Source: Scott Meyer’s Basic Instructions comic.

As it happens, in my senior year of college, I took a year-long coordinated studies program called Computability and Cognition. We studied the mathematics, logic, computer theory, cognitive science, philosophy, and ethics behind artificial intelligence, including what were then very early models of modern Large Language Models.  Mind you, the average smart thermostat today has more processing power than my school’s whole computer department had back then.  But the principles, *and problems* of AI were substantially in place even then.

So does that mean I completely embrace AI?  Nah.  Run screaming from the room?  Yeah, not that either.  Instead I find that it’s a useful tool for things I actually trust it to do.

In addition to tasks outlined in Scott Meyer’s comic, above, AI helps me outline, research, catch typos, and find better words & phrases. It’s very useful for summarizing Zoom calls with clients.  It’s very handy for tedious jobs like grinding through massive error logs when I adopt a new client’s website, as well as summarizing the hundreds of (mostly redundant) status messages I get.  It’s also convenient for analyzing and debugging code in client websites that other people built.  But in general, I like to do things the old-fashioned way, by hand, unless or until I thoroughly trust my ability to automate it.  (In other words, I use AI pretty much the same way I’ve always used non-AI.)

Bottom line: whether or not I use AI, I take full personal responsibility for any code I may add to a client’s website as well as any content I publish here or elsewhere (the good, the bad, and the em dashed.)