here's a take that'll please no one: the belief that computers are difficult to use—especially foss and tools related to programming—is mostly constructed by politics and marketing, not by any inherent characteristics of the software. reasoning: (a) people master very complicated and user-hostile processes *all the time*. knitting, driving a car on the interstate, playing a musical instrument, being a baseball fan, using the subway—all just as complicated imo as (e.g.) learning the command line
1 Boost
1 Like
@lrhodes i don't know, i don't think that indirect feedback is unique to computers, or understood to be something that limits the ability of people to understand or engage in an activity. gardening might be a good example of another complicated activity with indirect feedback, or repairing a car engine, etc.
"People master very complicated and user-hostile processes *all the time*"
As someone who has been programming for 50 years, I learned how to drive ONCE. I also learned woodworking ONCE.
That is absolutely untrue for programming.
@bitsavers i don't know. i don't drive, but i do hire a rideshare maybe a half dozen times a year as needed, and i feel like every year i have to learn the newfangled way to *open and close the car doors*. i know i wouldn't be able to figure out how to, like, turn on the parking brake or whatever in a tesla with a touch screen
Computers are difficult to use (and program) because the people who build them have zero empathy for the user, in particular changing behaviors required to use and program them to intentionally make them different
I'm glaring at you, Apple...
@bitsavers @aparrish I have to agree with you here. And this is not even talking about programming, it’s about just using the computer to do whatever you want to do. “Doing what you want to do” with a piece of software or an operating system may be different today than it was yesterday because of an update, or a change in network configuration somewhere, etc.
(b) this is totally anecdotal, but as someone who has been teaching computer programming to adults for ~20yrs or so, the biggest obstacle that students face is not that the material is difficult (even when it *is* difficult), but that students are *culturally predisposed to a belief that they're not smart enough to program*, which acts as a permission structure for giving up. (this is peculiar, because no one having trouble in a knitting class thinks "i guess i'm not smart enough to knit")
(c) programmers are predisposed to convincing others that programming is too complicated for "normal people" as a means of job security and achieving social status; computer manufacturers and their software-making friends are predisposed to a marketing message that states: computers are too difficult for normal people, so buy this new machine/software, which is finally the computer/software that normal people can use! together both sides form a kind of mutual ratchet imo
(the trope of the "computer nerd" in media is such an effective embodiment of both sides of this rhetoric that its prevalence almost seems like a conspiracy imo)
anyway, whenever someone says something like "LLMs have finally made programming accessible to normal people!" I try to remember that the Silicon And Fossil Fuel Pyramid Scheme Industry has spent literally hundreds billions of dollars in an effort to convince you that computers were *already difficult to use* (a grift that computer programmers are largely in on!) and to politicize LLMs as being "easy to use" in contrast
@aparrish I don't disagree, but I think the case is complicated by the fact that computer feedback is often indirect. If you mess up a stitch in knitting, it's possible to look at the fabric and determine where things went wrong. If you mess up code, the thing going wrong is happening at a level of abstraction we can't see, and we have to infer the bad process by looking at output that is often the indirect result of the dysfunction. I'm not saying that makes computers hard, but it does reinforce the perception that they're hard.