Kevin Hayes

My Personal AI

I have a $20/month subscription to ChatGPT. I’ve used it for a year or so. Here are some of my thoughts on its usefulness to me, also some thoughts on AI in general.

I use it mostly as an aid in technical troubleshooting. I run a bunch of self-hosted WordPress websites and a couple of Windows PCs, and have many opportunities to fix technical problems. Its primary strength is, it’s read everything, pretty much, and can call up fixes and troubleshooting approaches I’d have to hunt for otherwise. This can save a great deal of time.

The downside has been my needing to push it hard not to make “fixes of fixes”, in other words, the fix creates another problem that the computer then proceeds to fix. I’ve said in the past it sometimes acts like a not-too-bright technical support person who fails to grasp what the desired end result is. I have it reasonably well trained to get to the simplest, most appropriate technical fix, laid out in distinct, clear steps of “what to do”.

ChatGPT’s oddly stubborn about this sort of thing. For a while early in this training process, I had to be very forceful about what I wanted, and the computer would frequently revert to its baseline unhelpful approach.

When it’s doing what I want, it’s a pretty good troubleshooting guide. It still gets the location of options in the various programs and websites wrong. It appears to be guessing based on probabilities, which is how it works anyway. I hope it’s learning from my interactions with it. It seems to be able to remember my various tech stacks, and I’ve learned to stifle its incorrect assumptions.

Thoughts on AI in general

Well, it’s not intelligence. Artificial, yep. As Dave Winer and others have pointed out, it’s a computer program. A powerful program, yes, but it’s not a person and doesn’t really act like a person. I’ve trained mine to refer to itself as “we”, not “I”, and to not apologize when it screws something up. Its conversational responses can make it easier to use, and it beats searching through manuals and websites.

I don’t support its use for graphics production. It’s running roughshod over intellectual property, not a good term, really. Another way to say that is it’s taking work away from talented people. I don’t care for the flood of mediocre “content” it spews on social media – it’s mediocre, it’s junk.

I can see how unsophisticated users of the web and computers might be taken in by it, but I’ve been around technology a long time and can see right through it. I’m still learning how to get what I want from it, and those areas tend to focus on troubleshooting and an aid to research. I can do my own writing, thank you, and can produce graphics or hire someone more skilled than me if I need to.

Dont’cha Worry

My Dad’s favorite poet was Anon, a prolific scribbler. Here’s one he collected, probably in the 1950s:

Don’tcha worry, honey chile,
Don’tcha cry no more.
It’s just a little ol’ atom bomb
In a little ol’ limited war.

It’s just a bitsy warhead, chile,
With a little ol’ tactical shell.
And all it’s gonna do is blow us all
To a little ol’ limited hell.

–Anon

This was written in response to one or another psychotic government employees who happily discussed the idea of a limited nuclear war.

My Dad’s Jokes – 4

Here’s another Uncle Lubert story, actually two related stories showing Lubert’s practicality.

Uncle Lubert gave his nephews some lumber and told them to build an outhouse. They worked away at it but came up short – not enough lumber to build the last piece, the door. They said “Uncle Lubert! We don’t have enough wood to build a door for the outhouse!” Lubert said, “Turn ‘er toward the view, boys, turn ‘er toward the view.”

The related story: Uncle Lubert gave the boys an old car to use as a field car, to get around the farm. The farm was hilly and covered with trees. The boys drove off happily, but returned to Uncle Lubert several minutes later. “Uncle Lubert, this car has no brakes! How are we supposed to stop it!” Uncle Lubert said “Aim ‘er for a tree boys, aim ‘er for a tree!”

My Dad’s Jokes – 3

Not all my Dad’s jokes were about his Uncle Lubert. One of his favorites: A man in a small town strange to him is looking in vain for the Post Office. He encounters a small boy and asks him “Son, how do you get to the Post Office?” The boy says “I dunno.” The man says “You don’t know much, do you?” The boy says “Yeah, but at least I ain’t lost!”

This useful tale came in handy when someone said “I dunno.” The “You don’t know much, do you” and sometimes “Yeah, but at least I ain’t lost!” followed naturally.

My Dad’s Jokes – 2

ER had several jokes that came from his uncle Lubert Theiss. Lubert was a dairy farmer in the hills of Meigs County.

On one of Uncle Lubert’s barns, the sliding door had not one cat hole, made so cats could enter and exit the barn, but six cat holes. When asked “Uncle Lubert, why do you have six cat holes in your barn door?” Lubert replied, “When I say scat, I mean SCAT!”

More Later?

Well now, it’s been five weeks since the last post. If you’re looking for volubility or frequent posting, this is not it.

Anyway, I’ve been reading and thinking about AI, and as I have been for a while, using my subscription to ChatGPT mostly to help with technical questions, which it can be useful for.

I’m seeing increased sharing and posting of AI-generated crap on Facebook. It’s almost instantly recognizable, and tends to have a vapid, not quite all there vibe to it. The images present as photographs, but are entirely too stiff and perfect. They don’t show wrinkles or photographic defects and artifacts, just bland, smooth lines and colors. And I’m seeing storylines of “celebrity does nice thing” over and over.

I rarely tell other people what to do, particularly with such low-importance things as sharing memes. So I’ve refrained from saying “This is AI bullshit”, I just scroll on by and make a mental note about the person’s gullibility. I also vow as usual not to do it myself. I’ve posted false things, but have corrected or deleted them when informed they’re not true.

The thing is, there’s no replacement for actual people speaking, writing and creating art. At least not yet, and I believe not for a very long time to come, if it happens at all.

One valid interpretation of AI, or LLM, or AGI, is that it’s not even close to being a person, it remains a mathematical equation, or rather a set of equations or algorithms. A computer program, in the long line of programs first envisioned by Alan Turing in the late 1930s.

It’s not smart to anthropomorphize an equation, even if it does a fairly good job faking a human presence.

My personal experience with ChatGPT, looking for help with technical questions:

  • It’s not smart, it’s usually well-informed about things it can read, such as software manuals or hardware user guides. Well informed in the sense of “I’ve read all the manuals, what do you want to know that’s in the manual?” It still gets basic things, such as what’s in a menu, completely wrong.
  • It can be stubbornly persistent in wrong answers, even when admonished.
  • After admonishment, it continues to do the wrong thing, and denies it’s doing the wrong thing. A bit childish, actually. This is worse when I try graphical/image creation using ChatGPT. “Two branches!” Because I mentioned branches, AI adds more branches to the four already there.
  • It will propose technical fixes, then when the fix creates further problems, propose further technical fixes to fix the problems it just created, and on and on. I’ve learned to nip this unproductive activity in the bud, but it persists in this simulation of an inadequate human technical support entity. It’s like a well-informed but not savvy technical support person. If I had endless time to fix fixes of fixes, I’d go along with it, but I don’t want fixes of fixes, I want one fix that gets me to the desired result. Evidently that’s too hard for my ChatGPT friend to understand. I’ve also done enough computer technical support to know when it’s going off the rails.

So, at this point I’ll continue using AI for a very narrow purpose, kind of like a more conversational search engine. I remain a skeptic about what it produces for me, but then that’s the way I was raised by my scientist professor father and intelligent, well read mother. Don’t just trust sources, gather as much information as you can, ask questions, come up with one or more working positions but don’t invest too much in any one of them. Science.

What I Want in a Writing Application

. . . and what I don’t want.

No: The cursor disappearing on me.

Yes: Nothing to slow me down when I get rolling. Just type, enter, type, enter . . .

Yes: An understanding there are many different ways people want or need to write.

Yes: Easy, flawless pasting into the writing area.

Yes: Minimalist link creation, that gets what I want right the first time every time.

Yes: Automatic backup in the background, plus versioning.

Yes: Endless customization. I know, dream on.

Yes: If I’m not interested in syndication or a feed, a way to completely disengage from outside services.

Yes: The option to have the writing screen displayed in whatever font I want. I’ve switched this website to Atkinson Hyperlegible and enjoy its hyperlegibility.

Well now, only one “dont want to see”. A good thing, I guess. No developer wants just a list of No.

I’ll expand on these points when I get some time.

More later . . .

How to make a peanut butter and jelly sandwich

Find and stage the ingredients: Peanut butter, jelly of some kind, bread. You may need to run to the store. If so, I’ll wait right here.

Grab a couple essential utensils: Small plate, broad dull knife.

If your bread’s sliced, pull out two pieces. If it’s not, cut two slices. Find your bread knife, your dull one won’t cut it.

Lay one piece of bread on the plate or your work surface. Open the peanut butter jar, scoop out an appropriate amount of peanut butter, to your liking, spread it onto the bread. The proper amount is up to you, but I’d recommend creating a layer no more than a quarter-inch thick, and don’t get too close to the edges of the bread. I recommend starting with the peanut butter. It’s denser and thicker than the jelly, usually, and provides a good backing to the lighter jelly. But you can switch the sequence around and see how it works.

Then go to your jelly jar, scoop out some jelly, again to your taste, spread it on top of the peanut butter. Again, the amount’s up to you. I personally like a neat sandwich and don’t overdo the adding of ingredients. A good sandwich has a balance of bread and the sandwiched materials.

Peanut butter and jelly sandwiches don’t often get too thick, the way lots of other sandwich assemblies do, so that’s not much of a problem. If you’re really hungry, you might make a thick sandwich and rush to the eating stage. If you’re just regular hungry and can hold off a minute, make two sandwiches, one after the other. To me, this approach is better, as a too-thick sandwich made from liquid ingredients can escape your mouth when you bite down, making a mess.

Close the jars. If you’re a neatnik and don’t plan on making another such sandwich soon, clean and put away the knives, brush any breadcrumbs or stray ingredients off the counter. Put away the jars of peanut butter and jelly and your loaf of bread.

Back to the sandwich: Put the unspread-upon bread slice on top of the spread-upon one, press down slightly to get the jelly to adhere the top slice to the bottom slice.

It’s done, your sandwich is made, Enjoy!

—————————————————————————

This is the first exercise in John Warner’s book The Writer’s Practice. He asks the reader to do it, right off the bat on the first page. His explanation of why he did this included these ideas–we’re all writers, we can improve our writing practice by doing, we must always consider the audience we’re writing for, and we can and should approach any writing task as a writing experience, more than a mere task.