Predict the Future of Your Industry

Honestly? I think my industry (software development) is heading for a cliff. A complexity cliff. We're already struggling to build abstractions well enough to keep our heads from exploding with the complexity we all seem to love. Without a drastic change in direction away from OO, leaky abstractions (DAO and ORM come to mind), and a rejection of incidental complexity, we're going to collapse under the weight of our own complexity.

I know, the prediction seems pretty dire, but I'm seeing it in both new developers and experienced developers all the time, at multiple companies, and in multiple languages. What we've built and are now building ARE TOO COMPLEX. Not too complex for a computer to parse, but too complex for our (feeble in comparison) minds to be able to comprehend. We need a return to the fundamentals of computing: theoretical math, lambda calculus, pure logic. Maybe it's just me, but when I'm working on a problem where I'm comparing the outputs of two programs and can't satisfy the compiler, let alone get actual matches, we've got a huge problem. I'm no computer genius, but I've been doing software development for at least 10 years now.

I can't shake the feeling that as an industry, we've seriously fucked up, and need to go back and understand what and where the hell we went wrong and fix it, and soon.