In 1994, I was told by a visionary professor of Computer Science that I was a fool for going into CS because the combination of component software design and offshoring was going to eliminate jobs.
I remember being pale in the face and sticking with it. At the time I graduated, there were 13 CS graduates, of whom two had cross-disciplinary fields. That class had the guy who invented Hadoop, and the folks who invented dtrace, and me (yes, I am putting myself in the same breadth, but that’s because we graduated at the same time).
Thirty-one years later, I see the same kind of fear-mongering.
The notion that computers will do software engineering or that there is a finite demand for engineered products remains the dumbest and most ignorant take in the history of takes.
AI is just the latest iteration in making each unit of software we write more efficient. In the 1980s, it was the move from assembly. In the 1990s, it was the move to garbage-collected programming languages. In the 2000s, the emergence of databases, hypervisors, and the web occurred. In 2008, it was the emergence of public cloud.
Does that mean that there aren’t dislocations and changes? No. In fact, in those transformations, jobs stopped existing, and folks had to retrain. And some of it was unfun.
But the idea that tool-making, design, and construction don’t require human beings is the fevered dreams of AI advocates.
Leave a Reply