max weight.
> the price of the machine includes the cost of supplying the above service contract for 30 years
This seems like a recipe for bankruptcy though. For example, suppose Acme Appliances has a lot of customers, and then that country elects a crook as President who abruptly declared an (illegal?) import-tax (tariff) of one bajillion percent on all the spare parts.
> How do you keep a company running if it is so good it captures the whole market in 15 years, but it takes 30 years before it's products need replacing?
I think that you don't. Things don't have to be infinite money-spinners to deserve doing. Use the googobs of money you made to invest in something else, and let the parts and repair for the old product take over the majority of the old business. Fire up the lines once a year to make new ones, and keep a few engineers working the rest of the year to look for improvements. Or just keep a tiny shop assembling new ones all year, whatever's cost efficient; if you focused on durability and repairability, your best customers probably could put their own unit together from your parts catalog, so you certainly can at any scale.
That might be a good metric for repairability. Could one of your customers build one of your products if they purchased everything from your parts catalog (assuming prequisite knowledge)? If so, you can scale down indefinitely.
>Things don't have to be infinite money-spinners to deserve doing.
The point here is to optimise societal benefit of the production.
I like your ideas about repairability and such. One thing in the back of my mind here is the issue in the UK where we used to lead, somewhat, in nuclear power. We stopped building new power stations and lost the ability as a nation to do so. One would want not just the 30 year old washing machine to be manufacturable, but also for tech improvements to be added in over time. One would also want to extend the manufacture from washing machines to other goods in order to extend the societal benefit. If it can't sustain personnel, then that impacts societal benefit too; as well as impacting ability to attract talent. People being able to insert a replacement control board is different to the company making new control boards that include necessary modifications for changes in the power supply system or improvements to the motors, say.
I appreciate your response, thanks.
that's because you bought samsung products. If you don't buy samsung products, you won't have this issue.
Source: boycotting samsung because all of their products are like this except their cellphones, those just explode. Boycotting for 12 years, this year.
Also front loading machines are notorious for having a multitude of issues, from mold, to leaking seals, to bearing issues. There's a reason laundromats don't use them.
In fact, get a speedqueen if you want reliability. I bet they even have HE2 washers, now.
I've read somewhere, that washing machines need to be run on hotter programs (at least 60°C) from time. This dissolves detergent accumulation on the bearing, slowing it's wear down. Apparently bearing failure is more common with modern washing machines as eco 30°C programs are used more often.
By the time detergent and water reaches the bearing, you've already lost. There is a seal that prevents that from happening and when that seal fails, the bearings will be not far behind.
It is a grease-lubricated bearing; 60°C detergent-laden water would be slightly worse than 30°C detergent-laden water, but neither is helpful.
No need to go anonymous. I fondly remember my achievement of taking apart my washing machine to replace the drum bearing. Thanks to the interweb gods for making blogs and videos that helped me.
> QR properly
I'm pretty sure that's the fault of terrible tooling being available to most people. No devices have built-in easy-to-find QR generating abilities, so to create a QR code most people end up searching the Web which is overrun with trashy URL-shortening-and-analytics services, freemium or paid, that wrap your link in their crap and make the URL expire or die with their fly-by-night website. Hackers know that it's possible and free to just make QR codes of the right type, and are able to find proper software to do it, but most people are throwing darts with the assistance of Google so they end up with crap usually.
Don't know about washing machines, let alone this particular one, but my reasoning for replacing a twenty year old machine is that a new one will hopefully use less energy and might be better for the environment thus?!
[Author here] The quote from the repair guy was 270€ for the motor replacement. The new machine was €450. We ended up paying the repair guy €59 for coming, detecting the error, and telling us to let it go. We had the same repair guy exchange the carbon brushes earlier in the year. We really did not want a new machine, but, hey…
Do you know where I can read more about QR code protocols? I was under the impression that a simple URL (with http/s) is common, and I've never had it not on any device.
Not OP, but I found this article recently: https://cubiclenate.com/2024/03/22/generate-qr-codes-in-the-...
Includes patterns for wifi credentials, calendar events, and a few more.
There's no special URL there either, though. The parent specifically mentioned a protocol with semantics for "web links", hence my question.
Same, am I making qr codes wrong? Like you, I've never seen a simple https link in a qr code fail
I tried hitting the "identify" button but got the error "LanguageModel is not supported." This is on latest Safari, iPadOS.
[Author here] Safari is not supported. See https://developer.chrome.com/docs/ai/prompt-api#hardware-req... for the requirements.
chrome://flags/#prompt-api-for-gemini-nano to enable this on current Chrome (at least on Linux). It worked after I enabled that flag and restarted Chrome, though it required two tries (apparently it needed to download a model).
I tried to figure out if this stuff was available in other browsers but unfortunately came up short.
Googler, opinions my own.
The Prompt API [0] works in Edge, though it uses the Phi-4-mini model instead of Gemini Nano. [1]
Currently it doesn't work in Brave (at least on my machine), and I can't find anything online suggesting whether they plan on supporting the Prompt API. You can go to brave://flags/ and it shows "Prompt API for Gemini Nano" and "Prompt API for Gemini Nano with Multimodal Input", but it doesn't seem to actually work.
0. https://chromestatus.com/feature/5134603979063296
1. https://learn.microsoft.com/en-us/microsoft-edge/web-platfor...
Yeah, that's the irony here.
Article: "Listen to this overcomplicated warranty registration process with some jabs at overcomplicated IoT bullshit. They print the custom unit serial number on the sticker anyways, why not just also customize the QR code to also embed the serial number?
"Anyways, here's an overcomplicated way that uses AI ML to parse that custom unit serial number from a picture.
"Oops, it doesn't work for complicated browser jerry-rigging reasons."
I believe it's the nightly build on windows currently. Actually just looked and came back I'm mistaken it's on mac looks like too. Not sure why it's not working.
PromptAPI seems to be the usual fire-and-motion Chrome move.
It doesn't work for me in Chrome either. Oh irony.
Same on Vivaldi, macOS.
Same on Fennec
For such an operation like an ownership registration for a washing machine, this is totally true. But for some things (e.g., asking details about an invoice) I need to speak to someone, not to fill a form I receive an answer to in 3 days (and possibly not what I need). I say this because some companies are actually removing the option to call them in the first point, or hide it so it's very hard to find the number to call.
I’ve dealt with multiple companies that hang up after hearing what your issue is, before and after AI became common.
Since it is standard industry practice, I think draconian regulations are appropriate.
Off the top of my head, if you are caught having a policy to do this, you pay all customers $100/hour retroactively for the total time they spent on hold (clawing back executive comp if necessary).
An amount equal to the total automatically goes to the whistleblowers that reported the behavior (even if they engaged in it).
> even if they engaged in it
I can't imagine any way that could be abused... (Cue "I'm going to write me a new mini-van this afternoon.")
Yeah; to avoid trouble, the call centers would need to avoid the appearance of impropriety, which is a higher bar than not engaging in the behavior.
> For all that's holy, can you just leverage the Web, please? Don't make me talk to people!
Twenty years ago I would have agreed with this sentiment.
Today... definitely not.
Leveraging the web is an unfortunately-implicit invitation to all kinds of tracking about me, my computer, my stuff. I already get too much spam, no I don't want your f***ing newsletter that I will not read when you send it to me after I already declined receiving it.
> They could still offer to register the machine by telephone as an alternative
They "do". You had the experience of that call. They literally do not want you to actually call. That's a tragedy because even in 2025 there are plenty of people who don't have smartphones, don't have computers, and many of them don't even want one. I certainly don't want to have to install some f***ing app on a smartphone just to register a washing machine. Do you?
Quite ironic for someone at Google to be complaining about a lack of long-term support for customers, when Google seems to have basically no customer support at all.
The author's example serves as a counterargument to their point.
Relying on a feature that requires enabling an experimental flag in the latest version of Chrome to work is "leveraging the web" in the worst way.
[Author here] The AI feature is their implementation on the server. I was just trying to see if the same could be implemented client-side as a good use case for the, as you say, experimental, Prompt API. (It's in origin trial, so you can use it already on production sites.) Downloading a model for just this use case is obviously stretching it. You'd use it if it's there, but not download it on purpose just for this convenience.
There is something I can't seem to understand in the relationship between technology and non-tech people / companies.
As a tech-savvy person and someone who relocated recently, I rely a lot on tech solutions for daily problems. When I was choosing my medical insurance company, I concentrated on one that has a good app and English interface. When I need a barber, I search Google Maps and look for barbers who have a website where I can book an appointment. Same for doctors, workshops, or leisure activities. And most of the websites are crap. It's some obscure reservation system, or a generic website with poor design built on Wix or something.
And from one side, I feel like there is an untapped potential here. Any good developer who can build and design a website, can just offer this to their barber or dentist. But on the other hand, I also understand that I'm not the only one who came up with this idea, so there must be a different reason why non-tech people / companies seem to neglect their digital presence. Could it be because they still think that digital presence is not important (and only tech-savvy people like me search for a barber on Google Maps)? Or maybe it's because most small businesses can't / don't want to pay a few thousands for a website and a Google Maps presence (but rather pay $39/mo for Wix), meaning that despite the fact that we are basically surrounded by tech 24/7, the general population still expects tech to be free, or nearly free.
Some thoughts based on talking with small business friends:
- most business comes from local word of mouth, so digital presence is not a priority
- existing solutions are good enough; contact info and maybe a scheduling flow is all the website needs, regardless of UX polish
- changing tech solutions (e.g. website migration) is painful; they have more important things to do with their time and energy
- will investing a few thousand dollars in a website pay off? Probably not, and hard to convince them even if true; they have more important things to do with their tight budget
- I think there's a lot of mediocre devs and agencies offering mediocre work; the talented devs delivering great, user-friendly work are probably not pursuing these projects
Most AI agents and chatbots should be a web form. Make it easy to self service!
For all that's holy, can you just say use instead of leverage, please?
[Author here] Hah, our tech writers make me replace this word every single time with "using" in our company blog posts. But this is my blog, my rules ;-)
One of the better things about the web is it's staying power. A thing you do today with HTML, plain JS and CSS will work the same after 10 years.
Contrast with apps that force you to update and redeploy every few years.
I was thinking the samething about subway. Ideally everyone would just order a sandwich online and by the time they arrive its ready to go. You scan a QR code(or enter a code) and your sandwhich pops out of hole. Having the option to make or in front of you slows down production and is awkward talking over fans and such.
Inside the heart of every customer support department there are two wolves. One is "do whatever we can to prevent them from phoning us: use every resource, every trick and every stalling measure to try to make sure they don't call". The other is "meh, just direct them to the call centre".
> The other is "meh, just direct them to the call centre".
I worked at a large insurance company and this was definitely the approach. There was a website, but you had to call to realistically get almost anything done.
One product manager's big innovation was to completely remove passwords. Every time you wanted to log in, you had reset the password and be sent a link via email. Of course the didn't announce this, so you would be probably spend 20 minutes frantically looking for your password that didn't exist.
Or, since they're already printing something unique when they print out the product number with the qr code, just print a qr code that goes to a website with the product number already encoded. No AI necessary.
edit: I see a number of people have already suggested this. Clearly a very obvious answer.
I'd say this goes for a lot of developers too. The amount of push back I see online when Chrome implements something that was previously only available in app land is weirdly high. Like do you seriously want to write separate ios app for everything?
You could just add the serial number directly to the URL in the QR code too.
I have a better proposal. Let's scrap the web. It's corrupted and broken, it hasn't been fun since MySpace. We serve native apps like VNC over kitty graphics protocol + mosh/ssh + ios/android/windows/macos mosh/ssh client. Even if we need to distribute a customized terminal to make it work. Get rid of the browser and the app stores at the same time.
How is your "customized terminal" not yet another browser?
Because it has no HTTP or HTML
So you're replacing HTTP with SSH and HTML with a VDI. How are these technical changes supposed to fix the non-technical corruption of the Web?
How about a minimum warranty period on white goods of 10 years mandated by government?
Convince me it's a bad idea?
Define white goods. Define warranty.
Would this guys experience meet your criteria? What are the consequences for failing?
Is it 10 years of normal use or still 10 years if you basically run a laundrette. What if you never used it at all. What if (and where I live this is a serious use for a washing machine) you only used the washing machine for making a special rum based cocktail?
This stuff really annoys me but the problem is really well solved elsewhere.
When you buy an Apple product it has a number on it Axxxx and a serial number somewhere. That's all you need to identify your product to anyone. That includes service manuals and spare parts.
And as far as warranty registration goes, they register it at the point of sale/activation as the warranty starting. Job done. No humans / lookups / anything required. It just happens.
Product/warranty registration is often a way to collect more data about you as a first party.
I mean, the thing is, the warranty starts automatically, there's no need to register or activate anything. As per EU consumer protection laws (the author lives in Spain).
As for finding service manuals and user manuals - maybe what they need isn't the web but FTP. I mean, if it were still supported by browsers. I remember when some vendors just used to have folders with PDFs you could browse.
The problem with "the web" is that this is no longer a website, but a content management system, or worse, a "customer engagement platform" that is hostile to creating a folder full of PDFs that have stable links. They probably still have that FTP site in a webified form somewhere for service partners, just not for Joe Public.
Author chose a brand based on loyalty instead of doing research. Leverage the web indeed - Google Search would have quickly revealed Electrolux is mediocre.
Plus, the manufacturer can't run a simple website to look up serial numbers, so author prefers their AI-based solution? And then builds one himself? This post reads like satire.
I hate the word "leverage" as a substitute for "use".
I remember this being absolutely endemic in the Microsoft world in the 00's but the first time I heard it was in the 1980s when the current commander-in-thief was a young upstart punk who leveraged the word "leverage" every chance he could get. For him the real use of leverage in business (e.g. "debt") rapidly turned a moderate-sized real estate empire into a hole in the ground and he recovered by becoming a television performer.
> LanguageModel is not supported.
Fail.
Pay for an ad, Google.
I wonder if I could register the display washers sitting in Home Depot
Your demo just returns "Identify product number LanguageModel is not available".
AI?
It's just a QR-code
If I had to guess, they don't want to leverage the web because calling the phone number to get the 10 year warranty is a dark pattern they hope most people give up on. And they know the current version of the Electrolux washer is hot garbage that won't make it 10 years, compared to the previous one that made it 20 years. Now this is not great long term for the brand, but Yannick Fierling, the new CEO, just started in January 2025 after stepping down from being CEO of Haier Europe in March 2024, a job he kept for 9 years. The previous CEO, Jonas Samuelson, started on February 1, 2016, so why would Yannick care? Yannick will probably be out with a nice golden parachute and have a shorter life than the washer. Yannick is 54 years old and I guarantee the only thing he thinks about every day is retiring soon with as much money as possible.
This enshittification cycle runs every day in the world and I don't know how we can stop it.
They may also try to use the warranty registration as an opportunity to sell insurance products, such as "protection" for other white goods. I had this with a new boiler (American English: "furnace"). They took the registration but wanted to sell me a further warranty on related heating system components not directly part of the boiler. Interestingly, it was a different company that specialises in such products. I think they probably have a business arrangement with the manufacturer to handle the warranty registration admin and possibly the warranty servicing too, presumably at a discounted price compared to the raw cost of merely servicing the warranty itself, in return for the opportunity for upselling their own products.
My guess would be more that they don't have a support org in house. Were likely running very lean on a manufacturing pipeline and are adding a lot of other services on through partnerships.
But they don't have in-house capabilities precisely because of the above-mentioned enshittification cycle: CEO-monkey does not care about long-term. CEO-monkey lowers costs today. CEO-monkey moves on to new job soon.
Maybe? It is as likely that the company focused on manufacturing for so long that they decided to keep their focus there.
Electrolux has existed for almost 100 years, and on its way it acquired a lot of other companies. The chances that none of those companies ever developed any sort of in-house assistance, over that time, are close to zero.
That is precisely a path that would lead to you not having a coherent support network, though? My assertion is that they did mergers to consolidate manufacturing. Any support orgs that they acquired along the way were not the aim of the acquisition. Quite the contrary, they complicated things in ways that are not as easy to work with as getting new manufacturing expertise.