On one hand, it's a "people problem"... on the other, it's a software problem too.
"back in my time", you'd have some c code, one .conf file, you'd "make", edit the config (or hope it works with default settings), run, and you'd have a program running. Now you need five different services running, it comes in a docker, running on some random port, proxied to another random port, the configs are split into 12 yaml files, plus it needs 7 gigs of hdd space...
..sometimes providing the same functionality as the old 300kB software of the yesteryear.
The purpose of what he describes is to preserve the structure of denial around the implications of possessing, and the consequences of using, any or all technology. Politics as found -- war, law, class, etiquette -- depends entirely upon that structure of denial.
> These are the same people who would lose their minds if their city government told them they could only buy food from vendors the city had approved, licensed, and taxed
But it is exactly like this in the developed world, and not many would buy food from a trunk of a roadside car.
You see it in how services have twisted the meaning of "saving" something. Very rarely does it mean actually putting a file on your computer for you to access on your own terms as long as you have possession of it. More often, it just means associating something to your account, which is ultimately subject to the whims of the service provider.
I see this even within the developer community, people who understand systems programming is becoming rarer and rarer. Developers are constantly looking for GUI solutions and don't bother understanding how to script. I think this likely stemmed from a similar issue as introducing the smartphone, we software engineers separated the disciplines of DevOps from Software Engineering, this led to people not caring about the infrastructure at the background of their solutions.
"Don't Make Me Think" wasn't descriptive. It was prescriptive that became descriptive. An entire professional class, PMs and UX designers, adopted as axiomatic that cognitive effort is friction, friction is bad, therefore understanding is a design failure.
Then they spent 25 years engineering understanding out of every single interaction, and now point to the resulting learned helplessness as validation. "See? Users don't read!" No, you spent decades training them not to by ensuring that reading was never rewarded and never necessary.
My read has always been it was painful for a certain type of PM to think and so they assumed "minds like mine" and ... here we are.
> Mobile Platforms Did the Most Damage, and They Did It on Purpose
So true, but this has been going on for quite a while. Phones accelerated it and I have seem many of the concerns come up in IT where I worked.
A couple of examples:
1. My favorite, about 10 to 15 years ago. A user said this finance report always had 2147483647 in the total. This was looked at for weeks by another group.
After a few weeks our manager's manager called a meeting with everyone to look at the issue. Everyone had no idea what to do. When I saw the number it look real familiar to me. I then released it was the max value if an int. I told them the issue was its variable could was too small. A simple change fixed the issue.
Another old programmer who was not at the meeting asked me what happened. I showed him the report and he know instantly what it was too.
2. hex dumps, no one can read them now. About 25 years ago I was looking at a dump to see where a packed numeric value was, people who saw be thought it was magic. I had to explain how that number was read and what the hex represented.
I fear what will happen if AI becomes a real thing.
Most people, when they turn the tap on, they don't know where the water comes from. Try asking someone "What is the physical principle that makes the water come out of the tap? How do they make it come out?" you might be surprised how many people don't know and most importantly don't care.
The water comes out. The water has always come out, every time, so it's not really a thing worth investigating. Like the sunrise.
In many many domains I am that person.
If a person doesn't know (except in the vaguest terms) where their water comes from, where their poo goes when they flush, where their food comes from (the supermarket!), or the energy that heats their home... what do they really know? Most of us know very little about the concrete networks and systems that keep us alive.
> All of this was sold as a feature. “It just works.” Safety. Privacy. User experience. What it actually was, was control — Apple’s control over what you could do with hardware you supposedly bought.
Can't both of these be true?
Apple, Microsoft, and the evil tech companies us nerds love to vilify actually brought computing to the masses. In the early 1980s only tech geeks and corporations used computers—the Apple Macintosh changed that. In the early 1990s, only tech geeks and universities used the internet—Microsoft Windows changed that. In the mid 2000's, only tech geeks and business people used "smart" devices[1]—the iPhone changed that.
At every technological leap, business savvy entrepreneurs saw an opportunity to expand their markets by making their products enticing and useful to millions of more people than the previous generation of products did.
Unfortunately, this also came at the expense of the apparent "dumbing down" of computers, as every new abstraction hid more of the actual computer users had to interface with. And it also made things easy to control and lock down for corporations.
But I don't think we would've seen the explosion in the popularity of computing had this played out any other way.
I also disagree with the article's premise that power users are dying. We're still here, but we're a tiny minority of computer users now. We're both amused and frustrated at the insanity of where technology is taking us, and who is leading us there, but we still have our corners of computing we can retreat to.
And I also disagree that our favorite layer of computing is somehow more "real" than anyone else's. We scoff at Gen Z's inability to use the terminal as much as Baby Boomers scoffed at our inability to program in assembly. It's all relative. Except "AI". That is more of a disabler than an enabler, even though we're too hypnotized to see it now.
[1]: Yes, the BlackBerry was a cultural phenomenon, but it didn't have the capabilities nor mass appeal of the iPhone.
I read the text. It's something I would agree in principle, but the text is exceedingly whiny and in many ways wrong.
Power users, tinkerers, and so on were always extremely niche. By definition they were always only a few. They are still a few, probably in similar numbers as before.
The only thing that changed is that normal people now have access to computing devices. My wife does not want to know what a file system is, or what happens behind the scenes when an app is installed. She has no idea what a DNS is. Why would she? She is a lawyer with little interest in technology. She wants to use Instagram, not self-host a Matrix instance.
The normal users are the majority, and it's more profitable to serve those users than assholes like me who get pissy when they can't sideload an apk.
> Ask a twenty-two-year-old to connect to a remote server via SSH. Ask them to explain what DNS is at a conceptual level.
Modern IT has become a ubiquitous commodity, much like the car. You don't need to know how an engine works to drive; while that knowledge might make you more efficient, it isn't strictly necessary to get from A to B. Besides, most twenty-two-year-olds ten years ago didn't know how to use ssh, either.
However, if you want to call yourself an engineer (and work in the field), you must understand the underlying mechanics. IMHO if you want to defeat a competitor today, you don’t need industrial espionage - you just have to cut their internet and/or AI subscriptions. Modern vibe engineers would struggle to function.
> The man page is dead for most users. The RFC is unread by most developers who depend on the protocols it describes.
Well, those who are accustomed to using man pages still use them today. I find them far more accurate than whatever an AI might spit out at any given moment. As for RFCs, they were always read by a small population - either those implementing the protocols or the few of us who like to brag about obscure technical details.
> You can now write complete programs without understanding what a single line of them does... until something goes wrong in production at two in the morning and you are completely without tools to respond.
I’m not worried about this. When things go south, there will still be experts who will know how to fix them. But since those experts will be fewer and farther between, they will likely charge $1k/hr, and rightfully so. If you are in that field, more power to you! :D
I take issue with this. Knowledge like connecting to a remote server via SSH or explaining what DNS is have ALWAYS been niche topics. The article claims these are things you learned in the first week of "seriously engaging" with a computer twenty years ago, and that's just false.
I find the author is projecting heavily having entirely bought into the unix way of doing things he's become his own jailer. That a filesystem is somehow "fundamental' to the function of a computer. Nope, it's an abstraction designed to help the programmer, it should be thrown away. These kind of unified approaches are going to fail to specialization every time. There are endless things to understand, deprecated tech is at the very end of that list.
If it was only the Apples' and Google's who thought sandboxed apps were the future you might have a point, but most tech savvy people arrive at something that looks and awful lot like sandboxed apps. You see power, most see[1] a[2] dumpster-fire[3]
So true that's why I love the Google? interview question: how does a modern browser work (basically one needs to understand the entire journey from a file on the server, through DNS, and browser rendering engine). Sadly, this is becoming lost to many folks.
Pre-AI, I worked with Devs that didn't even know what an HTTP request is (difference between GET/POST/etc) - we were building an enterprise software where higher level libraries abstracted that away..With AI, it's becoming even worse now - just ask Claude
I don't know if I agree. Look at modern application development: even building the most basic GUI application in any platform, but especially the web, is significantly more "power-user" than Visual Basic ever was.
People are still building these apps at every level of developer experience so the kids must be alright.
Yes people, in general, have less knowledge of file systems and networking now because things just work. Every LAN party as a kid took at least an hour of networking to get started. Now kids don't do LAN parties because everything is already networked by default. But there is also an order of magnitude more people doing these things now -- in the past those people just went without.
The vast majority most of knowledge I ever had from the time the author describes is obsolete now anyway. I can still remember so much but that's not going to help me with my React app.
> “I have nothing to hide” is the response is the response, which is not an argument — it’s a thought-terminating cliché that makes it socially awkward to point out that privacy is not about criminality, it’s about power. Whoever holds your behavioral data holds power over you. That’s true whether or not you’ve done anything wrong.
To put a slightly finer point on it for puncturing the cliche, "needing to hide" has a time-component. Everyone has something to hide from a potential future, whether they're good at predicting it or not.
I "have nothing to hide" about my religion today, but when if extremists seize power and declare "death to apostates", the exact same fact-pattern will very very much need hiding.
> The industry isn’t going to fix this. Every financial incentive points the other way.
Cory Doctorow has a hopeful--perhaps over-hopeful--idea that a disruptive wedge can be created, where a profit-motive will promote breaking the system of control. Specifically, that some place with a legal haven for tinkerers and wall-breakers will reap benefits from letting them openly sell device-unlockers, export-your-data tools, etc.
Thank you for sharing this article. I found it very insightful while eloquently proposing root causes. In particular, I found this bit to be wonderful: "The smartphone didn’t just shift computing to a smaller screen. It replaced a computing paradigm — one built on ownership, modification, and composability — with a consumption paradigm built on managed access, curated experience, and dependency."
That's... Normal. Technology has always been moving towards higher-level abstractions. In terms of software, many engineers nowadays know how to code in high-level languages like JS or Java, while maybe 30-40 years ago many folks probably knew C, assembly, and all the low-level stuff like e.g. explicit memory management that most modern devs never deal with.
While on one hand I suggest revisiting the old Ironies of Automation by Lisanne Bainbridge on the other I've noticed that what's really missing everywhere is IT education. I come from a Computer Engineering background and I saw absolutely NOTHING regarding IT except for a few mentions here and there; I had to teach myself. Those coming from CS see even less (at least speaking for EU universities, I don't know about the US but I suspect it's not much different).
There's an attempt to deny the need for IT knowledge and expertise at every level, Big Tech does it out of self-interest, while most others do it out of ignorance. They often claim, "Oh, it only takes a few minutes or days on your own; just a couple of clicks and you can do everything." Yet, those who say this don't actually know what they're talking about and refuse to even try to prove their own theory.
The outcome is even worse: nowadays, doing it yourself is a struggle even when you have the right skills. All recent software is built to be unmanageable because there's no operation/infra vision. Don't even get me started on documentation; everyone talks about the need for a "documentation culture" yet what actually gets documented ranges from nothing to total garbage (basically text that's useless unless you feed it to an LLM and hope it can make some sense of it).
To make matters worse, standard hardware is getting more and more expensive, first it was graphics cards, then RAM, and now NVMEs, with the result that many people simply don't want to or can't afford to buy, so they're literally living on someone else's computers even if they don't like it. This is especially true for students, who are at the best stage of their lives for learning and who won't have the time or energy to do it later on.
To complete the picture, the business model just isn't sustainable; no matter how much is invested, a real digital evolution isn't possible while living on the computers of four giants limited by their own services, and this implies that a social collapse awaits us regardless.
For me, the solution is managing to have enough leverage so that we can push for mandatory FLOSS and open hardware de jure in response, in order to limit the damage and geopolitical upheavals who push anyone to relocalize, which necessarily implies starting over on a small scale. I see something coming: Nostr, Meshtastic, the Fediverse, the rise of self-hosters and their average age show that there's still an active group of people who want a different world. But they are few and far between, burdened by significant technical debt in a world that's becoming increasingly hostile,and that's exactly where things need to change.
The problems caused by centralization, from various companies getting burned by relying on giant third-party providers, to banking scandals driving crypto (not stablecoins), to the need for resilience that requires cutting down on SPOF might actually make a difference. I hope it'll be enough, and I hope anyone who gets it does their part to spread that understanding while we still can.
36 comments
[ 2.8 ms ] story [ 58.7 ms ] thread"back in my time", you'd have some c code, one .conf file, you'd "make", edit the config (or hope it works with default settings), run, and you'd have a program running. Now you need five different services running, it comes in a docker, running on some random port, proxied to another random port, the configs are split into 12 yaml files, plus it needs 7 gigs of hdd space...
..sometimes providing the same functionality as the old 300kB software of the yesteryear.
> These are the same people who would lose their minds if their city government told them they could only buy food from vendors the city had approved, licensed, and taxed
But it is exactly like this in the developed world, and not many would buy food from a trunk of a roadside car.
It's engineered dependency.
These are the type of persons who would get a girlfriend without a master degree in psychology.
Then they spent 25 years engineering understanding out of every single interaction, and now point to the resulting learned helplessness as validation. "See? Users don't read!" No, you spent decades training them not to by ensuring that reading was never rewarded and never necessary.
My read has always been it was painful for a certain type of PM to think and so they assumed "minds like mine" and ... here we are.
So true, but this has been going on for quite a while. Phones accelerated it and I have seem many of the concerns come up in IT where I worked.
A couple of examples:
1. My favorite, about 10 to 15 years ago. A user said this finance report always had 2147483647 in the total. This was looked at for weeks by another group.
After a few weeks our manager's manager called a meeting with everyone to look at the issue. Everyone had no idea what to do. When I saw the number it look real familiar to me. I then released it was the max value if an int. I told them the issue was its variable could was too small. A simple change fixed the issue.
Another old programmer who was not at the meeting asked me what happened. I showed him the report and he know instantly what it was too.
2. hex dumps, no one can read them now. About 25 years ago I was looking at a dump to see where a packed numeric value was, people who saw be thought it was magic. I had to explain how that number was read and what the hex represented.
I fear what will happen if AI becomes a real thing.
The water comes out. The water has always come out, every time, so it's not really a thing worth investigating. Like the sunrise.
In many many domains I am that person.
If a person doesn't know (except in the vaguest terms) where their water comes from, where their poo goes when they flush, where their food comes from (the supermarket!), or the energy that heats their home... what do they really know? Most of us know very little about the concrete networks and systems that keep us alive.
But this is what civilisation is.
Can't both of these be true?
Apple, Microsoft, and the evil tech companies us nerds love to vilify actually brought computing to the masses. In the early 1980s only tech geeks and corporations used computers—the Apple Macintosh changed that. In the early 1990s, only tech geeks and universities used the internet—Microsoft Windows changed that. In the mid 2000's, only tech geeks and business people used "smart" devices[1]—the iPhone changed that.
At every technological leap, business savvy entrepreneurs saw an opportunity to expand their markets by making their products enticing and useful to millions of more people than the previous generation of products did.
Unfortunately, this also came at the expense of the apparent "dumbing down" of computers, as every new abstraction hid more of the actual computer users had to interface with. And it also made things easy to control and lock down for corporations.
But I don't think we would've seen the explosion in the popularity of computing had this played out any other way.
I also disagree with the article's premise that power users are dying. We're still here, but we're a tiny minority of computer users now. We're both amused and frustrated at the insanity of where technology is taking us, and who is leading us there, but we still have our corners of computing we can retreat to.
And I also disagree that our favorite layer of computing is somehow more "real" than anyone else's. We scoff at Gen Z's inability to use the terminal as much as Baby Boomers scoffed at our inability to program in assembly. It's all relative. Except "AI". That is more of a disabler than an enabler, even though we're too hypnotized to see it now.
[1]: Yes, the BlackBerry was a cultural phenomenon, but it didn't have the capabilities nor mass appeal of the iPhone.
Power users, tinkerers, and so on were always extremely niche. By definition they were always only a few. They are still a few, probably in similar numbers as before.
The only thing that changed is that normal people now have access to computing devices. My wife does not want to know what a file system is, or what happens behind the scenes when an app is installed. She has no idea what a DNS is. Why would she? She is a lawyer with little interest in technology. She wants to use Instagram, not self-host a Matrix instance.
The normal users are the majority, and it's more profitable to serve those users than assholes like me who get pissy when they can't sideload an apk.
And I am okay with that.
Modern IT has become a ubiquitous commodity, much like the car. You don't need to know how an engine works to drive; while that knowledge might make you more efficient, it isn't strictly necessary to get from A to B. Besides, most twenty-two-year-olds ten years ago didn't know how to use ssh, either.
However, if you want to call yourself an engineer (and work in the field), you must understand the underlying mechanics. IMHO if you want to defeat a competitor today, you don’t need industrial espionage - you just have to cut their internet and/or AI subscriptions. Modern vibe engineers would struggle to function.
> The man page is dead for most users. The RFC is unread by most developers who depend on the protocols it describes.
Well, those who are accustomed to using man pages still use them today. I find them far more accurate than whatever an AI might spit out at any given moment. As for RFCs, they were always read by a small population - either those implementing the protocols or the few of us who like to brag about obscure technical details.
> You can now write complete programs without understanding what a single line of them does... until something goes wrong in production at two in the morning and you are completely without tools to respond.
I’m not worried about this. When things go south, there will still be experts who will know how to fix them. But since those experts will be fewer and farther between, they will likely charge $1k/hr, and rightfully so. If you are in that field, more power to you! :D
In my time, not being able to read assembler code meant you weren't a power user.
If it was only the Apples' and Google's who thought sandboxed apps were the future you might have a point, but most tech savvy people arrive at something that looks and awful lot like sandboxed apps. You see power, most see[1] a[2] dumpster-fire[3]
: [1] ; ls /usr/lib | wc -l
: [2] ; ls /usr/bin | wc -l
: [3] ; find /usr/share/man/man* | wc -l
Pre-AI, I worked with Devs that didn't even know what an HTTP request is (difference between GET/POST/etc) - we were building an enterprise software where higher level libraries abstracted that away..With AI, it's becoming even worse now - just ask Claude
People are still building these apps at every level of developer experience so the kids must be alright.
Yes people, in general, have less knowledge of file systems and networking now because things just work. Every LAN party as a kid took at least an hour of networking to get started. Now kids don't do LAN parties because everything is already networked by default. But there is also an order of magnitude more people doing these things now -- in the past those people just went without.
The vast majority most of knowledge I ever had from the time the author describes is obsolete now anyway. I can still remember so much but that's not going to help me with my React app.
To put a slightly finer point on it for puncturing the cliche, "needing to hide" has a time-component. Everyone has something to hide from a potential future, whether they're good at predicting it or not.
I "have nothing to hide" about my religion today, but when if extremists seize power and declare "death to apostates", the exact same fact-pattern will very very much need hiding.
> The industry isn’t going to fix this. Every financial incentive points the other way.
Cory Doctorow has a hopeful--perhaps over-hopeful--idea that a disruptive wedge can be created, where a profit-motive will promote breaking the system of control. Specifically, that some place with a legal haven for tinkerers and wall-breakers will reap benefits from letting them openly sell device-unlockers, export-your-data tools, etc.
There's an attempt to deny the need for IT knowledge and expertise at every level, Big Tech does it out of self-interest, while most others do it out of ignorance. They often claim, "Oh, it only takes a few minutes or days on your own; just a couple of clicks and you can do everything." Yet, those who say this don't actually know what they're talking about and refuse to even try to prove their own theory.
The outcome is even worse: nowadays, doing it yourself is a struggle even when you have the right skills. All recent software is built to be unmanageable because there's no operation/infra vision. Don't even get me started on documentation; everyone talks about the need for a "documentation culture" yet what actually gets documented ranges from nothing to total garbage (basically text that's useless unless you feed it to an LLM and hope it can make some sense of it).
To make matters worse, standard hardware is getting more and more expensive, first it was graphics cards, then RAM, and now NVMEs, with the result that many people simply don't want to or can't afford to buy, so they're literally living on someone else's computers even if they don't like it. This is especially true for students, who are at the best stage of their lives for learning and who won't have the time or energy to do it later on.
To complete the picture, the business model just isn't sustainable; no matter how much is invested, a real digital evolution isn't possible while living on the computers of four giants limited by their own services, and this implies that a social collapse awaits us regardless.
For me, the solution is managing to have enough leverage so that we can push for mandatory FLOSS and open hardware de jure in response, in order to limit the damage and geopolitical upheavals who push anyone to relocalize, which necessarily implies starting over on a small scale. I see something coming: Nostr, Meshtastic, the Fediverse, the rise of self-hosters and their average age show that there's still an active group of people who want a different world. But they are few and far between, burdened by significant technical debt in a world that's becoming increasingly hostile,and that's exactly where things need to change.
The problems caused by centralization, from various companies getting burned by relying on giant third-party providers, to banking scandals driving crypto (not stablecoins), to the need for resilience that requires cutting down on SPOF might actually make a difference. I hope it'll be enough, and I hope anyone who gets it does their part to spread that understanding while we still can.