I actually relate to that. It used to be cool and exciting to update to see what new things are coming for free. I was just looking up what Fedora Workstation 45 might bring.

Updates on phones, TV's, cars and non-Linux operating systems on PC's/laptops are downright scary to me now. What will they add that I don't want? What more connectivity to the external world are they going to wire in?

MacOS lost everyone's trust by ruining the experience, but even before it happened that trying to resize a window makes it feel like life will be at 80 years-old, hunting a tiny invisible border, even before that there wasn't anything exciting.

What Microsoft I think did (and I can't prove it), before I moved off of Windows 11, was to bundle all changes in "required security updates." So you have no choice but to update (or go through special settings to turn it off, or just defer it), but when the machine comes back you have a new Welcome prompt to some new connectivity to something nefarious. AI features I never asked for, that's the most recent dark pattern here.

The people who notice broken or degraded behavior (devs and power users) and conceive of good designs are not in charge of what goes into the product. So you shouldn't expect it to get better, only to continuously change so that the imposters can point to a difference that they were responsible for, knowing that upper management also has no idea whether that difference was actually an improvement.

The solution is to gut your product org, replace management with leading engineers, and hire some of your most fanatic users. They don't need to do anything other than give their opinions, and use the product every day. Put them in a room with your top engineers, and the software will mysteriously get better.

All of this was obvious decades ago, but now everyone wants to have a do-nothing job at a tech company, and to help their friends get one too. The result is imposters everywhere; "imposter syndrome" was brought into the vernacular to normalize a lack of expertise. "No one knows what they are doing, it's okay to not know how to do your job". That's what your (product) manager tells themself in the mirror each morning. Turns out: skill predicts quality, and an org ships quality proportional to the skill of the decision makers.

One of my absolute favourite features on KDE Plasma with Wayland is the global setting to control what can steal focus. It works wonderfully, and I always miss it when I have to use my work mac or windows computers.

See here for docs, under "Focus stealing prevention" - https://docs.kde.org/trunk_kf6/en/kwin/kcontrol/windowbehavi...

I still don't understand how a full focus stealability debounce is not the norm. If I am in the middle of writing, or interacting with one UI bit in any way, no other app should be able to steal focus, full stop. Add an annoying popup if you want, make angry computer noises, flash icons, I don't care. 99.9, if not 100% of the time, it's completely unrelated to what I'm doing and can absolutely wait. While some OSs have something along these lines, they also give the tools to bypass it (I'm looking at you, Cisco AnyConnect! Being connected after clicking on connect is not an emergency!)

Imagine you're working in a terminal and an app decides to steal STDIN. Why do we accept that for a GUI?

My experiences as a user haven't really changed. Streamers fail to cast to TVs. Browsers are devices which reliably and efficiently propagate red console text and 500s from companies to you. I still see bluescreens on public touchscreens.

Uncle Bob nailed one reason why the industry isn't getting better: the growth of programmers is exponential, such that half of all programmers have less than n years experience (5?).

I offer two additional reasons:

Firstly, you hone your skills doing individual algorithms on a single thread, where it's OK to get it wrong and change it as many times as you like.

When you get a job, you don't even realise you're in a distributed system and the getter/setter model you practiced on is woefully inadequate for interacting with partner systems, cqrs, event-sourcing, auditability, idempotency or whatever else is required.

In short, 10 years of practicing writing for-loops and data-structures will not translate into building a system that customers can access via their phone and multiple tabs of a browser, at the same time as a database schema is being updated.

Secondly, while I don't think there is any realistic alternative to agile development, it seems that nontechnical POs have captured the software lifecycle. 'Focus on happy path only, for the MVP' is how you build buggy software. Good developers can learn that this will guarantee the need for a rewrite in the future, but POs never will.

It seems like the languages have been increasingly targeting the beginner segment (C++ -> Java -> JS -> Python), while the tasks have been getting more complicated (b2b distributed systems, 24/7 uptime, millions of users, resistance to hackers, machine learning, etc.)

It turns out that the lower the barrier to creating "high quality" code is, the worse the software is on average.

Who remembers the Photoshop lens flare?

It's not that the effect sucked - it did lens flare perfectly. But who used it sucked.

Instagram and the Nashville filter - "we're all photographers now"

Code quality not equaling software quality gets at the fun part of the modern AI-labor debate. Where does talent and efficacy really reside in terms of software? Surely not in the (apparently unworthy?) mind of a human? Surely it can be extracted, distilled, automated – liberated! - unto a reactive web interface for some middle manager to invoke (at some expense to some account this folk loses zero sleep over), because it is after all this flimsy, transferrable - not worthless! Not worthless! I have a stock portfolio to maintain for God's sake. But mine. Nonetheless.

Code quality doesn't equal software quality.

People get promoted, get bonuses based on some shiny new tool they developed. Every year, they create something 'new', claim lots of features. Their managers/directors/VPs also encourage this because they need some 'winning' and 'automation' slides to brag about. Nothing gets solved. In fact, things get worse. The new tool and old tool and oldest tool all coexist with 20 other tools developed for the exact same thing in a different org. Each org (SVP reporting to CEO) is MASSIVE and everybody wants headcount and success stories. All these things will be happening at the same time. Those who are capable of making the biggest mess get the biggest pay and bonus, they get promoted and they move nearly year 1 - 2 years, within the same company or to another. Repeat. The ICs (individual contributors) who have values and explain why this is a bad idea are usually moved to Support or Ops roles or migration projects from old to new.

I've seen perfectly working systems continuously being replaced with worse and worse tools. Most of the new tools require a lot of people from several departments, just a guy who understands business and develops a perfectly working PHP/CRUD is a no-no. It has to be replaced with a complex monster with react, apis, mongo crap. And on kubernetes. Hybrid cloud.

This was true already before AI code generation, but AI code generation means the definition of “fast” has shifted like crazy. An experienced engineer can now build in an hour what would have taken them a week.

But it (right now) does nothing for the confidence about being correct - you still have to put in the extra time to ensure the software is correct to your desired degree of confidence.

Unfortunately for people who like their software to be correct (stable, performant, bug free, etc), it seems like way too many software developers right now are happy to take the gains of the former but ignore the tradeoffs of the latter.

(at the same time software quality for mainstream consumer goods like the kind of thing OP complains about in their post has been down the drain for a long time now, and it seems likely many of the examples they give are pre-AI)

There is a saying

Any idiot can build a bridge that stands, but it takes an engineer to build a bridge that barely stands

We found the limits of what it takes to make software stand, and what software paper cuts people are willing to deal with

There was a brief stint of UX, ergonomics and HMI getting traction in enterprise, and, imo, the biggest technological achievements in consumer software has been in the realm of UX in twenty first century. But it lost its importance, I suppose, as you don’t need to make good software to make money off consumers.

Whatever capabilities AI gives to programmers in terms of speed of development, etc, doesn’t change the fundamental behavior —- commercial software is not developed for humans. It is developed for investors, markets, around operational needs, based on CEOs whims, developers’ personal itches, relying on approximation of human experience that gets regurgitated through layers of commercial processes of product managers, UX researchers (who, if present at all, often don’t get invited to the conversation until there are buttons to move in a GUI of an already existing product). Where human experience is considered, it’s from the perspective of our natural weaknesses with the goal to exploit them to maximize profits. Hence why 21st century tech gave us all the dark patterns, feeds, social media, LLM chatbots, etc.

Even when I talk to individual engineers who are committed to creating something good, wholesome, something that they can be proud of -— many think and reason about the problem in a way that sounds like they are trying to engineer human experience the same way one would engineer instructions for machines.

It will never get better, IMO. AI only exemplifies what already exists.

- URLs commonly auto-link to the things behind them, even including spaces and words.

- Pinch to zoom on the homepage causes the page to reload. Forcing you to lose your place/content.

- I sometimes cannot click to play videos on the feed. I can click through the comments and then click play on the player.

- Comments pasted from perfectly valid markdown render as whitespace deleted garbage

- Clicking "reply" then tapping in the box makes it disappear. I literally cannot see what I'm typing. I have to scroll to the bottom of the page to find it.

- After typing into the comment box the "comment" button to submit it is not visible

- Being a mod: The dropdown for selecting a removal rule will not let me scroll past what shows on the initial dropdown.

- Rare but critical interfaces like banning a user (or managing an existing ban). Essentially cannot be used on mobile.

- When typing responses to people as a mod, depending on the box sometimes I "lose" the comment button.

Optimistically, I think we might just enter a new phase at some point, where businesses make very conscious decisions about what to invest in to differentiate. Like, everyone has support bots now. Where do we invest, absolutely stunning human support? Or is support not that key and we should invest into being the most reliable? The most enjoyable to use?

One can hope.

I come from sciences, where the problem of AI is more linked to AI-generated papers rather than coding. Again, the real underlying problem is the pressure to publish at all cost and the ranking grant system which rewards number of publication rather than quality.

I'm sure you could say the same of AI-generated images, sound or video used in advetising, or any field where AI generation offers a shortcut. The problem is always that you need to produce as much results as fast as possible and quality is not the objective. The problem is always the KPI.

It was simpler, better, more stable, _and_ had better UX.

Complexity does not equal quality. In fact, it usually points to poor design. Programs used to be seen as utilities, which enabled the same kind of mentality you have to any specialised machinery: "Here's a thing that can make your life easier. Learn how to use it". Now programs try to be all-singing-all-dancing-all-the-time-changing.

I keep thinking about another thing as well. Software development might be the only place were stability is punished. If your software works swellingly well, there is less people needed. You'll lose your job when the software is stable. So you either need a new project or the software needs to be crappy enough. And the companies don't give a damn about you.

Keep in mind we did have WhatsApp created with just 32 engineers until facebook bought it. Lichess.org has 2 people working on it serving to the number of people it does (millions?). It's the way we are making it. "Move fast and break things"is the shittiest way to make software.

All of the above works when the desired outcome is money. Not when it is quality of the product.

Because the premise that coding has been solved is false. :-)

It is little appreciated how USB has regressed. If you have a full-size top-of-the-line desktop machine with plenty of I/O channels you have a huge number of USB 3 and USB C plugs and will have mostly a "just works" experience -- like you can plug in an external hard drive and not regret it. Try a laptop or a pint size machine (say a Mac Mini or a Dell Optiplex 1.5 l) and you'll discover there is no reliable USB 3 hub on the market Maybe you can plug in your keyboard and mouse, maybe, but plug in a USB DAC and a memory card reader and an external drive... "BONG!" "BONG!" "BONG!" Windows starts screaming that devices are dropping out and reattaching and God forbid there is a filesystem you care about on one of them. (MacOS is too smart for that shit, it just assumes the hub is crap and won't reconnect until you reboot)

Plug a cooler master keyboard into you dev workstation and you'll find it overloads the interrupt handling system and causes everything to be janky.

Probably the worst thing about Electron is that the #1 applications use it, instead of making a web application, is that they can't stand that you might be able to find the application and start it the way you start a usual application (or web page) or be able to find the window when you are using it or be able to close the application or move the application window like a normal application window or resize the application like a normal application window or that you might be able to see anything in the tray -- so applications like Slack automatically become crapplications by virtue of inhabiting your tray. It makes everything worse. It reminds me of the later 90's when every crapplet for your printer had to have a completely custom widget set and window chrome.

The ability to generate a lot of code will not solve any of these issues.

Microsoft didn't need AI to make windows 11 as bad as it is.

The incentives are large companies are fundamentally not in align with making high quality software people actually enjoy interacting with.

Today, developers assemble a Tower of Babel built of libraries they do not understand. Those libraries were largely built the same way.

There is “no value” in a company maintaining their own purpose-optimized HTTP client, database, GUI framework, or anything else.

Each layer adds features but subtle crappiness in the details not understood.

I shudder to think how many string libraries end up in a modern system…

“Practical” solutions like “this WXYZ fails 5%” of the time, let’s add a retry loop. Of course, this the next layer sees random failures that are hard to reproduce, so they add the same.

At some point our modern supercomputers are repeating millions of operations because of a stupid bug. The user just sees a sluggish app and gets annoyed.

We could write simple, understandable, efficient software that is a joy to ship and use. Unfortunately that doesn’t add value so we are extremely careful to never waste resources on such!

If you're making a Linux distro, please don't do this. I will not watch your video. I was interested, and now I'm not.

The smaller the org, the better the chances of having only reasonable levels of misalignment... but without taking on high risk, it's just better pat to work on a large, worse org. Frustrating.

Firing people to bet against a downturn is self fulfilling. Firing people to bet for a thing everyone says is good isn't.

Obviously you cannot do that for everything. But any software you get from Debian repos (or similar distros) is at least safe from new bugs from all this new code.

Imo we know the tools (social, + technical) to make good software that works. But in the face of massive complexity, people just want the next feature, the next thing sales can use as a win

And yeah I agree with the other commenter here - I don't think this is AI. We had terrible software before, even declining quality.

I think that points do the solution. We should not make stuff more complex than we can the level of complexity at which we can make it reliable.

Programming is going to be like ceramics or pottery: you can grab a nice set of bowls at IKEA for $10. We are doing the artisanal pottery that no one really wants or cares for.

This is hilarious.

The hard part of programming almost is never how ONE function or ONE class works. It's how they all interact in a much larger picture. You can try to mitigate this, but it's a pretty fundamental concept that software is big and just too large for an LLM context window.

Until that's solved, coding has not been 'solved' by '''AI'''. They're useful for greenfield stuff, but new software was always easier to smash out quickly than working on older code.

Plus, it's not like they're building understanding like a person does regardless, even if the context window was somehow enormous.

What is coding? It is writing down specifications. Imagine a simple computer with just a 1920x1080 pixel screen, a 104 key keyboard and 3 button mouse. That is 2^107 inputs and 2^(1920x1080) outputs. The scale of specification of this (now over time) is already not handle able.

So, we invent concepts like taskbars, windows, shortcuts, and so each spec is different. People think they are writing down code, when they are actually just making a lot of decisions.

Unlike transformational software that can use mathematics as a conceptual framework, representational systems have no such thing.

Yet it is complex representational systems that allow the operation of our modern world.

Human coding effort without such a framework is problematic, and that is just exacerbated with AI generated code. Should we be surprised that for this type of software AI makes thing worse?

That's a different, equally big problem. Security mafia at every company will continue to rule everything and everyone. They do not have goals or constraints that are tied to productivity and user experience.

We are heading to a situation where the security teams are causing much more damage compared to the possible attack vectors. While this damage is real and certain, the damage due to attack vectors is only hypothetical.

Ofcourse, we do need security. But currently no one including CEO can dare to define how much security is needed. On top of that, regulators do their own share of damage, piling up regulations. No one takes any risk. They will ask you to tie yourself down until you can't move. Your mobility is not their problem.

XFCE has worked more or less the same for me for the last 20 years.

Sure, nothing lasts forever, but XFCE has been rock solid and absolutely works as expected.

I've used it successfully on both FreeBSD and various Linux flavors.

"They may have upgraded the average skill set of software teams in a way we have not seen before."

Well, no, not really. On one hand, it is true that AI makes it easier for experienced software developers to find information outside their existing knowledge boundaries. (i.e. when I need to use a language I've used before, but am not so proficient, it is faster with AI to get started). At the same time, it creates an apparent upgraded skill set. But, on the other hand, the more we rely on AI, and worse, the more we rely on AI and only on AI, the less skills we end having ourselves (we become skill renters, not skill owners). By this, we risk getting to the point where we don't really understand the code that AI generated. This is a very dangerous aspect and it might be a contributing fact to the lack of quality the author diagnosed.

It depends on whose perspective you consider the problem. An enterprise produces software for one reason only: make money out of it. And it works great!

if you are a paying customer, and if you can choose among a few different solutions, then you will get the good stuff. But for the rest...

1. Free apps? They just need to take ads in front of your eyes.

2. Software embedded in cars? You choose the car based on other merits, the software just needs to not by horrible.

3. Automated customare support? When was the last time you choose a fridge over another because of the quality of its customer support?

Isn't that how it's gone before, though? People used to make beautiful furniture by hand and leave something of themselves in it. Now it comes out of a factory, nobody knows exactly which part they made, they just fit pieces together, and it isn't built to last. Software is at that same point: nobody knows exactly what they've made or how. They describe the problem, assume they've found the solution, ship it, and there is no undoing it. The sloppiness will keep compounding. But that is also the point where people start thinking much harder about how to make it good again. A cycle, I suppose.

Everything is a minimum viable product that is constantly in flux these days. The decision makers and money movers are prioritizing rapid results and turnaround over quality product, maintainable codebases, institutional knowledge (debatable), etc

The incentive isn't focused on "Good" software, it's on making "profitable" software as fast and cheap as possible.

If anything it shows that people exposed to LLMs for the longest lost the most IQ. Googlers had access to Bart preview before LLMs were public so that checks out. I just hope those people don't get to redesign other tools as otherwise we will all need to talk to the chat box to use a calculator or a spreadsheet soon.

My other biggest pet peeve in this space is the annual iOS updates where Apple arbitrarily moves buttons back and forth between top and bottom of screen.

They really like to screw around with the photos app such that "show me my photos most recently taken in time order" is somehow different set of taps year to year.

You see flavors of this in most other apps/websites.

I could go on. That’s the network effect for you. The only reason I ever touch MS (or Meta or whatever) software is because someone I’m connected to cannot switch. And they can’t switch because others are in the same situation. At some point you don’t have to care any more. There’s no money in working software compared to igniting a massive feedback loop that just can’t stop regardless of if and what you do.

I work in mediation and have recently seen a few cases where I’ve wondered if the participants were suffering from AI psychosis as related to their claims: where they’d constructed their legal cases hand-in-hand with an LLM and had become absolutely convinced despite a contestable set of facts.

People have always been subject to delusion, but in these cases, all documentation submitted was obviously created by chatbots.

On top of that software has to be build with more users in mind so you are often going to get bloated software. That why enterprise software is often horrible because they have to support multiple industries that all need different types of workflows and UI.

I have been using software for decades and there has always been horrible software. Like I'm mind blown that people think human always wrote good code, and that AI is somehow worst than human code. AI writes better code then the vast majority of programmers that's a fact. Yes, including you. Yes there are certain areas of programming that it sucks at but 99% of programmer aren't writing that code.

But in our software we actually want language - high level expressive declarations and descriptions of process (which actually automate work).

Code is like a recipe that is also a poem.

If it wasn’t, we’d not invent higher level more dynamic languages over time. Expression is power.

You want the dual function so it’s both descriptive (documentation and digestible knowledge) and yet has productive power as an instrument.

Higher-ups can’t think of code like that - like writing. One reason I think corporations still need to evolve to become more technical and made up of more creative people.

Run it like an editor at the local paper.

Back when I was using IRC from Emacs (with erc) I had written a safeguard in elisp against that. Haven't been on IRC in a while but this helped me prevent sending to the "group chat" (the IRC channel) a great many private infos.

Same for my real name for example: I had a filter where anything containing my actual name would be blocked.

I think a great many software where you "chat" with "groups of people" would benefit from having both default and configurable filters that catch obvious mistakes. Say you probably want a confirmation before attaching a private SSH key, or before sending a password in the clear (which could be filtered by blocking words matching the hash of the password), etc.

Not that it solves the fundamental problem TFA is talking about: nothing works.

But that is a given: that's why we need safeguards.

If it's open source and I can review it before compiling and installing, and I can modify it and recompile if I want to remove/change/add something, then I feel like I have some control

This is one reason I like relatively smaller software, less code

I don't necessarily believe in 10x devs, but there is a very clear output difference between various developers. During my career I've been both in a position where I was delivering much higher quality code at a similar pace to other coworkers and also mediocre code at a slower pace. Sometimes at the same time.

Now, AI itself is a force multiplier. Depending on the task, your setup, your skill level (both as a developer and with LLMs), the weather, the model and so on, it can range anywhere from actively hindering you, to solving the entire thing for you. Let's say that on average you get a 20% productivity boost.

If you take a 0.8x developer and give them a 1.2 force multiplier LLM, you get a 0.96x developer. Worse, the axis that has moved is speed, not quality.

All this to say that if you let a below average developer vibecode, you get a much higher volume of slop. You can make an agent create good code, but this developer didn't care before so why should they now?

*Note that this is from experience. I've started getting feature PRs from our devops team. It's been a very enlightening experience.

It's just that complexity has skyrocketed compared to the past. The same goes for released software. In the old days, even small software could be competitive, but now the minimum level of complexity required to sell a program is completely different.

The best example of this is mobile games. Just look at the structure of early mobile games compared to modern games.

AI is precisely what is needed for high speed of coding.

Yes, coding is solved, for the developers, not for the users.

If you look at the open source space and review the contribution of AI in it, you'd think that AI is barely able to create functional code and the people who peddle it are solely slop authors. More than that, you might think it barely exists, because there's simply vanishingly few PRs landing which are AI generated. It seems in open source projects, for some reason, AI is barely able to produce any useful code at all

But apparently for closed source proprietary code its like magic and you don't need to write any code anymore and it is a solved problem. Mhm. Totally willing to believe that

My suspicion is that a lot of proprietary software engineering has become purely performative. It no longer matters if the resulting product works, or has any quality, because its not about that. The system is being burnt down to extract wealth from it instead, as we create the illusion of software development as these companies are asset stripped. Line Must Go Up. There's no simply incentive to make a real product anymore, when the valuation of our economy and management compensation is built solely on hype instead of results

Making legacy code however is now very fast. So fast you can practically ship it straight to production.

Expectations is neither about creation not use. "May" isn't a revolution outside of the historic calendar, so on both counts it's trivial to deny an ungrounded statement.

> So why does software keep getting worse across the board?

Because the premise on which the expectation is built is wrong?

Also plenty of the issues are prehistoric with all the past meat bag generations to blame, case in point: > Just then, the Slack window appeared, stole focus from Ghostty

So who stopped that huge fruit corpo from doing the very basics of managing user experience in the past few decades minus a few AI years?

It's glaringly obvious that poorly checked, AI-written, code is the culprit, I'm just wondering when companies will wake up.

Nobody gives a flying fuck about resiliency, good practices, security or any of that now that AI is there. All anyone cares about is pure output of features.

Any and all "productivity" gains attributed to AI efficiency are actually due to the complete disregard for any care or standard in the software industry in the past 2 years. Yeah you can output more shit when you don't give a fuck, big surprise, AI doesn't change anything here.

A product isn't necessarily designed by someone who cares about, or even understands, the problem at hand.

Part of it is incumbancy - much of the Mac OS level stuff is 30 years old and is not going to change.

But the real concern is that even new-builds are not particularly great - and its because we dont have the abstractions in place to help us scale complexity

And the lack of backlash from anyone with real influence when software fails.

That is software is bad because organizations make it so not because the coders are incompetent.

I will say that in my experience with AI my code and that of my employer has greatly improved. So maybe there’s a split between big companies and small as far as quality gap?

Now AI tests? Except not really. It claims to perform testing. It’s still up to us to review and remediate.

to the article: Yeah, it's different world out there. We need to filter ideas instead of focusing on more features, as lack of staggering revenue growth suggests it might be that more doesn't equal better. The stability of apps is getting worse lately. I couldn't pair my sony camera and it was stuck on pairing screen :(

It won’t change cause nobody cares to do anything other than complain, I don’t see developers quitting en masse from companies that refuse to pause feature churn and focus on quality. Why should they? They get a paycheck, everyone ships crap, the world moves on. The people making the frameworks and now even runtimes are of the same cloth, overworked and underfunded.

If we cared we’d treat software like real engineering. We don’t and therefore we’re quite cooked. Just be happy that airplanes aren’t as bad as your average CRUD. For someone who does care, look at the SQLite codebase, every project should be like that even if development cycles would be 3–10x longer, but that’s delusional to say out loud.

AI just revealed that - with slop being produced at increasing pace and test coverage and verification methods of things working properly across all states and inputs being nowhere to be found.

IME this happens more often than people think. If you’re leading a team part of your job is to manage the expectations of those around you and explain your choices. If you do that well you can absolutely have quarters where you only pay down debt, it doesn’t require some huge attitude shift in the industry. Just explain why it’s the right thing and then do it.

None of this is to say that I think this is a good thing, or even that it's necessarily an accurate viewpoint. Considering who decides how much to spend on LLM products in companies that are potential customers for the ones saying "coding has been solved", I do genuinely think this is what they mean though. It's of course nonsense to actual developers, but we're not the ones who they're talking to when they say this.

A lot of this is about bureaucracy

These products are bad because noone who works on them understands the whole system infact

"The only trades which it seems possible for a joint-stock company to carry on successfully... are those of which all the operations are capable of being reduced to what is called a Routine, or to such a Uniformity of method as admits but little or no variation." - Adam Smith

CPUs, dams, can be designed with software. Plenty of programs can be designed with software but most useful, interesting, valuable ones cannot. Plenty of programs are designed with software, or more generally a routine, and they are incomprehensible as a result, and clearly their programmers are the most stupid and ignorant people around.

This extends to any device which has software or firmware: banking websites, banking apps on mobile phones, car firmware, etc.

I find it very depressing, especially when I think of the people who design, develop and deploy these apps. We are talking about large companies and teams of people. Entire chains of well-paid people designed, looked at, developed, tested, and then accepted these apps. And yet they break in the most obvious ways, or fail to be usable, instead pestering us with pop-ups and useless drivel and forgetting their state at every point.

These problems are not subtle. To take a glaring example from my everyday life, in my car there are buttons for moving windows up and down. Sometimes when you press the button to move the window down, it will actually start moving up. We are talking about a simple button that moves a window. And seemingly everybody thinks this is fine?

I don't know what to do about this.

The issue is only partially within those who code. They get directions from those in charge, and they are just making idiotic decisions. And those who code, for some reason don't stand against those idiotic decisions in defense of a common sense and a great UX.

A blatant, exaggerated parallel would be: you are designing your kitchen and you've chosen a place for your washing machine such as that it will be unable to open its door. When inviting a professional, you expect them to correct such wrong decisions and explain to you that it won't work. But in programming, those who program, calling themselves "software engineers", would just place this washing machine where they're told, because it's "not my job". It's not a feature of an "engineer". It's being a coder lemming.

Do you A) tell your management hey it's absolutely perfect as is, let's just chill, leave it alone and let the money roll in. Or B) tell management about the need for whatever new hotness and you're going to get it implemented ahead of schedule and under budget. And then you do! Oh but the software is objectively worse because it was already perfect.

It's pretty clear what the overwhelmingly popular answer is to that question.

At the end of the day, fixing bugs, especially those bugs only happen in certain environments, is a very tedious and thankless task.

> My LG fridge started making weird sounds, so I tried to file a warranty claim, through a multi-step form with countless fields. It failed with a submission error at the very end. And I only found out because I looked at the JavaScript console.

Has that JavaScript code been written by AI? It may be more likely that it was written by humans. Well, the design of JavaScript really sucks, the language is both successful and horrible to use, in my opinion. And we have that issue across the board. Most programming languages really suck. C, for instance - that language is an abomination. But it is even more successful than JavaScript. Today I ran into the issue that I can not compile libpng; I keep on bailing out at it unable to find setjmp. About two months ago compiling it worked fine. I most likely did screw something up on my local system but ... the compiler is an idiot. It really does not tell me WHY it can not compile it anymore. Yes, it shows on the commandline that it assumes setjmp to exist but it can not find it. The error messages are pure and bogus. And the whole design of C cascades this horrible crap downwards. The languages that supposedly try to replace C also suck. Go? Yikes. Rust? Even more verbose. C++? Trying to replace C since decades. There is a ton of inertia.

The whole stack is utter gutter trash really. Everything is build up atop layers of complexity that in turn are not really well designed at all. Meanwhile more and more software is added. Now I need python due to meson, and meson is better than what came before. But it adds complexity. And has its own problems too. AI does not really fix anything here, it just AI slop spams on top of this. We kind of lost big at engineering. People are in general also not critical anymore. The design of most programming language shows that - including ruby, but ruby is barely used, so really the issue is mostly about C. The fact that C is so successful makes this even worse, since nothing replaces it. We just keep on adding more and more. Or we go the C++ way - and make it about 50x as complex. Or Rust - that syntax is even worse than C syntax which is also horrible. Of course even with great syntax and a great language, horrible software will be written. Companies want to offer a billion useless features. Many years ago I thought it was cool if my fridge tells me what I should order. Since the day of companies spying on people, I go back to simplicity at all times.

I'm confident to say that right now, AI already does a better job at turning a well-specified requirement into working code than the average meatbag I've personally encountered.

It's still useless at many other things that make a complete Software Developer, but let's be real: Programming (requirements in, code out) is solved for the stuff I'm involved in.

Coding has been solved in the sense that, per dollar spent, companies can get 10x or even 100x the amount of code of roughly the same terrible quality.

I mean, he has given the answer in the article: Because "coding has been solved" is from the perspective of upper management, not devs or users - and that cares about software quality and usability (as standalone metrics) about as much as the McDonald's board cares about nutrition value.

Name and shame! I'm in the market for a car. Which one should I avoid?

> They have raised upper management’s expectations for team output.

That is the worse part, it's not great at all dealing with this kind of expectations

Hint: mid-level engineers in tech hubs wouldn’t be getting 500K salaries if that was actually the case.

So taking it at face value is kinda part of the problem here.

The only way this can get better is by completely suffocating it off attention. It doesn't exist. It's not real.

Euphoric, who?

Other than techbros and a small group of AI-delirious fanboys, no one is euphoric. Normal people in the industry are between anxious, scared, exhausted, resentful, resigned or just plain sad. People outside our industry? They see the enshittification clearly, and are most certainly the opposite of euphoric.

But I would say the engineering barrier and the nerd barrier did provide some amount of sociopathic resistance in the world.

Llms are going to enable the absolutely immoral business people to do whatever they want. And that's going to mean enshittification on a massive scale

Thanks capitalism, for ruining literally everything yet again

Let's have a look at a few rarely quoted examples from the last decade, shall we?

Just six years ago, a big company rolls out a line of new chips so good that geeks and technology experts all get ecstatic and can't stop churning endless raving reviews. Those chips are fast, small and consume little energy. On top of that, the big company reworks their entire OS (OS!), the whole suite of software it has for it, the developer suite with compilers and SDKs and god knows what else and also ensures that everything ever written for that OS for the old chips continues to work and it indeed does.

Then there's another big company that launched a file storage service around 20 years ago. Back then it was reliable and only gotten better over the years. They started offering their storage for a nanocent per gajibyte and twenty years of hard work allowed them to lower that cost to a picocent per gajibyte. They don't lose data, they are highly available, half of the internet stores their data at their servers.

And then there's a -- gasp! -- consortium of big companies (fools would reply to this with an XKCD standards strip) that works on improvements to a shared set of standards for 4G, one of the most widely used technology stacks in the world, agrees on it, calls it 5G and rolls it out across the world and the whole spectrum of hardware and software vendors in a manner that resembles replacing a wing of an aircraft mid-flight so that not a single pasta strand falls from the forks of the business class customers.

And here we have someone who thinks that a customer support page for his fridge (fridge!) thrown together by a low-paid second-grade developer in a third-world country is indicative of the whole industry decline.

"Nothing works", my hairy bum. Get real.

It was always better in the past. The capitalists always ruined it in the past. Only I am the true bearer of the artisanal software flame. /s