So get interested in the needs that will never go away. I consulted once in the ready-mix concrete industry: surprisingly deep and intriguing. Another time it was a manufacturer of remote-sensing equipment for airports. Another time it was paper mills. None of those things are going away. Build a reputation as a troubleshooter.
> None of those things are going away. Build a reputation as a troubleshooter.
Not going away, sure, but are they getting automated?
Even troubleshooting.
This is where I'm at, right now: I could learn whatever, but how do I rate the various options' chances of being automated? Not "what do I think the chances are?", but how do I even sensibly approach the question?
I agree everything is kind of crap nowadays, but I also think that by 2016 we were already well on our way to the current situation. Adtech, social media, the iPhone (and subsequent smartphones), all were major contributors to the decline.
I wouldn’t call what we have now “dead” (though you seem to be using a loose definition). Exploitative, boring, harmful, sure. But it’s very much alive and sucking the life out of us.
Indeed. A lot of the stuff that's come out in the last decade has felt like people reinventing existing wheels, badly, because they didn't understand what the existing wheels were for, how they worked, or why they were there.
For example (while I don't know more than the basics of React on JS/HTML), SwiftUI is Apple's way to be reactive on iOS, but it comes with lots of "magic" that doesn't work right, missing bits, and feels like it only exists at all because enough people in the right places liked the right buzzwords.
While there were issues with version control on the old Interface Builder files (all that big blob of mystery XML), I've never seen a codebase or development process which was actually made easier (neither to develop nor to debug) with SwiftUI. If it had been up to me, the change there would be to have Apple buy up Figma and have it export UI directly: at least there, there (used to be, pre-AI) an obvious needlessly expensive duplication of work, with engineers replicating a design document.
So get interested in the needs that will never go away. I consulted once in the ready-mix concrete industry: surprisingly deep and intriguing. Another time it was a manufacturer of remote-sensing equipment for airports. Another time it was paper mills. None of those things are going away. Build a reputation as a troubleshooter.
> None of those things are going away. Build a reputation as a troubleshooter.
Not going away, sure, but are they getting automated?
Even troubleshooting.
This is where I'm at, right now: I could learn whatever, but how do I rate the various options' chances of being automated? Not "what do I think the chances are?", but how do I even sensibly approach the question?
I agree everything is kind of crap nowadays, but I also think that by 2016 we were already well on our way to the current situation. Adtech, social media, the iPhone (and subsequent smartphones), all were major contributors to the decline.
I wouldn’t call what we have now “dead” (though you seem to be using a loose definition). Exploitative, boring, harmful, sure. But it’s very much alive and sucking the life out of us.
Indeed. A lot of the stuff that's come out in the last decade has felt like people reinventing existing wheels, badly, because they didn't understand what the existing wheels were for, how they worked, or why they were there.
For example (while I don't know more than the basics of React on JS/HTML), SwiftUI is Apple's way to be reactive on iOS, but it comes with lots of "magic" that doesn't work right, missing bits, and feels like it only exists at all because enough people in the right places liked the right buzzwords.
While there were issues with version control on the old Interface Builder files (all that big blob of mystery XML), I've never seen a codebase or development process which was actually made easier (neither to develop nor to debug) with SwiftUI. If it had been up to me, the change there would be to have Apple buy up Figma and have it export UI directly: at least there, there (used to be, pre-AI) an obvious needlessly expensive duplication of work, with engineers replicating a design document.