This isn’t really a new problem, we just have new tool to highlight it with.
People have been developing tools to democratize software development for decades. No-code, low-code, “Watch me do” scripts, etc.
The gap is the understanding that the computer can do work for you, and you can control what work it does automatically. I’ve seen non-coders hack stuff together with a little Google-fu 20 years ago.
Generally people need a problem that’s annoying enough to need solving, and they need enough of those problems to start to build that muscle to start solving more minor things.
In college I had programming classes. I did well in them, but it felt like a paint by number. I wasn’t solving my own problems and after the class I never touched it again. It wasn’t until years later at work where some things became annoying enough to start automating stuff. Now automation is my full time job. But even with that, it took years for any of that to transfer to my personal life. I just didn’t have any problems that were annoying enough to solve.
These days when I reach for a little code to help solve a problem for someone, it’s not that they lack the tools, they lack the mental framework to make something custom, and even the hour or prompting the author mentions is too much. The problem isn’t that big for them.
I think things like Apple’s writing tools are going to be more what people actually use. For example, telling it to turn data in one format into a table. This is a really annoying task to do manually that AI can take care of with a quick prompt. Even stuff like this, I think it will take a while for people to see it enough that they think to use it.
While there are things I have today that wouldn’t exist without AI, because I wouldn’t have bothered, I think it will be the largely the same people making stuff. They’ll just make a little more. Maybe some people on the fringe will start making stuff, but I suspect they’d make stuff either way, even if it’s just some Apple Shortcuts or a little bit of Powershell.
al_borland•6mo ago
People have been developing tools to democratize software development for decades. No-code, low-code, “Watch me do” scripts, etc.
The gap is the understanding that the computer can do work for you, and you can control what work it does automatically. I’ve seen non-coders hack stuff together with a little Google-fu 20 years ago.
Generally people need a problem that’s annoying enough to need solving, and they need enough of those problems to start to build that muscle to start solving more minor things.
In college I had programming classes. I did well in them, but it felt like a paint by number. I wasn’t solving my own problems and after the class I never touched it again. It wasn’t until years later at work where some things became annoying enough to start automating stuff. Now automation is my full time job. But even with that, it took years for any of that to transfer to my personal life. I just didn’t have any problems that were annoying enough to solve.
These days when I reach for a little code to help solve a problem for someone, it’s not that they lack the tools, they lack the mental framework to make something custom, and even the hour or prompting the author mentions is too much. The problem isn’t that big for them.
I think things like Apple’s writing tools are going to be more what people actually use. For example, telling it to turn data in one format into a table. This is a really annoying task to do manually that AI can take care of with a quick prompt. Even stuff like this, I think it will take a while for people to see it enough that they think to use it.
While there are things I have today that wouldn’t exist without AI, because I wouldn’t have bothered, I think it will be the largely the same people making stuff. They’ll just make a little more. Maybe some people on the fringe will start making stuff, but I suspect they’d make stuff either way, even if it’s just some Apple Shortcuts or a little bit of Powershell.