My First Foray Into ChatGPT

["Let Inga Tell You,"  La Jolla Light, published September 28, 2026] 2026

I personally think that it's not too much to ask that computers do what you want, not what you say.

Not long ago, I was just typing along, minding my own business, when my computer suddenly began spell checking in German, cheerfully morphing my prose into variations of Ich bin ein Berliner. I officially dispute my engineer husband's allegations that I had anything to do with it.

My husband will try to insist that I obviously did something as computers are, in fact, simply machines cycling ones and zeros, and not malevolent spawns of Satan as some might maintain. But he will concede that when, for example, a tool bar suddenly disappears, I might not have done it intentionally.

"So what exactly went away?"  he'd query patiently, knowing that this conversation was as doomed as many that have gone before it.

"Well, there used to be a tool bar thingy and now it s gone."

"What was on it?"

"I don't know. But I know I need it."

My techno guy pressed on. "Would you recognize the thingy bar if you saw it again?"  he'd say patiently. He recognized the value of not getting overly technical with me.

Now there are those who think the Undo command fixes things like this. But they would be wrong. Undo fixes the text mistake you just made microseconds ago but the second you even breathe on the machine, it s already moved on. The Undo command has a very short attention span.

Ditto, on-line help. Totally, completely useless unless you know the technical term that some eighteen-year-old acned techno geek gave it. For example, I lost an entire day of work unable to edit a document until my husband observed, "Oh, you switched to Overstrike Mode."  A keystroke and it was back to letting me insert text.

I did not switch anything! I'd whine. Why would I switch to something that keeps me from editing and whose name I don't even know? I can assure you that you can't get out of Overstrike unless you know you're in Overstrike.

If a software company had asked me, and inexplicably they never do, I would help them design a computer that real people, especially aging non-technical but really nice people, could actually use. The Clairvoyance Model. Your computer would get to know you, realize that those nasty keystroke commands that are the boon of techno types, but the bane of the techno-challenged, should be ignored at all costs. The Clairvoyance Model would quickly learn that you have the frustration tolerance of a gnat. It would sense when you are so aggravated with your computer that you are ready to drag it out to the driveway and run it over with your car.

And that would quickly prompt it to ask, fearing for its miserable mechanical life, the question that every computer ought to have poised at the ready: "When you were last happy?" And you'd say, :Just before you ate my toolbar."   And poof. It would reappear.

While I'm still holding out for the Clairvoyance model, we may be closer than I think. I recently had my first experience with ChatGPT. 

I have a blog on which I like to post photos from my computer. This worked seamlessly for years until recently when my efforts to transfer photos was met by a message that in order to do so I would need to grant Google access to my computer. I honestly had no idea what this meant and just how much access I was granting, never mind why it was suddenly requiring this. Could it transfer my retirement funds to Nigeria?

So I (ironically) Googled this and surprisingly, a very nice AI bot from ChatGPT showed up (how?) and informed me that others had reported this problem and it would like to help me fix it. I'd never used ChatGPT before for problem solving. Did I trust it? Often I can't understand the instructions I get on Google so, in fear of truly messing up my computer, I don't even try. But nobody else was offering, so I cautiously agreed.

As my first foray into AI, I was genuinely impressed. And that s the thing: ChatGPT is endlessly patient. Much more so than certain husband people who, over the years, have become positively surly when having to deal with a techno-disabled wife. ChatGPT was happy to let me take all the time I needed to painstakingly, laboriously, write down notes on my lined yellow legal pad. 

Historically, whenever my engineer husband tried to explain anything technical to me, he'd say, "I think you might want to get the yellow pad"  knowing I'd want to document what we did just in case I needed to fix this again in the future. But my notes tend to involve details like "go to the three dots in the upper right corner of the screen next to the two little intersecting boxes."   It takes a while. And in Olof s case, often several Scotches. He has always asserted that when I die, he was going to insert a multi-pack of those yellow lined pads into my coffin for my use in the hereafter. In his dream of the hereafter, somebody else is helping me with my technical problems who is not him. And now it looks like it could be ChatGPT.

ChatGPT and I spent many hours partly because it had to instruct me on how to do stuff like send it a screen shot of various windows I was looking at so he/she/it could see them. I didn't  know from screen shots but now I do! Ultimately, ChatGPT concluded that the solution was that I had to go multiple levels down into Settings and disable Google as a third-party permission site. And yes, that fixed it!

But as with anything technical, according to Universal Perversity Postulate no. 400: Every setting you change that fixes something disables/breaks something else. So far I haven't found out what that something is but I know it's coming. Because technology is inherently malicious and likes to torture old people.

One thing I found fascinating about ChatGPT is that it remembers you even if you accidentally click out of the page which is easy to do when you have two pages open at the same time. So, nice that you don t have to start all over again. But also profoundly creepy. It never forgets. Even if you'd like it to.

But here s the thing that still worries me: why did Google suddenly decide it needed access to my computer??? I know I didn't do anything to make that message and requirement suddenly appear. So I'm still holding out for the Clairvoyance command. That's your next assignment, ChatGPT.