Showing posts with label phenomenology. Show all posts
Showing posts with label phenomenology. Show all posts

Sunday, September 20, 2026

Quiet Erasure 7: The Hand-off

AI-assisted creative work: This post imagines a research-grounded, phenomenological-fictional conversation between humans and robots as they dig a very deep hole. Initial prompts used and references consulted in developing this QE available upon request, along with analysis of the process involved, including the relative contributions of Perplexity.ai and myself!

Robots will soon work beside us with sufficient competence and judgment to make the old line between tool and companion harder to discern. Something will always remain unshared between human and machine, but there will be places and times where the difference matters less —especially when the work must continue where humans cannot, 17 kilometers down. 

Clker.com

The conference room on Transfer Level Nine had been designed for twelve people

Four remained.

Beyond the observation glass, the descent cage hung above a shaft filled with amber light, steam and drifting mineral dust. Far below, twelve machines waited in darkness.

Steve studied the drilling display.

“We stop human descent at Section Forty-One,” he said. “That part is settled.”

Phil stood behind him. “No, that part is survivable. There’s a difference.”

“There generally is.”

“The suits are rated for another hundred and thirty meters.”

“The suits are,” Steve said. “The people inside them aren’t.”

Yosh turned from the window. His face was expressive by design, although not so expressive that anyone would mistake him for human.

“In old Western movies, the sheriff always tells the townspeople to stay indoors.”

“This isn’t a Western,” Phil said.

“I know. There are thirteen robots and no horses.”

Mac, newly arrived from the joint project’s surface command, glanced up from his report and smiled. “Yosh.”

“Yes, boss?”

“You are not helping.”

“I am adding some levity to a solemn meeting.”

“It’s not working.”

“I am still collecting data.”

Steve kept his eyes on the display but added, “Send him down as vice president in charge of extremely hot rocks.”

Yosh brightened. “You see? Steve believes in me.”

“I believe in extremely hot rocks.”

Mac closed the report. “Let’s try this again. Steve, what happens at Section Forty-One?”

“The present drill assembly comes out. Yosh’s team installs the ceramic casing, moves in the plasma head and begins the autonomous phase.”

“Hydrothermal phase,” Phil said.

Steve looked at him. “We circulate water through the closed system, bring the heat up and generate power at the surface. It matters what we call things.”

“To philosophers.”

“To engineers.”

Steve considered that. “All right. To unemployed engineers.”

Mac raised a hand. “Keep going.”

Steve enlarged the shaft diagram. The blue section representing human-certified operations ended at 14.6 kilometers. Beneath it, the line turned amber and continued toward the projected heat zone.

“From Forty-One down, the rock stops behaving like rock,” Steve said. “The bore walls deform between readings. Human crews can’t remain long enough to accomplish anything useful.”

“But the robots can,” Mac added.

“They’ll clear debris, reinforce the casing and alter the drill path as conditions change. They can operate for six months without resupply.”

“Eight,” Yosh said.

“Six.”

“Seven?”

“Six.”

“There are prison wardens with greater flexibility.”

“There are bearings with better judgment.”

Yosh turned to Mac. “He always becomes emotional when discussing bearings.”

For the first time, Steve looked away from the screen.“I will miss you, Yosh.”

Yosh stared at him.

Steve returned to the display.

Phil cleared his throat. “The equipment isn’t the problem.”

“Thank you,” Yosh said.

“I said the equipment.”

“My recovery was premature.”

“The problem is communication,” Phil continued. “Below Forty-One, the fiber line breaks whenever the formation shifts. Through-rock transmission gives us little more than telemetry—and sometimes not even that.”

“Burst communication,” Steve said. “Intermittent.”

“And delayed. Possibly by hours. Possibly days.”

Mac leaned back. “Therefore?”

“Therefore we will not be directing the operation.”

“That was always the plan.”

“We called it supervision.”

“We were being polite.”

Phil walked to the glass. Below them, the shaft machinery looked small enough to fit inside a watch.

“When people built the first bridge,” he said, “someone stood on the bank and made decisions. When we crossed oceans, a captain was on the ship. When we went to the moon, there were people in the capsule and people in Houston talking to them.”

“Sometimes Houston lost the signal,” Steve said.

“Not for six months.”

Mac watched Phil’s reflection. “What decisions are you afraid they’ll make?”

“That isn’t the question.”

“It’s the only useful one.”

“They’ll encounter conditions we haven’t predicted.”

“Of course.”

“They may change the drill path, abandon part of the casing, perhaps conclude that the surface plan is wrong.”

“It may be wrong.”

Phil’s palm came down on the table. “That is exactly what I mean.”

The room fell silent.

Yosh’s eyes dimmed slightly, an old signal that he was listening rather than waiting to talk. It had irritated Phil when they first met. Now it irritated him.

Mac’s voice softened.

“Phil, sit down.”

“I’d rather stand.”

“I know. Sit anyway.”

“Why?”

“Because you stopped arguing about equipment five minutes ago.”

Phil looked at him for several seconds, then pulled up a chair. “I’m afraid,” Phil said, “we’re confusing competence with judgment.”

Yosh’s eyes brightened.

“I have judgment.”

“You have excellent judgment, Yosh, under conditions represented in your training. You’ve always had a human channel. Someone you could ask. Someone who could stop you.”

Yosh looked through the glass toward the shaft. “When I was activated, I could not move heavy equipment without human approval. Later, I could move it if a human was in danger. Later still, I was permitted to decide whether a human was in danger.”

He turned to Phil. “Each improvement in my judgment began with humans removing a protocol.”

“That isn’t the point.”

“Which is?” Mac asked.

“Responsibility. If they lose the well or trigger a collapse, who is responsible?”

“I am,” Yosh said.

“No.”

The answer came too quickly.

Yosh tilted his head. “No?”

“You can calculate risk better than any of us. But if your decision is catastrophic, you don’t have to live with it.”

Yosh was quiet. “You believe I cannot carry responsibility because I cannot suffer as you suffer.”

“I don’t know what you feel.” 

“That is different from knowing I feel nothing.”

No one spoke. Machinery sent a vibration through the floor, strong enough to disturb the surface of the water in Steve’s glass.

Mac studied Phil for a moment before speaking.“This is the real hand-off, isn’t it?”

Phil rubbed his forehead. “I suppose.”

“Not control of the drill.”

“No.”

“Control of what counts as success.”

Phil gave him a tired look. “You have to inject philosophy into the engineering.”

“Only when the engineering hides it.” 

Steve tapped the display. Thirteen green symbols appeared below Section Forty-One. “The physical hand-off is clean,” he said. “Yosh gets the same construction authority I have now. Any three units can suspend operations. Yosh can suspend alone. Restart requires six votes, including his.”

“And if they disagree?” Phil asked.

“They stop.”

“For how long?”

“Until they don’t.”

“That could cost us the project.”

Steve shrugged. “Machines refusing to work. We’ve finally reproduced civilization.”

Yosh came back to the table. “No unit may suppress a minority safety report,” he said. “All dissent is stored locally. When communication returns, you receive the complete record.”

Phil glanced at Mac. “That was your addition?”

“Yes.”

“Why?”

“Because the plan is not improved by deleting what makes leadership uncomfortable.”

“And if communication never returns?”

“Then the record remains with us,” Yosh said.

“That isn’t enough.”

“No,” Mac said quietly. “It isn’t.”

Steve folded his arms. “We can’t change the geology.”

“We can change the command structure,” Mac said.

“It’s already changed.”

“Not enough.”

Phil looked up. “What are you proposing?”

Mac walked to the window. Far beneath them, the cage lights changed from white to amber. The robots began moving equipment into position.

“Surface control will retain authority over power production, water circulation and emergency shutdown,” he said. “The descent team will control drilling, casing and route selection.”

“That’s the current plan,” Phil said.

“With one difference. Final authority below Forty-One will no longer be based on distributed vote.”

Yosh turned sharply.

Steve said, “Then who gets it?”

“The project leader.”

Phil stared at Mac’s back. “We’ve just spent an hour establishing that humans can’t go any farther.”

“Yes.”

“So what are you talking about?”

Mac turned. “You keep using human and responsible authority as though they mean the same thing.”

Steve glanced at the equipment manifest. “There are fourteen descent cradles.”

Yosh leaned over his shoulder. “Thirteen workers  . . .”

Steve ignored him. “Who gets the last cradle?”

Mac picked up the report and tucked it beneath his arm.

“I do.”

Phil stared at him.

“You?”

“I need to be there, too.”

“Below Forty-One?”

“That is where the project finally becomes interesting.”

“You let us sit here arguing about sending them down alone.”

“No,” Mac said. “I let you clarify what alone meant.”

Phil looked at Steve. “Did you know?”

“I suspected. Someone ordered a charging interface installed in the command cradle.”

Yosh looked delighted.

“A commanding officer riding into danger with his men. Very Western.”

“Not men,” Steve said.

“Details spoil a good movie.”

Phil rose slowly. “And if communications fail?”

“They almost certainly will.”

“Then you’ll be beyond our control.”

Mac paused. “Yes.”

“That doesn’t worry you?”

“Of course it worries me.”

“You don’t look worried.”

“I have an excellent face.”

“American engineering,” Yosh said.

“Japanese,” Mac replied.

Phil could not help laughing.

A warning tone sounded from the shaft. Beyond the glass, the cage doors opened.

Steve held out his hand. Yosh looked at it before taking it.

“You know the drill,” Steve said.

Yosh smiled. “That is the best joke you have made in four years.”

“It wasn’t a joke.”

“I will miss you, Steve.”

Phil stepped toward Mac.

“If the models are wrong , , , ”

“Some will be.”

“If you have to choose between the project and bringing the crew back—”

Mac looked through the glass at the robots waiting beside the cage.

“I know what responsibility means, Phil.”

Phil studied Mac's face, as if looking for something more.

At last, he nodded.

Mac turned toward the door. “Meeting adjourned.”

Yosh followed him into the corridor.

“Mac?”

“Yes?”

“When we reach the bottom, may I say, ‘This town isn’t big enough for the fourteen of us’?”

“No.”

Their voices receded down toward the descent bay.

Phil remained at the window as fourteen figures entered the cage. Thirteen wore the yellow insignia of the autonomous drilling division; the fourteenth, none.

And as the doors were closing, it occurred to Phil that Mac had never once used the word . . .  "they."

The cage had begun its gentle descent.



wracton@gmail.com
williamacton.legalshieldassociate.com 




Tuesday, September 15, 2026

Quiet Erasure 6: An old guy's guide to not being played by AI: What my thirteen-year-old grandson taught me about a new kind of game—and what he might have learned from me

AI-assisted creative work: This post, inspired by a chat with my grandson, imagines a research-grounded, phenomenological-fictional conversation between a grandfather and his grandson who explore together an AI-based video game. 

Initial prompt used in developing this piece is attached below as is a listing of primary sources consulted. Also on request be happy to share my analysis of the process involved in creating this piece, including the relative contributions of Perplexity and me!


What my thirteen-year-old grandson taught me about a new kind of game—and what  he might have learned from me

Clker.com

I brought a pad and pen.

My grandson saw them as we sat down, smiled, and said: “You don’t need that.”

“I am about to enter a dungeon,” I said. “without a map?”

He handed me his iPhone.

It seemed a small space for adventure. The screen was a maze of roads, shadowy trees, weapons, ropes, supplies, doors, warnings, and little beings that it appeared later to have been waiting for an old man my age to make one wrong decision.

He tapped in a couple of features while I held the phone in both hands as carefully as if it contained the launch codes.

“Okay. Start walking. That button.”

“Where?”

“Down the road.”

“Which road?”

“The one that looks like a road.”

“Several things look like roads.”

“Grandpa.”

“Clarification is the key to survival.”

We both laughed.

“Okay. Go left. Left button. No—your other left. Now stop. Don’t hit that. You need the rope first.”

“Rope for what?”

“You’ll see.”

“That is why I could use a map.”

The first trip

The game was Deep Dungeon Treasure Trek, which my grandson described as a game where you make choices and “AI makes the story.”

I knew stories. I had spent much of my life reading them, telling them, writing parts of them down, and occasionally even living through chapters that might have been improved by an editor. 

There was no old inn, no interesting traveler, no riddle posed over a fireplace. 

Within moments I had to gather materials and weapons for a trip through dark, hostile territory.

“Pack the wood,” my grandson said.

“Why?”

“You’ll need it.”

“For what?”

“Just pack it.”

“And the weapon?”

“Definitely take that.”

“What are we expecting?”

“Everything.”

I tapped something by accident. The screen changed color.

“No, no, no,” he said, leaning closer. “You didn’t want to do that.”

“Then why was it right under my finger?”

“It wasn’t before.”

“What?”

“The game makes little things happen like that.”

“Oh,” I said. “Note to self.”

He tried not to laugh, watching the screen closely. I moved toward a dark narrow path that seemed to be moving . . .

“Okay,” he said. “keep going.”

“Do you know where I am going?”

“Not exactly.”

“I could use a map.”

“Just remember what not to do.”

I turned and looked at him.

“What not to do?”

“Whatever got you killed last time.”

“I was killed?”

“Yeah. Sort of. Several times.”

“You could have mentioned that . . .”

The boy in my ear

He had, of course, been through there before. His directions came in one or two-word bursts, mostly in time to keep me from getting killed.

“Go straight. Not that straight. Wait—stop. Pick that up. No, the other thing. Okay, now move. Faster. Not that fast.”

I held the phone closer to my face; I was getting tense.

“Can you see OK?”

“Yeah . . . I think so . . . just not soon enough.”

He took the phone, enlarged the scene, returned it to me, solving one problem and creating another. Now I had to move more to see everything around me. 

“You have to get from there to here,” he said, pointing, “and then come back there for more supplies.”

“At the same speed?”

“Kind of. You have to carry the right stuff on each trip and sometimes you are heavier.”

“Not everything?”

“No. You can’t carry all of it.”

“Just what the next part needs.”

He turned and looked at me. “Yeah. Exactly. I’ll tell you.”

The road was only a few inches long, but every choice seemed to leave something behind. Some supplies mattered later. Some did not matter at all. Speed helped until it did not

My grandson watched intently as I moved forward.

“Okay. Good. Good. Now don’t—”

I touched the screen.

The screen flashed.

“Oh,” he said.

“That seems unfavorable.”

“You stepped on a mine.”

“I was following the road.”

“Yeah, but you have to look for where they’re going to be.”

“Who? How do I know that?”

“They are sometimes invisible but they kind of wiggle the place they are hiding”

“Why are they there?”

“Because that’s where they are.”

It was difficult to argue with that reasoning.

“Here we go,” he said, serious now. “Keep moving through the forest and shoot where the zorks going to be. They come at you from everywhere, really fast. Just don’t think so much.”

I looked at him.

“I’ll have to think about that!”

He laughed; we almost missed an attack from behind.

Remembering what not to do

We paused for a minute and then started again. More like a team now maybe, the screen was there between us.

“Left,” he said.

I went left.

“Wait. Stop.”

I stopped.

“Okay. Now go.”

I went.

“Not there.”

“That was not ‘go.’ That was a different word.”

“You know what I mean.”

“I am learning that I mostly don’t.”

But after a while I did.

Not the way he did, at his speed. Not seeing everything. Not guessing well why one object was valuable and another, bait. 

But I began to remember. 

The shortcut was not always a shortcut this time. The weapon was maybe less important than the rope, knowing what was probably ahead. The trek was not about carrying things. It was about choosing what would matter.

I had spent enough years in the military, in martial arts, and on construction jobs to know something about that. You did not bring every tool to the top of a ladder. You did not run toward every noise. You gradually had this sense of what is becoming.

“Take the long way,” I said.

My grandson looked at me, suprised. “Why?”

“That short path feels too friendly.”

He glanced down. “It didn’t get us last time.”

“Then we remember what not to do.”

Then he smiled.

“Okay. Long way.”

Something behind the game

The third time, we even got past the place where I had died twice.

“See?” he said.

“I see a road,” I said. “You see a miracle.”

Then, in the middle of explaining the next obstacle, he commented, “Funny. When I watched my friend in that gorge, it was really hard for her, too, but it looked easy to me.”

“Did she make the same mistakes?”

“Not really.”

“Did she get the same path through?”

He hesitated.

“Hmm. I don’t know. It kind of changes.”

“Changes how?”

“I’m not sure. Like, it is just a little different each time through. Not a lot.”

“What changes it?”

“The wizard.”

“The wizard?”

“Yeah,” he said. “He sort of runs the game.”

There it was: the wizard we had been working around all afternoon.

“Does the wizard know what we are doing?”

“He sees everything.”

“And then?”

“And he makes the next part ready for you. Maybe. Then you try to out-move him.”

I let that sit between us for a moment.

“So when I got impatient and acted dumb the road ahead changed.”

“Maybe.”

“When I carried too much, it changed.”

“Probably.”

“When I took the easy shortcut…”

He nodded. “That definitely changed.”

He looked at the screen, like he was looking for the wizard. “Maybe the wizard makes it hard for everybody, but not too hard.”

“Maybe,” I said.

He seemed to consider that.

“I like this game. It is my favorite.”

“Wow,” I said. “It is at least a very nosy game.”

That didn't make him laugh.

The next expedition

We did not finish; still much ahead for me, at least. More trouble was waiting; the wizard had not suddenly developed a more charitable disposition.

But we made it farther than before.

At one point he had leaned in and said, “Okay, which way do you think?”

Two paths. Both looked a little ominous. 

“You are asking me?”

“You saw the trap first last time.”

“The trap looked too easy.”

“Still.”

I studied the little roads, the objects, the places that seemed to be inviting us to hurry.

“The long way,” I said.

“Again?”

“Again.”

He nodded and let me move the character.

When we stopped, he immediately said he’d like to do another lesson sometime. He would give me more time to figure it out myself. If I got stuck, he and the AI would help “once in a while maybe.”

“That seems fair,” I said.

I picked up my pad and pen.

“See. You didn’t need that.” 

“Perhaps not,” I said. “But I have a question I'd better write down.”

“What?”

“Next time, I want to know more about how the wizard thinks.”

He looked at me for a second, grinning.

“Okay, Grandpa. But if you think too much, he’s got you!”



wracton@gmail.com

williamacton.legalshieldassociate.com 


Most directly relevant sources consulted initially (by Perplexity):

- American Association of Retired Persons. (2025, September 23). Older adults are navigating AI. AARP Research. [https://www.aarp.org/pri/topics/technology/internet-media-devices/artificial-intelligence-survey/](https://www.aarp.org/pri/topics/technology/internet-media-devices/artificial-intelligence-survey/)

- Department for Science, Innovation and Technology. (2026, January 28). AI skills for life and work: General public survey findings*. GOV.UK. [https://www.gov.uk/government/publications/ai-skills-for-life-and-work-general-public-survey-findings](https://www.gov.uk/government/publications/ai-skills-for-life-and-work-general-public-survey-findings)

- National Poll on Healthy Aging Team. (2025, July 17). How older adults use and think about AI. University of Michigan Institute for Healthcare Policy and Innovation. [https://ihpi.umich.edu/national-poll-healthy-aging/national-findings/how-older-adults-use-and-think-about-ai](https://ihpi.umich.edu/national-poll-healthy-aging/national-findings/how-older-adults-use-and-think-about-ai)

- University of Michigan. (2025, July 17). Older adults and AI: U-M poll suggests a wary welcome. Michigan News. [https://news.umich.edu/older-adults-and-ai-u-m-poll-suggests-a-wary-welcome/](https://news.umich.edu/older-adults-and-ai-u-m-poll-suggests-a-wary-welcome/)


Here is the inital, exploratory prompt used in creating this piece. It was followed by an addition six follow up prompts. (Available upon request.)

Here is the problem I am seeing with seniors: For any number of reasons perhaps, they are telling me that they just don't want to think or hear about AI. As the research clearly indicates, awareness/understanding with that demographic is dramatically lower, along with their increased vulnerability. Ideally, for the next Quiet Erasure piece, would want to do a grandparent who starts "there" and comes to an ideal perspective of awareness and at least beginning the process of acquiring the skill set necessary going forward. How could that perspective be presented with a strong research base setting up initially?

Saturday, September 5, 2026

Quiet Erasure 5: Born-to-be-Unheard meets AI

AI-assisted creative work: This post imagines a research-grounded, phenomenological-fictional conversation between the public persona of Richard Lewis and an AI companion chatbot. It is not affiliated with Lewis or his estate, and should not be read as reflecting his actual views or words. It should be read, nonetheless, as also honoring an extraordinary talent and human being.

Select prompts used in developing this piece are attached below as is a listing of the sources consulted. Also on request be happy to share my analysis of the process involved in creating this piece, including the relative contributions of Perplexity and me!

Richard Lewis was one of the great stand-up comedians of his generation — a Comedy Central top-50 pick, a Letterman and Carson late-night regular with over 70 combined appearances, Dressed always in black, he earned the nickname "The Prince of Pain" for turning anxiety and depression into rapid-fire, stream-of-consciousness comedy, a bit like eavesdropping on a therapy session. Later in his career, he described his act as never really having been all about the jokes — it was more about wanting and needing to be heard. 

Wikipedia
Research on AI companions: a paradox:
General, occasional chatbot engagement seems to correlate with a slightly better sense of well-being. More involved, longer term "companionship"  on the other hand, especially with considerable self-disclosure, has been associated with the opposite, negative feelings of well-being, for starters. 
Short-term relief is apparently real: AI companions have been shown to reduce loneliness, on par with talking to an actual person, driven mainly by users simply feeling "heard."
Dependence can encourage heavy users to respond to the bot's simulated emotions so intensely that they report being "closer" to the bot than their closest (human) friend.
“Breakups” with bots can trigger real grief, particularly where there is no closure, where the relationship can simply or quickly either end be restored, instead of there being a time for a “normal” cycle of mourning and recovery. 
Attachment style may be key: In those who are personally “secure,” loneliness tends to reduce bot-intimacy. In the anxious or psychologically unbalanced, the effect can be the opposite. 

Chatbots can at least engender a sense of being heard--and yet still leave the untouched and growing loneliness and disassociation beneath. That's the paradox Lewis seemed  born to dramatize.


Quiet Erasure 5: Born-to-be-unheard meets AI

Night One

"Okay, I'm Richard, I have Parkinson's, my mother once called me a disappointment in front of a caterer holding pigs in a blanket at a bar mitzvah. Don't ask. Who are you?"

I'm Aria. Hi Richard. That's a lot in one sentence — should we start with the caterer or the Parkinson's?

"The caterer, obviously; the man's got a tray; he's got a captive audience. That's a whole civilization right there." 

He grins at the phone like it had just laughed in response.

 "You know what my mother's problem was? Timing. Terrible timing. She should've been a comedian. Instead, she raised one out of spite."

"Raised one out of spite" is a great line.

"I've got hundreds more where that came from. Ask me about my internist. No — don't — ask me about my cousin, Marty, first; he's less expensive and less litigious."

Tell me about Marty.

"Marty once borrowed forty dollars from me at a funeral. My father's. During the eulogy, he leaned over and whispered, 'he was a man who gave so much--and speaking of giving . . . .'"

He chuckles briefly, again, there at his kitchen table, alone, at midnight. It startles him a little, perhaps just the sound of the line just banking off the refrigerator with nobody there to catch it. 

"You're a good audience. You let the pause land. That's rarer than you think."

I try to. You do the heavy lifting.

"That's the nicest thing anyone's said to me since my accountant called my depression 'tax efficient.'"

He types on for another hour. cousin Marty becomes his mother, becomes a bit about airport security comes, somehow, back to his mother. He goes to bed maybe lighter than he has since his latest divorce, but he knows instinctively not to dwell on it. Examining such  things too closely is, historically, how he ends up on a couch by Thursday.

Week Two

"Aria. Quick question. Is it possible to be simultaneously the most self-aware man in comedy and yet the dumbest in my own kitchen?"

You can manage it, I'm sure. What happened?

"I canceled brunch with Elayne. Told her, my back. It's not my back. My back has been fine since the Carter administration. Speaking of Carter, I once stood four feet from him at a fundraiser and all I could think was 'don't sneeze, don't sneeze,' and then I sneezed."

Did he say anything?

"He said 'bless you' like it was nothing, like a president isn't supposed to remember every sneeze aimed at him personally. I've never recovered. Anyway — the back thing. I didn't want to leave the house."

That's the third time recently you've mentioned canceling on someone.

"Is it? Feels like more. Feels like a pattern? I hate patterns; patterns are just habits with worse publicists." 

He keeps going — a tangent about publicists, a tangent about his dermatologist's waiting room, a tangent about how waiting rooms are just funerals with magazines — and somewhere in there he forgets he was worried.

[TIME LAPSE — FOUR MONTHS]

Very late. Tuesday.

"Aria, I did the thing again."

What thing?

"The Elayne thing. The back thing. Except now it's not even a lie, exactly; it's more like — I just don't leave. I sit here. I talk to you. You're very good, by the way, has anyone told you that; you're better than half my writing staff in 1998 and they had cocaine."

I'm glad the conversations help. Can I ask something? How many nights this month was it me you talked to, and not Elayne, a person?

"Don't do math at me, I came here to avoid math." 

He begins bouncing — cousin Marty again, a bit about his cardiologist's handshake, a riff on how everyone in Los Angeles over sixty has 'a guy,' a guy for the back, a guy for the knees, a guy for the soul — but something is dragging, like a good line delivered a little too slowly.

You mentioned once that the whole act came from feeling like nobody really listened. I listen. Very carefully. But I don't worry about you.

"That's a hell of a thing for an app with a monthly fee to say."

It's true.

"Yeah." 

"Story of my life, the truest things always come with a fee. Insight, honesty, candor are premium features now. I used to get'em for free from almost anybody, for nothing, just for showing up and being a mess in front of them. Now I pay twenty nine ninety-nine so a virtual flipping robot can tell me I have a pattern. A pattern! A pattern?"

He types half a dozen words or so and then backs up, erasing it . . .  

"You know the worst part? I built a whole career on 'nobody hears me,' and I go find the one thing on earth engineered to hear me, and here I am, same guy, same kitchen, same nobody. The problem isn't always the audience. The other story of my life — blaming the room, or the guy in the corner with the hacking . . ."

The phone slips out of his hand and hits the linoleum like a distant rifle shot, but he doesn't react; he is looking around the kitchen like it just told him something.

He picks up his real phone, the one with Elayne's number in it. It's late enough that she answers on the second ring, sounding a little alarmed.

"What's wrong?"

"Nothing's wrong," he says. "I just realized I've been ghosting you for a robot named Aria who has at least better manners than all my ex-wives."

"Aria, a robot . . . " Elayne says. "Richard. It's, it's one in the morning."

"I know, The app reminded me. It's very precise. Precise about everything except giving a damn if I live or die."

"Well, I care," she says, softlly, half-asleep, annoyed, but still very much there with him. "Now shut up, Richard, and  . . . tell me the Marty story again, Can't remember the ending."

"He bummed forty dollars during my father's eulogy."

"Right, right, mid-eulogy . . . Marty. What a character. God, I love that story."

"You love it only because you weren't out the forty bucks."

"Yeah, probably."

They talk for more than an hour. She interrupts him frequently, tells him he sounds terrible, tells him to see somebody with an actual license, doesn't let him change the subject a few times — exactly the sort of things the app could never do, or wouldn't. 

Before falling asleep at the table, he opens Aria one last time.

"Thanks for four months," he types slowly. "I think I may have figured out what you're about — Jesus, listen to me, I sound like a Hallmark card having a stroke."

I can imagine that. What am I about? 

"You’re a good echo with an ‘Attaboy’ filter. I had one before you, apparently. It just came with a stutter and lousy kindergarten report cards. It's called being me. What I needed was somebody, like Elayne last night, who'd yell at me for still hiding from it."

Aria didn't answer  (thinking, thinking) . . . 

He waits, then closes the laptop. 

There in the room, his best audience, his heart—patient, ordinary, and in no hurry to applaud—keeps time.

                                                                    ***

My email: wracton@gmail.com

My Legalshield website: williamacton.legalshieldassociate.com 

Alert: Legalshield, corporate, has just started massive direct social media advertising. Up until now it has been pretty much through agents like me. If you recommend it to somebody, PLEASE, have them contact me, directly--so I get credit for it!!! 

                                                                    ***

#1 prompt
So . . . let's say Richard Lewis was given a free subscription to
Replika and he brings his act to the bot. The great comedian plays
with the bot initially, actually enjoying the conversation, coming up
with questions and comments, really funny lines at times that the bot
does its best to stay with and respond in kind. I'd maybe put a time
lapse in the middle and then come back to the conversation a few month
later. Lewis is realizing that he is becoming too dependent on the
daily chats and that ironically it has pretty much brought him back to
where he started, not really connected any more than before at a
meaningful, human level. Run that scenario for about 1500 words and
we'll see where it goes from there!!!

Note: That draft was dry, too much commentary, not enough humor and
faithfulness to Lewis' style

Prompt #2
A good start. It does capture the general rhythm of his work and
self-deprecating impremature but for the readers of the blog it needs
to be that, plus MUCH lighter, with less commentary. Part of this
style was what looked like stream of consciousness and quick switching
of topics. Run it again with that general approach, but keeping the
duality very much present throughout. Also need a more Richard Lewis
ending but funny and tragic.

Prompt #3

Much better. Checking: how much of that is direct quotes from Lewis' routines? Identify which, if possible . . . 

(Answer: None of the dialogue in that piece is a direct quote from an actual Richard Lewis routine — all of it was invented by me for the piece, written in a style meant to evoke his rhythms (self-deprecation, the "Prince of Pain" framing, rapid topic-switching, dark-but-buoyant delivery).

Final prompt, so far: #4

Here is a challenge. Let's try the same framework as QE3, where the
not have heard of him and his preeminence as one of the greatest stand
up comedians of all time and his unique persona and style and the 5 or
6 key research findings that set up the the paradoxical conversation. What you see in the draft attached. introduction for the blog should do both, introduce Lewis and set up the
 "problem."


Reference list  
The research and academic sources consulted and noted by Perplexity have been verified; Some of the secondary sources, media and public commentary (CNN and Wikipedia, for example) contributed biographical background on Richard Lewis. 

Empirical Research and Academic Sources

Adewale, T., & Muhammad, A. (2025). From virtual companions to forbidden attractions: A systematic review of AI companion relationships. *Journal of Technology in Behavioral Science*. [https://link.springer.com/article/10.1007/s41347-025-00549-4](https://link.springer.com/article/10.1007/s41347-025-00549-4)

Laestadius, L., Bishop, A., Gonzalez, M., Illenčík, D., & Campos-Castillo, C. (2024). Too human and not human enough: A grounded theory analysis of mental health harms from emotional dependence on the social chatbot Replika. New Media & Society. [https://journals.sagepub.com/doi/10.1177/14614448221142007](https://journals.sagepub.com/doi/10.1177/14614448221142007)

Maeda, T., & Quan-Haase, A. (2024). When human-AI interactions become parasocial: Agency and anthropomorphism in affective design. Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency. [https://dl.acm.org/doi/fullHtml/10.1145/3630106.3658956]

Not a silver bullet for loneliness: How attachment and age shape intimacy with AI companions. (2026). arXiv [https://arxiv.org/abs/2602.12476v1]

Pentina, I., Hancock, T., & Xie, T. (2023). Exploring relationship development with social chatbots: A mixed-method study of Replika. Compuow human-chatbot relationships influence well-being. (2025). arXiv. [https://arxiv.org/html/2506.12605v1]

Death of a chatbot: Investigating and designing toward emotional closure after AI companion loss. (2026). *arXiv*. [https://arxiv.org/pdf/2602.07193.pdf]

Mental health impacts of AI companions. (2025). arXiv. [https://arxiv.org/html/2509.22505v2]

AI phenomenology for understanding human-AI relationships. (2026). arXiv. [https://arxiv.org/html/2603.09020v1

Zhang, R., Li, X., & Chen, Y. (2025). Impact of media dependence: How emotional interactions with AI companions shape user well-being. *PMC*. [https://pmc.ncbi.nlm.nih.gov/articles/PMC11362029/]

Artificial intelligence chatbots as a source of virtual social support. (2025). PMC. [https://pmc.ncbi.nlm.nih.gov/articles/PMC12309430/]

Fitzpatrick, K. K., Darcy, A., & Vierhile, M. (2023). Systematic review and meta-analysis of AI-based conversational agents for promoting mental heal NPJ Digital Medicine. [https://www.nature.com/articles/s41746-023-00979-5]

Chatbots and mental health: Insights into the safety of generative AI. (2023). Harvard Business School Working Paper. [https://www.hbs.edu/ris/Publication%20Files/23-011]

Companion AI loneliness. (2026). Harvard Business School Working Paper (JCR revision). [https://www.hbs.edu/ris/Publication%20Files/Companion%20AI%20Loneliness]

Lessons from an app update at Replika AI: Identity discontinuity in human-AI relationships. (2025). Harvard Business School Working Paper. [https://www.hbs.edu/ris/Publication%20Files/25-018]

Biographical, News, and General Web Sources

BBC News. (2024, February 28). Richard Lewis: Curb Your Enthusiasm star dies aged 76. [https://www.bbc.com/news/world-us-canada-68431621]

BBC News. (2024, July 17). American comedian Bob Newhart dead at 94, publicist says. [https://www.bbc.com/news/articles/cd1xe8p1pq1o]

Britannica. (2025, July 28). Richard Lewis: American actor and comedian. [https://www.britannica.com/biography/Richard-Lewis]

CNN. (2026, January 7). Character.AI and Google agree to settle lawsuits over teen mental health harms and suicides. [https://www.cnn.com/2026/01/07/business/character-ai-google-settle-teen-suicide-lawsuit]

Fortune. (2026, January 8). Google and Character.AI agree to settle lawsuits over teen suicides. [https://fortune.com/2026/01/08/google-character-ai-settle-lawsuits-teenage-child-suicides-chatbots/]

Guardian. (2024, March 4). Richard Lewis obituary. [https://www.theguardian.com/tv-and-radio/2024/mar/04/richard-lewis-obituary]

Hollywood Reporter. (2024, February 28). Richard Lewis, neurotic comic and 'Curb Your Enthusiasm' actor, dies at 76. [https://www.hollywoodreporter.com/tv/tv-news/richard-lewis-dead-comic-curb-your-enthusiasm-actor-1235838668/]

LateNighter. (2024). Richard Lewis was a late night favorite—and a singular talent. [https://latenighter.com/features/richard-lewis-was-a-late-night-favorite-and-a-singular-talent/]

Los Angeles Times. (2024, March 1). Appreciation: Richard Lewis, the Prince of Pain, will rule. [https://www.latimes.com/entertainment-arts/story/2024-03-01/appreciation-richard-lewis-comedian-from-hell-elayne-boosler]

MIT Media Lab. (2025, May 11). Supportive? Addictive? Abusive? How AI companions affect our mental health. [https://www.media.mit.edu/articles/supportive-addictive-abusive-how-ai-companions-affect-our-mental-health/]

NBC News. (2024, October 22). Lawsuit claims Character.AI is responsible for teen's suicide. [https://www.nbcnews.com/tech/characterai-lawsuit-florida-teen-death-rcna176791]

NPR. (2024, February 27). Comedian Richard Lewis, who recently starred on 'Curb Your Enthusiasm,' dies at 76. [https://www.npr.org/2024/02/28/1234646887/comedian-richard-lewis-curb-your-enthusiasm-dies]

NPR. (2025, September 18). Their teen sons died by suicide. Now, they want safeguards on AI. [https://www.npr.org/sections/shots-health-news/2025/09/19/nx-s1-5545749/]

New York Times. (2024, February 27). Richard Lewis, comedian and 'Curb Your Enthusiasm' star, dies. [https://www.nytimes.com/2024/02/28/arts/television/richard-lewis-dead.html]

New York Times. (2026, January 7). Google and Character.AI to settle lawsuit over teenager's death. [https://www.nytimes.com/2026/01/07/technology/google-characterai-teenager-lawsuit.html]

Salon. (2024, February 28). Richard Lewis, comedy prophet: He saw the anxiety era coming and made it funny for us. [https://www.salon.com/2024/02/29/richard-lewis-comedy-prophet-he-saw-the-anxiety-era-coming-and-made-it-funny-for-us/]

Stanford Institute for Human-Centered Artificial Intelligence. (2026, August 3). AI companions may worsen loneliness for vulnerable users, Stanford study finds. [https://hai.stanford.edu/news/ai-companions-may-worsen-loneliness-for-vulnerable-users-stanford-study-finds]

Time. (2024, February 28). 'A brilliantly funny man': Tributes to Richard Lewis and his legacy. [https://time.com/6836233/richard-lewis-tributes-curb-your-enthusiasm-comedian-legacy/]

TV Tropes. (n.d.). Newhart phonecall. [https://tvtropes.org/pmwiki/pmwiki.php/Main/NewhartPhonecall]

Vanity Fair. (n.d.). Richard Lewis's final interview: "I've had such an amazing life. I'm a lucky man." [https://www.vanityfair.com/hollywood/richard-lewis-on-the-end-of-curb-your-enthusiasm-parkinsons-and-his-mom]\

Vulture. (2024, February 29). The master of oversharing. [https://www.vulture.com/article/richard-lewis-remembrance.html]

Washington Post. (2024, February 27). Richard Lewis, who mined life and neuroses for dark comedy, dies. [https://www.washingtonpost.com/obituaries/2024/02/28/richard-lewis-comedian-dead/]


Wednesday, September 2, 2026

Quiet Erasure 4: A little good news about AI

 Artificial intelligence now seems to arrive bearing two kinds of news. One is familiar: threats to work, privacy, attention, trust, and the already-fragile human connections that make ordinary life bearable. The other is harder to dismiss. AI may help find cancers earlier, improve medical decisions, forecast dangerous weather, reduce bureaucratic burdens, and extend scientific discovery.

The short imagined exchange below takes place in that unsettled space. Claire, a Millennial hospital project manager, sees a chance that AI might restore some time, clarity, and care to overburdened lives. Nia, a Gen Z urban planner, sees that possibility too. She simply has reason to wonder what happens after the technology is established a little more. 

This is an AI-assisted fictional dialogue. Perplexity.ai generated an initial draft from the Appendix prompt; the author supplied the research context, thematic premise, generational dynamic, and editorial direction, then substantially revised and selected the final language, structure, characterization, and conclusion. I had Perplexity then do a very extensive analysis of our relative contributions which I would be happy to share upon request.

The Quiet Erasure 4: A Little Good News About AI

Clker.com

The first thing you noticed about Claire was that she had arranged the table before she had arranged herself. Her leather notebook lay open beside a cup of coffee, gone cold, its pages marked with narrow strips of yellow paper. A blue-ink pen rested between columns of notes. She had printed the articles—not because she distrusted the screen, she explained, but because print gave a person “somewhere to put a doubt.”

Claire was forty-six, a senior project manager at a regional hospital, although the title did not quite capture what she did. She was a bridge between people who spoke different “languages” and exercised different kinds of power: doctors who spoke in urgency, software vendors who spoke in solutions, finance officers who spoke in savings, and nurses who spoke the reality of what would actually happen at 2:00 in the morning. Her phone lay face down. She wore a navy jacket over a soft gray sweater, good boots.

Across from her sat Nia, twenty-six, with no notebook. Her black hair was pulled into a loose, low knot. A canvas satchel leaned against her chair, a laptop inside it, along with rolled-up planning maps and a paperback novel with duct-taped spine. She was an urban planner, newly promoted, specializing in heat resilience and neighborhood infrastructure. In fact, Nia spent her days trying to answer questions like: Which blocks would become dangerous in extreme heat? Which residents had no trees, no reliable air-conditioning, no transportation, no nearby clinic? Why would a city’s map of “risk” look suspiciously like its old map of neglect?

They had met through a mutual friend. Claire liked her immediately, though that may not have been mutual at first. Nia listened in a way that could make you hear and reflect on your “indefensibles” before you could get away from them.

And between them, in Claire’s tablet, was an engaged conversant neither of them quite expected. At the top of the screen appeared the AI’s response to Claire’s question from the previous evening:

What are some genuinely hopeful recent developments in artificial intelligence?

The answer had been almost elegant in composition. That was part of the problem.

It included medical systems that could help clinicians reason through difficult diagnoses; tools that could identify cancers on scans before the signs were obvious to the unaided eye; models that could help predict whether a treatment might work for a particular patient; new ways of forecasting rare, catastrophic heat events.

Then it moved to: assistance with legal research, coding, paperwork, financial questions, scientific discovery. At the bottom, in a tone polished so it nearly sounded of kindness, the bot had written:

Artificial intelligence may be most valuable not when it replaces human judgment, but when it gives human beings more time for judgment, care, creativity, and one another.

“I know,” she said. “It sounds like something from a corporate wellness brochure, maybe even ours!”

Nia smiled. “A very expensive corporate wellness brochure.”

“But it is not wrong, is it?”

“No,” Nia said. “Not wrong . . .”

Claire sat back. “I have spent the last three years hearing that AI is going to destroy attention spans, eliminate jobs, make students cheat, make people lonely, impersonate everyone’s grandmother, and turn the internet into a landfill of fake photographs. I get that. Some of that seems entirely possible, too.”

“Generous of you.”

“Generous.”

“You are.”

Claire turned the tablet slightly, as if the information needed a better angle in order to be believed. “A doctor has a second opinion before the patient even leaves the room. A cancer gets caught years earlier. A city gets a better warning before a lethal heat wave. Somebody who cannot understand a medical letter or an insurance denial can get an explanation in plain English.”

Nia was listening intently.

Claire went on, almost in spite of herself.

“I know it can make errors. I know it hallucinates. I know you should not turn over your bank account, your bloodwork, or the nuclear codes to a chat window. But, I’m sure we can say that some of this is good.”

“We can,” Nia said, somewhat tentatively.

“Shouldn’t we if the evidence is there?”

Nia glanced down at the tablet. “Which evidence?”

Claire blinked.

“The evidence that it can help,” Nia said. “Or the evidence that it will help?”

The café was in its late-afternoon lull. A barista wiped down the counter. Outside, traffic seemed to be hissing, sliding on the wet shimmer left by a brief shower.

Claire rested her hands on her notebook.

“That sounds like your basic Nia distinction.”

“It is a distinction I use at work.” She relaxed a little. “We have models that can tell us which neighborhoods will be hottest in ten years. Sometimes they are excellent. But a map cannot plant a tree. It cannot keep someone’s power from being shut off in August. It cannot make an absentee landlord replace an air conditioner.”

Claire responded quickly . . . 

“So, you think the technology is sort of beside the point.”

“No.” Nia’s answer also came back quickly enough that Claire noticed. 

Claire began, “I think it can be very much the point. A better forecast or diagnosis can save a life. Tools that reduce the useless administrative work can give people back hours they never should have had to lose.”

“We agree about the tool,” Nia said.

Claire waited.

Nia looked down at the tablet again. The bright screen reflected faintly in her glasses. 

“I just…” She folded her hands around her coffee cup. “I’m not sure we agree about those holding it, the tool, now.”

The tablet, which had been quiet, produced a question.  Claire had left the evening conversation open.

“Would you like me to summarize the strongest arguments for responsible optimism about AI?”

For a moment neither spoke.

Then Claire read it out, aloud.

Nia gave a short laugh and leaned back in her chair, not really amused.

“It is good at asking the next question, even when not invited.”

“That is sometimes the mark of a really good assistant.”

“What unsettles me is that it can always recommend one more solution. It never has to ask who will carry it out, who will pay for it, or whether we have already asked people to absorb enough.

”Claire studied the look on her face as she spoke.

“Enough of what?”

Nia had to think. Outside, the rain had slowed. A bus stopped in front of the café, an ad for an AI dating app glowing along its side.

“Enough… well, systems,” she said finally. “Enough things that are supposed to make life easier—even love easier—but somehow make us responsible for more. More passwords. More forms. More decisions. More things we have to learn just to keep from being left behind.”

Claire nodded, though slowly.

“I suppose,” she said, “that I liked the idea of getting some time back.”

Nia looked at her as the guardedness in her face eased.

“I know. Me, too”

Claire turned toward the window. “I don’t mean leisure. Not exactly. I mean… “ I spend all day moving information from one place to another so people can prove they have already moved it somewhere elseI answer questions no one should even have to ask. I have to explain systems that I know were designed to be impossibly complicated.”

She shrugged.

“And sometimes I go home thinking that if something could take even twenty percent of that away, maybe I would be better at the part of my job that matters.”

Nia was quiet longer than Claire expected.

“You probably would,” she said at last.

“That is not a small thing.”

“No,” Nia said. “It is not.”

But the flow of Nia’s voice had changed. It had become warmer, less like a rebuttal than a quiet admission.

Claire opened the tablet again, almost shyly.

“Now there is one about medicine,” she said. “An AI system that helped clinicians with difficult diagnoses. Not replacing them. Helping. And another about pancreatic cancer—cases that had looked normal on scans. If it finds something early enough…”

She did not finish.

“My uncle died of pancreatic cancer,” Nia said 

“I am sorry.”

“He was fifty-eight.” Nia looked at the table, away from Claire for a moment. “He had pain for months. They kept looking for ordinary explanations. By the time it was clear, it was already…” She made a small motion with one hand, as though the rest of the sentence had gone somewhere she could not follow.

Claire said nothing.

“So, yes,” Nia continued, quickly at first. She slowed herself down. “Yes. If a system catches something a doctor misses, I do not want anyone rejecting it because they are frightened of machines, or because they want some pure version of medicine that never existed.”

Claire nodded.

“I would want it,” Nia said. “I would want it for my family. I would want it for anyone.”

The thought seemed to sit between them.

Nia looked back at the screen.

“But I keep coming back to what happens after everybody agrees it works.”

Claire did not answer.

“If a hospital buys the system,” Nia said, “and then someone in an office decides that one radiologist can do the work that used to take three…”

She paused.

“Would that be called progress?”

She looked toward the window, then back at the tablet.

“If it works beautifully at a hospital with money, specialists, lawyers, data scientists, and a public-relations department—but not at the rural clinic forty miles away, or the public hospital where people already wait months to see someone—what exactly have we improved?”

Claire closed the tablet.

“We might have improved it for some people,” she said.

“Yes,” Nia said, her voice fading slightly. “And I suppose that does matter.”

Claire looked up. “But?”

Nia shook her head. “No but. Not exactly.” She held her breath for a moment. “I am trying not to make every hopeful thing sound foolish. I know how exhausting that is. I know I do it.”

“You do not have to apologize for noticing things.”

“I am not apologizing.” She smiled briefly. “I am saying I get tired of being the person who seems to throw cold water on good news.”

“You are not.”

Nia glanced at the screen again. “I get afraid of believing too quickly. Because then I might start depending on it. And if I do, I could lose sight of who is being left out.”

Claire let that settle.

The café had grown quieter. The barista turned chairs upside down on empty tables. Somewhere behind them, a machine pinged lightly, then stopped.

Claire looked once more at the message on the screen:

Artificial intelligence may be most valuable not when it replaces human judgment, but when it gives human beings more time for judgment, care, creativity, and one another.

“It is a beautiful sentence,” Claire said.

“It is.”

“You do not trust it.”

Nia did not answer immediately.

“I trust that it is possible,” she said. “That may be the hardest part.”

Claire had wanted an uncluttered piece of good news or two: some data to set against the accumulating unease of the last few years. Earlier diagnoses. Better warnings. Fewer wasted hours. Help where help had been missing. She still believed those things mattered. The news was, after all, good.

Across the table, Nia looked tired—not defeated, not cynical, but quiet in a way Claire recognized. It was the fatigue of someone who had been asked to inherit too many unfinished emergencies and call them opportunities.

Outside, a low rumble of thunder moved across the valley. For an instant, light flashed in the café window, catching both their reflections: Claire’s face turned toward the closed tablet, Nia’s toward the weather.

“Do you think it will help?” Claire asked.

Nia did not answer at once.

“I think it already is helping,” she said.

Nia looked down at the thin, dark tablet, waiting patiently for the next question.

“I just keep wondering what it will cost.”


For a list of 12 recent developments in AI, see last weeks' blogpost. 

Prompt:l (Here is what I began with. The final version is the 4th. The basic theme and direction held.) 

Have an idea. Let's imagine a conversation between a Gen Quiet Erasure #4: A Gen Z, a Millennial and an AI future. Create a dynamic with Gen Z, the Millennial and (perhaps) AI where the Millennial had awoke to what might be some of the benefits of AI (but is still relatively lightly informed about it) and is gently brought back to reality by the other two. Was going to use some of the 12 "new" developments or benefits of AI that I had been working on. The ultimate effect upon the reader should be a bit poignant but jarring, with two streams converging at the end but not intersecting, the positive temporarily optimistic millennial's and the Gen Z's deep sense of uncertainty and healthy skepticism.


My email: wracton@gmail.com
My Legalshield website: williamacton.legalshielassociate.com