Let Me Tell You a Tale of AutoCAD
A story about technological change
Hey y’all, it’s KimBoo! I’m an author and a podcaster who is also a librarian, text technology historian, and former I.T. project manager. I write about a lot of interesting things, I hope you agree! Please consider supporting me (and my dog!) so I can keep throwing errata & etcetera into the Scriptorium!
NOTE: This article is not particularly pro- or anti-artificial intelligence technology, but is simply looking at what is happening and how the history of technological change can inform us on what is going to happen next.
There are many pearl-clutchers out there railing on about how A.I. is not a human being and cannot replace the unique creative spark intrinsic to human beings. Putting aside discussions about the singularity or AGI or sentience in general, I agree that at the current stage of A.I. development, the peal-clutchers are right.
On the other hand, it doesn’t matter if they are right.
Gather ‘round and let me tell you the tale of my first brush with AutoCAD in 1995.
After my mother died in 1994, I found myself contemplating what I was doing with my future. I was 24 y.o. and living in a brief, golden and ultimately tragic moment of financial security. The economy was on the upswing, grunge music was all over the radio, the Internet had suddenly appeared, and my father was still alive. I was living at home with him and we easily managed to live off his military retirement despite Mother’s medical bills forcing us into bankruptcy.
I planned, like the naïve child I was, that I might have ten years or so to live with him in relative safety in order to “reboot” my life after bailing on grad school (on my whole life, really) so that I could take care of my dying mother.
A very grim introduction to a post about AutoCAD, but bear with me.
I decided to pursue my dream of becoming an architect, which I was woefully unprepared to do. To make up for my lack of anything resembling architecture, engineering, or art in my undergrad I went to the local community college and enrolled in their “construction and design” vocational program.
I learned a lot about concrete, but one of the course tracks was the more esoteric craft of drafting. I fell in love with it and bought my own drafting table to work on projects at home. I loved the mechanical pencils, the art pens, the vellum, the geometry, the lines… I spent hours in meditative contemplation of line width and perspective and shapes. I felt like I had found something I could do, and do well, and that serve my plans. Draftspeople were highly paid and I figured I could pay my way through an architecture program doing it.
The AutoCAD (“Auto Computer Aided Drafting”) lab was newly redone, expanded with a lot of “terminals” for students to work on. Taking one class in using AutoCAD was required but no one really wanted to do it since it was a maddeningly difficult program to learn and, at the time, most CAD software was only used by very large architecture shops and corporate construction firms anyway.
As I sat in that room with thirty or so of my fellow students in mid-1995, I looked around and realized my dreams had just been dashed against the rocks. This, I realized, was the future. Only a small fraction of us would ever be needed as draftspeople, and those jobs would likely go to people who were already experienced at it and AutoCAD. This software was going to decimate an industry.
I talked to my drafting instructor about it, a local architect himself who was very supportive of my goals, and he shook his head. “They will always need real draftsmen to draw up blueprints. AutoCAD will be helpful, but it won’t replace people.”
In the end, we were both right.
Gone are the days where every general contractor needs one or two draftspeople in-house. Gone are the days of large rooms filled with dozens of people working on drafting tables at the big architectural firms. Gone are the days I could have worked at a small firm and paid my way through school drawing elegant lines with a pencil.
Draftspeople remain, though, because they are still needed. However, businesses need one or two where they used to need ten or twenty.
AutoCAD did not replace people entirely. It just replaced a lot of them.
That is where we are headed with current developments in A.I.
The discussions surrounding this topic is very, ahhhhh… shall we say, hyperbolic? But like the discussion I had with that instructor back in 1995, the current reality of A.I. is somewhere in the middle.
AutoCAD is not an architect, nor is it a draftsperson. It has been around since the early 1980s but when the technology finally made it more widely available, even without a fancy-pants LLM attached it managed to completely reshape a long-standing and respected profession in just a few years.
The point I keep making to those pearl-clutchers, based on my own historical experience with AutoCAD, is that to replace many (many) white collar jobs, A.I. doesn't need to be better than the best human. It does not require a unique spark of sentience or creativity, it just needs to be as good as a mediocre desk jockey. Even then, it will be better because it is faster.
A.I. tech as it stands now is not a lawyer or a plumber, and none of the models in current use are a sentient being (yes, I’ve heard the rumors, and I’m not buying it…yet…). There is no single A.I. LLM that can write a whole novel without a lot of human intervention and ingenuity to do the real work.
But that is where A.I. tech stands now, and it is definitely not standing still.