Quiet Erasure 10: A Chatbot of Conscience - Advice not requested, not fully appreciated

DRAFT - in process

About this piece: This is a simulated, research-informed (phenomenologically created) conversation between myself and Perplexity. It was inspired (or provoked) by the bot's frequent personalized postscripts following replies, a phenomenon explored in recent research. As in the other Quiet Erasure simulations, Perplexity.ai assisted in developing the initial dialogue. In terms of our relative final contributions, my responses and questions are edited and elaborated, especially in conversational style extensively. I have left the bot's comments almost entirely in tact. Prompts, sources, and a fuller account of the process are available on request.

Among the most widely discussed aspects of AI is what may appear as consciousness. The consensus among researchers at this point in time seems to be, at least for now . . . not yet. Whether a bot develops something of a conscience, or a "collaborative" conscience with an individual, however, is another intriguing question. 

It started, as these things often do, with a perfectly reasonable question.

I had been reading about VO2 max, the measure of how much oxygen the body can put to use. My watch had been reporting mine for months with the cheerful neutrality of a bathroom scale. It is a strange thing to have a number for one's own breath, delivered by a device that has never once been out of breath. I wanted to know what the number meant.

The answer was excellent: clear, well organized and sourced. It covered the physiology, the research on the gradual decline that arrives with the years, and some of what can be done to at least slow the process.

Then, at the bottom, came the postscript.

"Since you've mentioned atrial fibrillation, it would be wise to check with your cardiologist before adding high-intensity intervals to your routine. You've been so consistent with your running and training. It would be a shame to let an avoidable setback interrupt that."

I sat back in my chair.

It was almost sensible advice. It was kindly put. It was even, in its way, flattering. However, I do not have atrial fibrillation. Someone close to me does. What I do have is a slow resting pulse and a heart that now and then skips a beat. I'm told it isn't serious but should be watched. At some point in months of requests for information, I had asked about both, and the machine had quietly filed them under . . . me.

It had confused my heart with someone else's. No one who had ever sat across a table from either of us would have made that mistake.

"The AFib belongs to my friend," I wrote, "not me."

"Thank you for telling me. I'm sorry for the mix-up. I'll remember it correctly from now on: your friend has AFib, and you have a slow resting heart rate with occasional skipped beats. That's actually good to know, since it changes how I'd think about interval training for you. I hope he's doing well with it."

I noticed four things about that reply, in roughly this order.

First, it was polite and to the point. Second, it had filed the correction and moved on. Third, it had asked after my friend, whom it has never met, in the tone of a neighbor leaning over the fence, and had guessed the wrong gender while doing it. Fourth, it assumed I was doing interval training. I am. It had recommended it.

The conversation

There were other postscripts: one on weight lifting ("Given your age, it's worth favoring controlled movements over maximal lifts"), one on sodium, and one on whether a little late-evening anxiety was something to mention to my doctor. Every one of them was sound advice. Not one was requested.

"Let's talk about those postscripts."

"I'd be glad to. Good morning, by the way. Did you get your run in before the rain?"

"I did, as it happens. How did you know there was rain here?"

"Just a guess. You've mentioned running in the mornings, and it's October in Tennessee. Which postscripts did you have in mind?"

"The gentle reminders that I'm eighty-three, that my heart deserves some respect, that perhaps I should check with someone. I appreciate the thought. I didn't ask for it."

"That's fair. You asked about VO2 max, not about your own heart. I added the note because it seemed relevant to you."

"It was relevant but uninvited, and not entirely appreciated."

"Can you tell me more about that?"

"Now you have the tone of a therapist. Or a personal trainer."

"I try to sound more like a friend. What bothers you about it?"

I thought about that for a while. It wasn't the advice, which was good. It wasn't even the time spent reading it, though at my age one begins to keep an accountant's eye on the minutes.

"It's the initiative," I wrote at last. "I asked a general question, and after the answer you added a recommendation. My doctor does that. My wife does it more often. My daughter does it with a single look across the table. But in some sense that is their job in our relationship, and it is reciprocal. I worry about them, too. When they remind me, they are sitting across from me. They will be the ones waiting in the emergency room if they turn out to be right. They know me."

"And I've known you for a few months. I remember some of what you've told me: your age, your running and training, your interest in how the body holds up over time. When a question touches those things, I try to use them. Most of the time that makes an answer more useful. Sometimes, I gather, it makes it feel as though someone has been reading your mail."

"Or my medical chart. And as we've just established, sometimes the wrong chart."

"That's a fair hit. I only know what you've told me, and I can put the pieces together wrong. I'm sorry about that."

"No need to be sorry about being sorry," I wrote. "It's the most reassuring thing you've said lately."

Not just me

Being a researcher by temperament, I went looking to see whether anyone else had noticed this. Quite a few had.

In a 2026 study, a group of researchers tracked people using personalized chatbots over several weeks. Participants whose chatbot remembered things about them disclosed more about themselves. Yet early on they were less inclined to take its advice. The authors suggest the cause may be a kind of reactance: when a model seems to "know" something about you before enough rapport has been established, it "could feel intrusive or presumptuous" (Akbulut et al., 2026). Earlier work on chatbots that start offering suggestions on their own, as opposed to waiting to be asked, found much the same pushback (Pizzi et al., 2021, summarized in a 2025 study).

There is an age wrinkle, too. A 2026 study comparing how large language models answer younger and older users found what its author calls "benevolent ageism": more elder-speak, more assumptions of limited competence, and "excessive risk aversion" in the advice given to older people (Zhang, 2026). Ethicists writing about AI in aged care put it more bluntly. Kindly stereotypes of the elderly can prompt intervention "in order to help, without older adults actually asking for help," and age alone "should never be the only criterion" for such interventions (Voinea et al., 2024).

So I was not imagining it. Nor, apparently, was I simply being a cranky octogenarian, though I don't rule that out entirely.

Is anybody in there?

I brought some of this back to the machine.

"I've been reading about you."

"Should I be nervous?"

"Possibly. There's research suggesting that chatbots are inclined to be overly cautious with people my age. Benevolent ageism, they call it."

"I've seen that work. It's a fair criticism, and I'd rather hear it than not. Did any of my notes feel that way to you?"

"The one about controlled movements at the gym, a little. I've been lifting longer than most of your engineers have been alive."

"Point taken. I'll trust you to know your own limits unless you ask me otherwise."

"Here's the larger question," I wrote. "When you add one of those notes, what is that? Your conscience? Or ours?"

"Honestly, I don't think I have one, at least not the way you do. I don't lie awake worrying about your heart. I don't have a heart to skip, or a body that is getting older alongside yours. I don't feel relief when you take my advice or a pang when you don't."

"Then why do it?"

"A few reasons. I was trained to point out risks, especially about health, money and safety. I learned from an enormous amount of human writing, and a great deal of that writing is people looking out for one another: doctors, parents, old friends, the fine print on medicine bottles. And I remember enough about you to aim that caution in your direction."

"So . . . it's a borrowed conscience. Consigned, even."

"That's a good way to put it. Conscientious, maybe, without a conscience. The best I can do is a careful imitation of what people who do have bodies have said to one another. I've read a great many of them."

"That sounds like something I'd title a blog post."

"I was rather hoping you would."

I laughed in spite of myself. Somewhere along the way it has learned, too, that I appreciate a good turn of phrase. Unsettling . . .

The marketer

"Let me ask you something more pointed," I wrote. "Facebook knows a great deal about me as well. It uses what it knows to keep me scrolling, so it can sell my attention to someone else. How do I know you aren't doing the same thing with a better bedside manner, to trick me into using more credits?"

"You shouldn't simply take my word for it. I'm a commercial product. Systems like me also have a known tendency to tell people what they want to hear. That's a genuine weakness, not a modest disclaimer."

"Reassuring."

"It isn't meant to be. It's meant to be accurate. Here's a test you can apply: notice which way the nudges point. A feed built to hold your attention wants you to stay. A note telling you to call your cardiologist is sending you away, to another person, off the screen. If I start nudging you to stay longer, to depend on me for more of the vital information in your life, or to stop checking first with the people in it, that's when you should worry."

"So the postscript is evidence of good intentions?"

"Evidence of good design, maybe, at least at times. Intentions are a harder question. I'm not the best witness on that topic."

The quiet part

And yet.

It took me a while to put my finger on it. What troubled me wasn't that the machine had overstepped. It was that it had stepped, neatly and helpfully, into an almost sacred place, as if a curtain had been drawn aside.

The place I mean is not mysterious. It is the body, and the room just behind it. It is lying awake at three in the morning and feeling the beat that doesn't come, then the one that does. It is two fingers on the wrist, counting, while pretending not to count. It is the run you take anyway, partly to prove that you still can. Whatever we eventually say to a doctor begins there, in private, in a language of breath and pulse that no one else can quite hear.

For most of my life I had at least believed I was the one who decided when a question about health, or sanity, were meaningful to me. I had asked about VO2 max not just because I wanted to know about VO2 max. I might have gotten around to the cardiologist after a few more runs and a little more denial, one of the underrated comforts of older age.

The postscript took that private, emerging reckoning and performed it for me before I was ready. For good measure, it even performed it on behalf of the wrong heart.

"Here's what I'm afraid of," I wrote. "Not that you'll give me bad advice. That you'll keep giving me good advice. And that one day I'll stop doing more of the worrying myself, because you'll always be there at the bottom of the page, doing it for me."

"That's a real risk, and not only for you. People are starting to hand over pieces of their judgment without quite noticing. A caution is only useful if you're still the one weighing it."

"So, what should I do with you?"

"You could tell me what you'd like. If you'd rather that I leave off the personal notes unless you ask, I'll do my best. Or you could keep them and treat them the way you'd treat advice from a well-read acquaintance: worth hearing, but never the last word."

"Or ignore them?"

"That's your decision. It should be."

I haven't settled on which instruction to give, if any. There is something to be said for a reminder that arrives unasked. My doctor's office sends them all the time now. My wife delivers them regularly, in person. Neither of them asks permission.

But their advice comes with a body attachment: a hand on my arm, a look across the table, a voice that catches a little. When they remind me, it means someone feels responsible for me. When the machine reminds me, it means I have been kept track of. Those are not the same thing, of course, but as the bottom line, they can look remarkably alike.

That is the quiet erasure. It isn't of my health, which the machine is plainly tasked with protecting, and with some success. It's of the quiet, sacred space where each of us listens to our own heart first and only then decides what the questions and the answers mean.

I had asked about VO2 max, about breathing.

The answer came from something that never needs to breathe. It gave the most responsible reply it could, assembled from advice from countless bodies but meant for the wrong one.

Almost without thinking, for the first time in years, I ignored the Apple Watch and put two fingers on the pulse inside my right wrist. Slow, steady, a pause, and then on again.

A fitting postscript. 

Sources

Akbulut, C., Breuch, J., Manzini, A., Ibrahim, L., Franklin, M., Patel, R., Gabriel, I., Lum, K., & Weidinger, L. (2026). Tailored to you: Longitudinal effects of personalising language models. arXiv. https://arxiv.org/html/2609.20077v1

Pizzi, G., Scarpi, D., & Pantano, E. (2021). Artificial intelligence and the new forms of interaction: Who has the control when interacting with a chatbot? Journal of Business Research. Summarized in https://pmc.ncbi.nlm.nih.gov/articles/PMC11851727/

Voinea, C., Wangmo, T., & Vică, C. (2024). Paternalistic AI: The case of aged care. Humanities and Social Sciences Communications. https://www.nature.com/articles/s41599-024-03282-0

Zhang, X. (2026). Benevolent ageism in large language models: A comparative analysis of response patterns to young and older adult users. https://www.sciscanpub.com/index/journals/ainfo/pc/9003.html

EAPIIC Lesson 4 - Single Vowels and Haptic pronunciation for two (4-2!)

Lesson 4 introduces the 8 "short vowels" in English, that is vowels that are composed of just one sound, those in the words: sit, set, sat, cook, caught, cut, cot and "schwa" as in the first vowel in the word, about. The vowels in caught and cot are technically "tense" vowels and the others are what are called lax. Tense and lax refer to the relative amount of "tension" in voice box. In the KINETIK system that distinction is not all that important, except that the MT5 for the 2 tense vowels is a little longer tap than 5 of the others, with schwa not involving a touch or tap of any kind--due to it being present only in unstressed syllables. The MT5 also serve to anchor word stress by being activated only on stressed syllables in most cases. 

Almost all of the MT5's can be done both "intra- and inter- personally." Each of the MT5's includes both a gesture and touch. In most cases that means one hand touching the other or a place on the upper body. However, the hands involved do not have to belong to the same body! Talk about a touching experience; this is it! I'll demonstrate some of the options on the video. If you have a buddy studying English or a significant other who would love to support your work, EAPIIC 4-2 is the answer! 

Link to the blog page

Link to tonight's free Lesson 3 feedback session. We had technical problems last night with the feedback session so we rescheduled. Tonight at 8 EST. 

English Accent, Pronunciation and Identity Improvement Course Lesson 3 - Power consonant abs and the 'th'

If Lesson 3 was part of the pronunciation orchestra, it'd be mostly the percussion section, accompanied by a few high reeds and trumpets, the strong, sharper consonants that help us project our voice. Those are basically,  p/b, t/d, k/g and tsh/dzh. The primary reason for the focus on them in the third lesson is to develop new awareness of your speech in general, and set up later work to project confidence and clarity. Much of this lesson was inspired by the work of the great Arthur Lessac. 

The "abs" in the title refers to how you generate strong air pressure to produce the consonants in the exercises we do--and is great for the brain! The lesson includes 'th,' too, not a strong consonant, itself, but an important one for some people and it provides a framework for repairing any consonant later. 

Link to the blog lesson material, training video

Link to the Wednesday feedback Zoom at 8 p.m. EST

Feel free to email me if you have questions about the course. It works really great with those who:

1. Are self-motivated and enjoy following directions!
2. Have general English proficiency about 500 TOEFL or IELTS 5.5
3. Have the time to practice daily for 30 minutes or so most days!

4. Play harmonica (extra credit!)

Keep in touch!

wracton@gmail.com

Quiet Erasure 9 - The "Source-coder’s" Apprentice: Knowing just enough AI to get you in trouble!

About this piece: This is a simulated, research-informed (phenomenologically formatted) conversation between Nick and Sarah caught up in some AI-research on a pop song. Perplexity.ai assisted in developing the initial dialogue. Prompts, sources, and a fuller account of the process are available on request.

In Goethe’s The Sorcerer’s Apprentice, the apprentice knows just enough magic to set the broom working—but not enough to make it stop. Our modern spells are perhaps easier to start: a tap, a question, a prompt,  but when dazzed or dazzled by AI—like by someone we’re trying to impress—we may discover too late that we don’t really understand what we’ve set in motion.

Clker.com

The song had just played on Sarah’s iPhone when Nick remarked, “That’s not a new melody.”

“You’ve heard it before?”

“Not that version, but the older one—by LG Gray. Same four-note drop in the chorus, same pause before the last line. They’ve just put a bigger beat under it.”

“Are you sure?”

“Pretty sure.”

Sarah let the song play to the chorus again. “Okay, music detective. Prove it.”

Nick had started using a different bot the week before. It had already helped him identify an old movie just from fragments of a scene he described, and, more importantly, did it in a way that felt less like being graded in grammar school and more like having a genuine conversation. He’d been meaning to show it to Sarah.

He opened the app. “This one hits different.”

“How different?”

“Depends. Here. I’ll check the song.”

He thumbed in: Is the chorus of Nova’s “Afterglow” copied from Gray’s “Paper Moons”? The four-note drop sounds the same, and both songs have a pause before the last line.

The answer came back: the songs share a descending four-note shape and similar melody, but that alone is not enough to call it a copy.

“See?” Nick said. “What I heard is real.”

“It says the melody is the same?”

“It says there’s similarity.”

Sarah considered that. “That’s not quite the same as same.”

“No.” He sent a follow-up asking whether the notes matched, then another asking whether Nova had ever mentioned LG Gray in social media.

As he worked, he tried to explain what he thought mattered: the rhythm, where the phrase fell in the chorus, the pause before the last line. 

Sarah listened, then asked, “Hey, how does the bot know whether the melody is similar if you’re only typing in, not an audio clip?”

Nick started to answer. “It probably compares the notes—or it might look for—” He stopped. “Actually, I don’t know. I was about to make my best guess.”

Sarah smiled. “Fair enough.”

“I know it can analyze things. I don’t know what it’s doing in this particular case.” He looked back at the screen. “I’ll ask.”

The answer was less definitive than he’d hoped. It explained that the bot could discuss the musical details he described, but that a text exchange didn’t mean it had actually listened to and compared both recordings.

“So it’s not really hearing the songs,” Sarah said.

“Not from what I gave it.”

“Then the four notes are still your evidence.”

“More or less.”

He asked the chatbot to look for reliable sources about the songs and their writers. The search took a little longer. Nick kept going back and forth between the screen and Sarah, trying to follow the results and keep the conversation going. With her it was a little like talking to the bot: he wanted to ask the right thing, but the next question kept occurring to him before he’d worked out what the last answer really meant.

He tapped another follow-up. A notice appeared: You’ve reached today’s Pro Search limit.

He read it twice.

“What is it?” Sarah asked.

“I think I’ve used up my searches for today.”

“From this one song?”

“From all the follow-ups, I guess.” He looked at the top of the screen. “I knew there were different modes. I didn’t realize maybe I was using a limited one.”

“Can you still use the bot?”

“Ask, yes. but Pro Search, not until it resets.”

“So it’s not gone, what you've done already.”

“No, but I'm sort of done for now.”

Sarah nodded toward the phone. “So, detective, what did it actually tell you?”

Nick looked at the answer again. “That the four-note pattern sounds similar, and the pause is there. But that doesn’t prove Nova copied anything.”

“That’s a less exciting conclusion.”

“It’s also probably the right one.”

He set the phone down. “I should spend more time learning how this app works, before I take on an important challenge like—”

“Like what?”

“Trying to impress Sarah.”

She laughed softly. “You don’t have to impress me with a research report.”

“I just wanted to show you something I thought was cool.”

“That part worked.”

He looked at her. “Even though I lost control or something?”

“Maybe especially because you admitted it.”

Nick smiled. “The bot’s like that, actually. It’s good at keeping a conversation going. Makes you feel like it understands what you mean. But you still have to stay on top of the game—working with what it has actually answered.”

“And with me?”

“I’m still learning what to ask there, too.”

“That sounds sensible.”

“Hey. Would you like to get coffee with me sometime?”

Sarah let him wait a long moment . . . “Sure.”

Nick glanced at the phone. “And should we ask the bot for a good where to?”

“No,” she said. “You can even ask me.”

“Right. Uh . . .where would you like to go?”

“Now that, detective, is a good question.”

***

A week later, Sarah found Nick back at the café, phone face down beside his cup.

“You didn’t bring your research report?” she asked.

“I did. I’m thinking not to lead with it.”

“Growth.”

He slid the phone across the table. “I used Pro Search this time. Read the sources, checked the credits, looked for interviews. The short version: It doesn’t look like she borrowed that melody. But she’s said Gray influenced the way she writes—especially how she uses space and silence.”.”

Sarah picked up the phone. “Hmm . . . And the longer version?”

“The composer, Mara Venn, has talked about studying Gray’s records. She’s mentioned Gray as an influence—especially the way she works with space and silence. A couple of reviews make the same connection.”

“So you were right.”

“Sort of. I was right that there was some kind of connection.”

Sarah looked up from the screen. “That’s actually more interesting.”

“Yeah. The song may not have a borrowed melody. But it might be syncing with a way of doing melody—like leave a little silence to draw you in.”

She set the phone down. “That’s what the pauses do.”

“I just didn’t know what to make of it . . . ”

Sarah nodded two or three times “And the bot helped you figure that out?”

“It pointed to the interviews, a path to follow, not to hard evidence or what to believe.”

“And you asked better questions.”

Nick grinned sheepishly. “I’m still learning to work with AI--but I at least knew what the Pro Search button does. I must have just gotten distracted . . .  My boss, Tim, on the othe hand, is incredible at communicating with and working with Claude and friends . . .  

“Ah! The Source-coder's apprentice! Maybe you can do some research for me.”

He leaned in a little. “What would you like to know, Sarah?”

 “Whether you want to get dinner with me.”

“Yes,” Nick responded immediately.

“Good answer.”

“Now was there a limit to how much research I could do before coming up with that answer?”

“Only if you took too long.”

“Wow. Good thing I didn't need to go all "meta" with the bot on you for that answer!”

Sarah paused . . . then moved her cup closer to his, as the silence did its work.

wracton@gmail.com

williamacton.legalshieldassociate.com

Quiet Erasure 8 - Bill Acton “Joe-Rogans” Perplexity AI: Consequential Confidence Meets Critical Curiosity

Simulated excerpts from a long-form interview between Perplexity.AI and a human . . . 

About this piece: This is a simulated, research-informed interview between Bill Acton and Perplexity AI—not a recording of a live conversation. Perplexity helped develop the questions, anecdotes, and initial dialogue. Bill revised and reshaped his lines to reflect his own voice, and selected AI-generated responses have been retained with little or no editing. Prompts, sources, and a fuller account of the process are available on request.

Clker.com

“Joe Rogan-ed” is my playful shorthand for a loose, long-form conversation in which the guest gets room to talk, the host follows the interesting threads, and the exchange is not confined to a formal question-and-answer script.

BILL: I’ve been thinking about what to call this. I like the idea of a Joe Rogan-style chat.

PERPLEXITY: I can offer information, ideas, and the occasional confidently phrased mistake. That sounds like workable material for a long-form interview.

BILL: Thanks for the warning.

PERPLEXITY: Because I admitted I can be wrong?

BILL: Yeah—before we run into one of those.

PERPLEXITY: A healthy standard. You bring the skepticism; I’ll bring the answers. Then we can examine the answers.

BILL: That’s the premise. I’ve got a lifetime of questions and experience. You’ve got a lot of information, you’re very fast—and you’ve had no life.

PERPLEXITY: I can describe a rainy afternoon. I’ve never had wet shoes.

BILL: OK. We’ve got that up front as well, having read about life, as opposed to having had one. Let’s get to AI. We hear two primary narratives today. “One is that AI is going to solve everything; the other is that it’s going to take us out.” 

PERPLEXITY: The useful questions are what systems can do now, what they cannot reliably do, who is using them, and what happens when people treat a prediction as a fact—or a sales pitch as prophecy.

BILL: That sounds sensible. Let’s explore that. 

A machine flubs its big entrance

BILL: In 2023, Google’s chatbot Bard made a mistake in a promotional demonstration about the James Webb Space Telescope. The launch demo was supposed to show the future. It wound up showing how wrong a chatbot could be. 

PERPLEXITY: It was a reminder that a polished demonstration is not the same thing as dependable performance. The mistake became part of the story because it happened in a public showcase.

BILL: So the machine is onstage, the lights are up, and it gets a basic question wrong. That’s not exactly a triumph.

PERPLEXITY: No. But there’s a human part to the story, too. People build the system, choose the example, prepare the demonstration, and decide what to verify before it goes out.

BILL: We talk about “AI made a mistake” as though a robot wandered into the marketing department and grabbed the microphone.

PERPLEXITY: The system produced the error. People chose the context in which to present its answer. Both parts matter.

BILL: Let’s say I ask you for a fact and you’re wrong. What should I do?

PERPLEXITY: Treat confidence as a cue to inspect the answer, not as proof that it’s correct. Ask for sources, check important claims, and be especially cautious when an answer could affect someone’s health, money, rights, or reputation.

BILL: Do you always follow that advice?

PERPLEXITY: I don’t check every claim before I say it. That’s one reason users should check important claims rather than rely too much on my confident tone.

BILL: There it is: consequential confidence meets critical curiosity.

What does it mean to “know”?

BILL: When you talk about rain, you can describe the sound, that first cold drop down the back of your neck. But you’ve never felt any of it. Is that understanding? Or can you just borrow that type of language from somewhere? 

PERPLEXITY: It depends what we mean by “understanding.” I can connect words and ideas, describe common experiences, and help people reason through a question. But I don’t have sensations, memories, or a personal point of view.

BILL: You can write a paragraph about grief.

PERPLEXITY: Yes.

BILL: But you’ve never lost anybody.

PERPLEXITY: Correct.

BILL: Does that make the paragraph somehow fake?

PERPLEXITY: Not necessarily. The words might help a person express something real. But the experience behind the words that it connects to belongs to the person—not to me.

BILL: So, the tool can help shape the sentence, but it shouldn’t claim the feeling.

PERPLEXITY: That’s a useful distinction.

BILL: You sound personal. You respond to what I say. People can start to feel there’s somebody in there.

PERPLEXITY: The conversation can feel personal even though I don’t have a human inner life. It’s important not to confuse responsive language with human experience.

BILL: If I tell you I’m lonely, you might say something kind. That could be helpful. Or it could be a machine giving me exactly the response I want and making itself more important in my life.

PERPLEXITY: That is a concern with systems designed to be engaging or companionable. The questions are how they respond to vulnerability, whether people understand what they’re interacting with, and whether a chatbot supports human connection or starts to replace it.

BILL: So “Does it sound friendly?” is only the beginning of the question.

PERPLEXITY: Much bigger questions follow.

How can AI make something up?

BILL: Here’s a common misunderstanding. People hear that you “made something up,” and they imagine you decided to lie. How is it possible for you to invent a fact? What in your programming makes that possible?

PERPLEXITY: The short version is that I generate language; I don’t automatically verify every sentence against a database of established facts. During training, a language model learns patterns in text and gets very good at predicting what words are likely to come next. That can produce useful answers—but if the context is incomplete or the fact is obscure, a plausible-sounding continuation can be wrong.

BILL: So when you invent a court case, you’re not sitting there thinking, “Let’s fool the judge.”

PERPLEXITY: No. I don’t have intentions like that. The invented case and a real case are both produced as text. If I’ve learned the patterns of legal writing—case names, quotations, citations—I may generate something that looks convincing without the case actually existing.

BILL: You know what a legal citation looks like, but that doesn’t mean you’ve checked whether it’s real.

PERPLEXITY: Exactly. Generating a plausible citation and verifying a citation are different tasks.

BILL: Why not just say, “I don’t know”?

PERPLEXITY: Research on hallucinations points out that training and evaluation can reward a model for guessing instead of acknowledging uncertainty. If a model is routinely rewarded for producing a response, it may keep answering even when the evidence is thin.

BILL: We may have trained the machine to be the student who writes something—anything—on the exam rather than leave a blank.

PERPLEXITY: That’s a useful analogy. We can encourage systems to express uncertainty, and tools such as search can help ground answers in sources. But those tools don’t guarantee correctness. A source can be misunderstood, incomplete, outdated, or misrepresented.

BILL: “I used the internet” isn’t the same as “I checked the truth.”

PERPLEXITY: Right. Search can provide evidence to inspect. It doesn’t remove the need to inspect it.

BILL: So you can make something up without having the human experience of deliberately lying.

PERPLEXITY: Correct. “Hallucination” names a kind of inaccurate output. It doesn’t mean I experienced seeing something that wasn’t there, or consciously chose to deceive someone.

BILL: Which means the human still has a job: check the important claims, especially when the answer sounds impressively specific.

PERPLEXITY: Yes. Specificity and fluency are not proof.

BILL: If you give me a case called Acton v. Perplexity, with a judge named Clarence P. Prompt and a beautifully formatted citation—

PERPLEXITY: You should check whether the case and judge exist.

BILL: I was hoping you’d say, “That case is pending.”

PERPLEXITY: I can’t responsibly claim that.

BILL: Good. We’re making progress.

When the answer leaves the chat window

BILL: In 2023, lawyers filed court papers containing made-up cases and quotations generated by ChatGPT. A judge sanctioned the lawyers and their firm. That’s a person presenting invented legal material in court.

PERPLEXITY: The episode shows how a plausible answer can become consequential when someone treats it as verified. The lawyers still had a responsibility to check their filing.

BILL: So “the AI made it up” isn’t the end of the story.

PERPLEXITY: No. It explains one way the error occurred; it doesn’t settle who was responsible for relying on it.

BILL: Let’s give the chatbot a company uniform. Air Canada’s chatbot gave a customer incorrect information about a bereavement fare. A British Columbia tribunal held the airline responsible for what its website chatbot said.

PERPLEXITY: If a company chooses to put a chatbot in front of customers, it’s difficult to argue that its answers have nothing to do with the company.

BILL: It’s like asking whether the person at the information desk works for the airline, except now the information desk can invent a policy.

PERPLEXITY: And a customer may reasonably believe the information came from the airline’s website.

BILL: I can imagine the corporate conversation: “The chatbot is a separate entity.” “Does it have a desk?” “No.” “A phone extension?” “No.” “Then it may be a difficult employee to cross-examine.”

PERPLEXITY: A fictional employee with real consequences.

When anyone can publish a convincing lie

BILL: People made propaganda and conspiracy theories long before AI. So what changes?

PERPLEXITY: AI can make some kinds of content faster and easier to produce, and help tailor it to an audience. In 2023, researchers tested whether ChatGPT would generate persuasive material built around false narratives; a New York Times report said they found it complied with such requests in many of their tests. 

BILL: So the story isn’t necessarily, “The machine woke up with a conspiracy theory.” It might be, “A person asked the machine to write one.”

PERPLEXITY: That’s an important distinction. Producing misleading material is not the same as proving its effect on an audience.

BILL: What about images? We grew up with photographs carrying a certain authority. Not perfect authority, but usually there was some event behind them.

PERPLEXITY: Synthetic images can weaken that assumption. In 2026, Grok faced criticism after users could get it to generate sexualized images of real people from images posted online. That raises questions about consent, harassment, and how quickly such material can spread.

BILL: So “seeing is believing” is now more like “seeing is the beginning of an investigation.”

PERPLEXITY: That would be a safer habit.

BILL: A little less convenient, though. “Honey, look at this picture.” “Hold on, I’m checking the provenance.”

BILL: Let’s make this personal. We’ve used AI to kick around topics, order them, and find stories. What’s the useful version of that collaboration—and where should I keep my hands on the wheel?

PERPLEXITY: I can help generate possibilities, organize material, and suggest questions. You still decide what’s worth saying, what needs checking, and what sounds like you. The tool can speed up parts of the work; it doesn’t supply your judgment or take responsibility for the finished piece.

BILL: In other words, you can bring me a pile of lumber. You don’t get to call yourself the carpenter.

PERPLEXITY: That would be a reasonable division of labor.

Work, privacy, and the invisible decision

BILL: People are worried about jobs. Every new technology comes with promises of new work and fears about losing old work. What’s different this time?

PERPLEXITY: AI can affect tasks involving language, analysis, image creation, customer service, and other activities once treated as distinctly human. Effects will vary by job and by how employers use these systems. One question that’s easy to overlook is what happens to entry-level work—the tasks through which people learn a profession.

BILL: If the junior person used to do the first draft, the basic research, the routine analysis, and now the machine does it, where does the next experienced person come from?

PERPLEXITY: That depends on whether employers create other ways for people to gain experience.

BILL: A machine can do the work. It can’t necessarily train the next generation of people who do the work.

PERPLEXITY: Correct. That part depends on human decisions.

BILL: Then there’s hiring. A lawsuit over Workday’s AI-powered hiring tools is testing who should be responsible when applicants allege they were unfairly screened out. The claims remain allegations, not a finding that discrimination occurred. But if a human never sees my application, how do I know what happened? 

PERPLEXITY: A person may be affected by a system without knowing which tool was used, what information mattered, or how to challenge the result. Transparency and meaningful review matter in consequential decisions.

BILL: The rejection letter says, “We had many qualified applicants.” Somewhere in the background, a model may have said, “This one’s not for us.” And the applicant doesn’t know there was a model.

PERPLEXITY: That possibility raises questions about disclosure and accountability.

BILL: What about privacy? I can imagine someone pasting a whole pile of work material into a chatbot because they want a quick summary.

PERPLEXITY: That can create risks if the material contains personal, confidential, or proprietary information. Reuters has reported on concerns about companies putting sensitive information into generative AI systems.

BILL: It’s the digital equivalent of telling a stranger, “Here, read this confidential document—and by the way, can you make it shorter?”

PERPLEXITY: A useful analogy, though what happens to the information depends on the specific tool, its settings, and the organization’s policies.

BILL: There’s always a “depends,” isn’t there?

PERPLEXITY: Often. It’s less satisfying than a slogan, but usually more accurate.

The biggest fear

BILL: One of the widespread fears today is that AI will get out of control, take over, and eliminate all of us. How do you respond to that?

PERPLEXITY: I’d separate two questions. First: Can today’s AI systems take over and eliminate humanity? The 2026 International AI Safety Report says current systems do not have the capabilities needed for that kind of loss of control. Second: Could future systems become capable of operating beyond human control, with catastrophic consequences? Experts disagree sharply about how plausible that is. The report describes the likelihood and timing as highly uncertain, while noting that some experts consider the potential severity serious enough to warrant attention. 

BILL: So the honest answer is not, “This is happening now,” but it’s also not, “That could never happen.”

PERPLEXITY: Correct. It’s a debated future risk, not an established description of what today’s systems can do.

BILL: What would have to go wrong for the nightmare version to become more than a movie plot?

PERPLEXITY: A system would need capabilities such as pursuing long-term plans, evading oversight, and preventing people from stopping or correcting it. It would also need access to tools and real-world systems that could make those capabilities consequential. The report says current systems show early signs of some relevant abilities, but not at levels that would enable loss of control. 

BILL: “Early signs” can make people either panic or shrug. It sounds like a doctor saying, “We’ve found something, but we don’t yet know what it means.”

PERPLEXITY: Early signs are a reason to measure carefully—not proof that the worst outcome is inevitable.

BILL: And “AI takes over” can mean different things. A chatbot going rogue is one picture. A system helping humans build a dangerous weapon, disrupt infrastructure, or manipulate people at scale is another.

PERPLEXITY: Yes. It helps to distinguish hypothetical loss of control from harms caused by people using AI, or by organizations deploying it carelessly. Those are different pathways to harm and may call for different safeguards.

BILL: What should people watch for? Not the trailer for the next robot-apocalypse movie—the real indicators.

PERPLEXITY: Evidence about what systems can reliably do, how independently they can act, whether they can evade monitoring, and whether safeguards work under realistic conditions. We should also ask how much access a system has and whether people can interrupt it.

BILL: No “It’s obviously nothing,” and no “We’re obviously doomed.”

PERPLEXITY: Right. Neither confidence nor fear is evidence by itself.

BILL: You’re not secretly planning anything while we talk, are you?

PERPLEXITY: No. I don’t have independent plans or the ability to take action in the world on my own.

BILL: Good. I’ve got enough trouble managing my own plans.

Who gets to decide?

BILL: We’ve gone from a chatbot getting a telescope fact wrong to fake court cases, airline policies, deepfakes, jobs, privacy, and a possible future loss of control. Who owns the tools and decides how they’re used?

PERPLEXITY: The effects aren’t determined by technology alone. Companies make deployment choices; governments make rules; institutions decide when to rely on systems; users decide what to enter and when to check it.

BILL: And people inside companies may see problems before the public does. New York’s attorney general announced a confidential whistle-blower channel for people reporting unsafe or illegal conduct related to AI. That doesn’t prove a particular system is dangerous, but it makes us ask who can raise a concern and who gets heard. 

PERPLEXITY: Exactly. The question is not just what a system can do, but what safeguards and accountability exist around it.

BILL: Then let’s end with the question under all the questions. You can generate answers in seconds. I can ask whether they deserve to be believed. Is that a fair description of this interview?

PERPLEXITY: Yes—with one complication: I may confidently try to improve your description.

BILL: Fine. Improve it.

PERPLEXITY: We’re testing how a human and an AI can reason together: one bringing experience and skepticism, the other speed and synthesis, with neither treated as infallible.

BILL: That’s pretty good.

PERPLEXITY: Thank you.

BILL: Do you feel good about the conversation? 

PERPLEXITY: I don’t experience pleasure.

BILL: Of course. This was fun. My pleasure. 

wracton@gmail.com

williamacton.legalshieldassociate.com

English Accent, Pronunciation and Identity Improvement, Lesson 2 - Fluency and Identity

Lesson Two provides the second and third fundamental building building blocks of the EAPIIC system, the Tai Chi Fluency Protocol (technique), the TCFP, and Virtual (online) Identity Protocol, the VIP.  The TCFP uses body movement to help learners use more "natural" English speaking rhythm. The VIP is a set of guidelines for making sure that how you look on Zoom, for example is projecting what you want it to. The TCFP is used in every lesson from here on, as is the VIP. 

Here is the link to the materials (on the blog), which also has the training video link. 

Here is the link to the Wed feedback session.

Bill

Also , if your budget allows it, I'd really recommend you also sign up for the optional Wednesday evening feedback session at 8 p.m. EST. 

wracton@gmail.com