Nothing like a clean batch of underwear. I can go a while as my family will attest. Apart from the fact they’re laying off tens of thousands of workers and trading in human capital for machines (and utterly evil), Amazon’s order history engine is so rich it can provide shoppers purchase detail from years back. That’s how I learned the last time I bought new underwear was May 5, 2020. I’d bought some in Germany but got the sizing wrong and there’s nothing worse than too-small briefs. Worse yet, the slim, Euro fit. I’d gotten fooled by the model on the cover that I could look like that but of course, I can’t. Those got put in the trash.
Sinking through the cracks of the day into a deep sleep. Thinking about things like anti-fungal scalp foam and pimple cream, the people I love and where in the world they are, the dog licking herself. This banal compilation of life. How to get rid of fruit flies. Work.
Thinking does you no good in bed and I’m pretty good at not thinking. So I slip between the cracks and dream. The dreams are less about thinking and more about letting go of your thoughts. Decanting or off-gassing. A younger version of mom crying, borrowed from a photo. It is the best picture of mom and John: the two of them unpacking, moving into that apartment in Collioure. John pointing at an empty bottle of Pastis, that south-of-France staple liqueur that turns yellow-green when you add water. Mom with her hair tied back smiling, starting a new life. We have a copy of the original photo and they used the same picture for the back of John’s book, Carousel.
I shake off dreams like that or they just evaporate. Each day follows the same pattern, beginning with the pile of clothes by the bed I put on in reverse order, starting with the underwear and socks: the last to come off are the first to go back on.
I check for work activity and if it’s there I note it and come back to it so it doesn’t dampen the most sacred part of my day. But I’m a tweaker, I don’t like surprises, so I am constantly checking for activity. Sometimes in the early morning I’ll respond to work activity to signal the fact I’m online at like 0500h but mostly it’s banal thumbs-up reactions to chats or accepting meeting invitations. Yesterday I tried doing something more cerebral, updating a video script, which I used an LLM to help me with because frankly I was feeling lazy and knew the LLM could do as well as me. But then I noticed hours later there was a problem with some phrasing and I had to redo it. The problem wasn’t the LLM’s—it just hadn’t caught an issue passed down through a reviewer who’d rewritten something I had. I’d taken the language from an approved messaging document and she tried to make it sound more conversational but botched some phrasing by mashing up two ideas: embedded AI and business AI. She made it embedded business AI, which is technically two different things. This is my work.
Our team has hired researchers to interview business leaders about how they’re using AI and last week I watched a guy say with glee how he’s going to fire a staff of writers who refuse to use the company LLM and insist they can do better. He insists they can’t, not better enough to justify their salaries. The cost-benefit analysis comes down to that, the enough. Does the cost of the AI deliver enough savings to pay for itself? Can it help people get more done? Since late 2022, when ChatGPT was released, this has been my mode: prove you can get more done than the machines. No one will ever win that game but I will spend the rest of my days trying.
I can’t get my anti-fungal foam because it’s on back order due to something our president probably did. Our local radio station lost all their funding and keep reminding us they’re not algorithms, they’re human-curated, please show us your support. We’ve been giving them the same amount we give Spotify every month, for years. We still get the Sunday newspaper too but just use that for fire starter.
Categories: Creative Nonfiction, Errata, Technology

I’ve been thinking a lot lately about AI, and how all the expectations of it ending so many jobs are based on the fact that it’s near free. However, AI is NOT earning money. It won’t stay free. Billionaire investors’ investments have been paying the operational costs. I am slowly diving deeper into this, but I expect that when OpenAI, etc, stop receiving enough investor dollars to cover operating costs, that we might find that humans are quite cost effective.
LikeLiked by 1 person
So many layers of illusion and fog. Looks like OpenAI may be converting to for-profit now too? You should check out that guy Stephen Klein on LinkedIn if you haven’t already; I like how he calls it like he sees it.
LikeLiked by 1 person
I’ll track him down. Thanks for the recommendation, Bill.
LikeLiked by 1 person
He’s the Stanford Stats professor.
LikeLike
This chap, correct? https://www.linkedin.com/in/stephenbklein/
LikeLike
Yup.
LikeLike
Oh my … when we humans become fully second-class to AI … hope I’m old enough to not live that long … I treasure my identity as pre-AI!
Stirring post. Jazz
LikeLiked by 1 person
It’s also self-imposed in some cases innit? I think there will be a correction, I hope…
LikeLike
You can’t return or donate undies, no matter how new and attractive, once they are out of the box.
A bit like AI.
~
Bait, hook, reel in, gut and monetise.
It looks like the gutting stage might be about to start in earnest.
LikeLiked by 1 person
Yeah scoop our heads out like pumpkins, re: the gutting stage. All that glows is the reflection of the screens on our faces
LikeLiked by 1 person
lamentable, perhaps
inevitable, probably
supply chasing demand
LikeLiked by 1 person
Not to mention the tech is pretty dope as we say! Effing amazing, real-time translation etc.
LikeLiked by 1 person
I’m just in the middle of reading about an AI training avatar for teaching about dementia care. It produces better comprehension and retention than the standard training approach.
PDG
LikeLiked by 1 person