Artificial Smarts

I had taken one computer programming class in college. I didn’t like it, and almost didn’t pass. They tried to teach me something called FORTRAN IV. I guess FORTRAN I through III were failures. In my opinion, so was IV. — ‘An Einganeer’s Tale’, page 37

They also realized they could do better with fewer employees and more automation. — ‘An Einganeer’s Tale’, page 67

Those first engineers who developed things like the wheel, the pyramids, the catapult, and the guillotine, had little more to help them with calculations other than their fingers and toes. Then came the abacus, the slide rule, the calculator, the computer, and now AI (Artificial Intelligence).

I became an engineer during the twilight of the Slide Rule Age. Throughout my college years, I proudly wore my slide rule holstered on my belt. I’d quickly draw it whenever I needed to calculate anything, and there was nothing I could not do with it. Lining up those slides and hairlines became second nature to me. I did not move onto advanced calculators until Texas Instruments came up with one I could understand. I just could not get my head around Hewett Packer’s concept of reversed Polish notation, no matter how hard I tried.

After college, my career pretty much coincided with the computer age. (Bill Gates dropped out of Harvard the year I graduated.) I worked through all its transformations: main-frames to smartphones, floppy disks to thumb drives, punch cards to the internet — finding that each advancement allowed me to do more with what I knew, and faster. Rather than becoming frustrated by all the changes (well, there was some of that), each advancement allowed me to be more creative, which lead to greater overall productivity and technology development. So, now we’re on the dawn of the next age, AI. Now what?

In case you didn’t know, the Deep Dive podcast on www.einganeer.com is AI generated. This is my first usage of AI tools for anything, and I am amazed by what they can do. Jake and Cary, the podcast’s hosts (whose names I borrowed from my niece and nephew) make me sound so much smarter than I really am. I can see that some day, AI tools could do a much better job with things like planning and scheduling a refinery’s operation, much better than the technology I helped developed over my entire career.

As automation has replaced so many workers, AI threatens to replace programmers and engineers. Tasks which are accomplished by following a defined set of rules, procedures, methods, or conditions are prime for AI solutions. Unfortunately, most programming and engineering tasks are exactly these. Furthermore, AI can potentially accomplish these tasks faster and with no errors, as well as determine if better solutions might be found by considering other factors. I understand that AI is already displacing some programmers by writing code better and faster than they can.

So what are engineers to do? Well, the same we have done when faced with other advancements in artificial smarts — CREATE! As powerful as AI is, it is limited by what is known, what has already been thought of in some form or fashion. We should continue to treat AI as only a tool, but a tool we can use to accomplish things we can now only dream of.

Moving to and learning to trust a new form of artificial smarts has always been a difficult and daunting undertaking. But it has also always been a rewarding and never regretted one. A colleague of mine framed his slide rule and hung it on his wall beneath a sign that read IN CASE OF EMERGENCY BREAK GLASS. The glass has never been broken.

Fun fact: The first humanoid on Mars will have an AI brain (Tesla’s Optimus robot, before the end of 2026, according to Elon Musk).