I saw this post on Mastodon, authored by @SirBoostALot, and got permission from the author to repost it here, because I think it makes some good points:

Struggling to remember all those #Linux commands? You’re not the only one – and the Linux “experts” need to understand that. Linux “experts” who look down on GUI users are missing the point – not everyone has a photographic memory.

For over a decade I have wondered why some Linux users, particularly those of the somewhat “elitist” variety, keep saying on the one hand that everyone should consider switching to Linux but then in the next breath suggest that they need to learn how to do things from the Linux command prompt or in a terminal window. You can almost hear the contempt in their voice whenever they mention a #GUI. Worse yet, some of them seem to have the belief that everyone over the age of 5 and under the age of 75 needs to learn how to write code.

The problem with such people is that they have the same logical fallacy that many people have, which is that we think everyone’s experience is similar to our own. Objectively we know this isn’t true (well, at least most of us do) but still there is a part of us that believes that if “if I could do it, anyone can”, and worse yet, they think that if “it” was easy for them it will be easy for everyone else. And even if “it” wasn’t easy, everyone should still be able to do it with sufficient dedication and motivation. And if anyone isn’t as motivated as they are, well then they are just being lazy, or are just wanting to be “spoon-fed” information.

What I believe such people don’t realize is they have one big advantage over many of the other people they share the planet with, and that is that they have good memories, maybe even near-photographic memories. When I was in school, in the pre-internet era, many of the classes involved a lot of rote memorization. I would generally pull A’s and B’s in any subject that wasn’t highly dependent on memorization, but if memorization was involved I was lucky if I could pull a D rather than failing the class completely. Some other students would ace those kinds of classes with no sweat, because they had good memories, and I imagine some of them may have gone on to become computer programmers. Nowadays they would be the kind of people who love using the command line rather than a GUI.

See, it takes a good memory to remember all those commands and all their options. In a GUI you are shown all your options, and often if you mouse over an option there is explanatory text. And some of us really need that, because again, we don’t have good memories. We simply can’t remember all those commands and options, even if we just read a man page or help screen ten minutes ago. Some people get around this by having “cheat sheets” of the commands and options they use the most, but then that limits them to a subset of all available commands (to the ones they have previously used).

When people with poor memories ask a question about how to do something in a Linux forum and the resident Linux “expert” respond in a dismissive way, implying they didn’t search hard enough or are just wanting to be “spoon-fed” answers, that is really upsetting. Not only do they probably have no idea which search terms to use, but that is not how users respond to each other in Windows or MacOS forums. In those forums, if someone asks a question and you don’t know the answer or don’t want to give it, you keep your hands off the damn keyboard and move on to the next post. This is a skill many Linux users don’t seem to have developed, at least in some forums. If you feel it is beneath you to “spoon feed” an answer to someone then just move on, because once you have insulted someone they will never appreciate anything you have to say, and they will always think of you as an arrogant donkey’s patootie.

And that brings us to #AI, and why it is becoming so popular. A lot of Linux “experts” seem to hate AI, or see no use for it. They correctly point out that it sometimes hallucinates answers, and if you ask it a coding question you have to check pretty much every single line of code it spits out because that code may contain errors. AI’s are actually pretty stupid sometimes; they will make logical errors that most humans with any experience would not make. But still, AI’s are becoming more and more popular, and leading to decreased usage of forums such as StackExchange, for better or worse, and in my opinion there are two big reasons why.

First, they know most or all Linux commands, and have nearly perfect memories. If you have an average to poor memory, and you ask them to write a bash script to do something, they will like spit out code in a few seconds that would have taken you hours to research, and they will sometimes use Linux commands or statements that you had no idea even existed (and as time goes by, fewer and fewer of those are hallucinated commands that don’t actually exist). You may not have even known that there was a Linux command that will do a certain thing, but the AI does, particularly if it was trained on code.

Second, they always just deliver answers, They don’t chide the user for not having researched the question before asking, or for not reading a FAQ or a man page or some other piece of documentation that the user may not have known existed, or might not understand even if he had somehow stumbled across it. And they never complain that the user is being lazy, nor do that answer in that condescending tone that a few Linux “experts” that haunt various online forums seem to have adopted. I have never understood why some Linux “experts” complain that users want to be “spoon fed” answers, as if that’s a bad thing. Not everyone wants their operating system to be a puzzle, with new challenges to solve on an intermittent and unpredictable basis. Especially in the case of people coming from MacOS or Windows, most people are wanting their operating system to be as trouble-free as possible.

If you have a good memory, and if you really enjoy being presented with new challenges that can only be solved after a lot of effort and frustration, then you won’t get that most people want their operating system to be as invisible as possible, and to “just work”. And that most folks don’t want to ever have to use the command line, and they certainly don’t want to “learn” an operating system. When was the last time you heard someone say that people should “learn” Windows, or “learn” MacOS? I’m not saying that never happens, but it is a pretty rare thing to hear.

What would really be helpful, and I know some old-time Linux guys will consider this anathema, is a freely-accessible AI trained specifically on Linux and related subjects (the various Linux shells and maybe even programming languages commonly used in Linux, but with preference given to subjects that matter to users). Someplace that a new Linux user can go and ask questions and just get correct answers all day long. But also one that would that would learn and receive correction based on user input, for example if it offers a solution and nearly every user that tries that solution reports that it doesn’t work, the AI would start to prefer alternative solutions. In other words, don’t force the user to “learn Linux”; instead let the AI do it and then it can assist users in their time of need, with no condescension or other snarkiness.

And for those who would complain that AI’s use too much power, I say, seriously? How much power is consumed when users search web page after web page for an answer, mostly finding irrelevant information? And isn’t it worth using a little power so people can get their issues solved and get on with their lives, rather than spending an entire day or two trying to resolve one issue (and being insulted in the process)? And maybe to actually help those of us with terrible memories, so that we are not made to feel as if we are unworthy of using Linux?

Not all Linux “experts” are of the type mentioned in that post. I have encountered many that have been willing to go out of their way to be helpful! But, unfortunately, I have also encountered the type described above, and that’s never been pleasant.

But speaking of AI, here’s what I’d like to see. Imagine having an AI assistant that can just dive into all your past notes, articles, and even those old BASIC programs from back in the day, and then use that knowledge to help you out whenever you need it. That’s like having a personal librarian, tech support, and brainstorming partner all rolled into one!

I mean, think about it – you could just ask that AI thing, “Hey, remember that old file renaming program I had back in the day? Can you help me update that for modern Linux?” and BOOM, it’s got your back. No more digging through piles of old files and trying to piece it all together. The AI just takes care of it, using all the info it’s got stored up from your past experiences.

And the security aspect is super important, too. I mean, we all want to keep our personal stuff, well, personal, you know? So having something that runs locally and doesn’t send your data out to the internet is a must. That way, you can just feed it whatever you want, and it’ll be your own little knowledge base, ready to help you out whenever you need it.

The more I think about it, the more I’m convinced this is the kind of thing that could really change the game. It’s like having a personal assistant, but one that’s tailored specifically to your own experiences and knowledge. Someone’s probably already working on something like this as I write this. At least I hope so, because it could make our lives a whole lot easier!