Episode 8: Why Does AI Content Bother People? And How Do You Show You Care?
Host: Joe Zappa
Guests: Diane Anderson & Bill Brazell
Length: 28 minutes
Published: August 31, 2026
First-time podcast guest Diane Anderson, Bill Brazell, and Joe Zappa discuss how to differentiate AI from human content, why AI content rankles so many people, and how to produce AI content with care. Plus: Joe's viral tweet on The Odyssey that pissed off Racist Twitter, what Heidegger has to do with AI content, and who the original marketing hack was (Cyrano de Bergerac!).
Listen: Spotify | Apple Podcasts | YouTube
Full Transcript
Joe Welcome to the AdTech Marketing Podcast, the only podcast devoted to the very niche topic of adtech marketing. Today, I'm joined by Bill Brazell, a veteran of this outlet, and Diane Anderson, who is new. Diane, thank you for gracing us with your presence.
Diane Happy to be here.
Joe Diane is usually in southern California in paradise, and yet this week she told me she wants me to have her head up the European office, and I was like, "You already live in a 50% tax rate paradise. Do you need two?"
Diane Yes, I do.
Joe She's like, "Affirmative."
Bill She's like, "I'm bored with these English speakers in California. They're annoying."
Joe Speaking of things that are at best tangentially related to our work, I did a viral tweet this week. Viral by my standards. I fired this tweet off in, like, 15 seconds. I was in the middle of my morning work, so I was tearing through, and I had seen The Odyssey the night before.
So this was the tweet: "Saw The Odyssey last night. If you don't appreciate that film, you're an enemy of storytelling, all of Western civilization, and possibly the very concept of craft itself." This got 2,100 likes and 229 comments, mostly from what I would call racist Twitter— —which was extremely mad, because apparently, you know, the film casts a few Black actors, and that's very bad, I learned.
Diane Ay-yi-yi.
Joe Did you guys see The Odyssey?
Bill Loved it.
Diane I did. So they're mad about Lupita being Helen, or...?
Joe Yeah, I think that one, and Athena in particular, angered them—
Diane Oh...
Joe —because, you know, beauty needs to be embodied by a white woman, I guess. And then, like, you still got Charlize Theron as Calypso.
Bill Right.
Joe That wasn't enough for you? I saw that and I was like, "I'm good." Like, "This is all I need." I'm like, "Give me the lotus flower, okay? Ship me away."
Diane You weren't like Circe? You didn't want to go to the shack?
Joe I was not trying to get turned into a pig, honestly. You know what? Bill is extremely erudite. I have a question for you, Bill. So there was a hilarious interaction on the Mentor thing. I think—did you send me this? Where someone...No, my friend Sava sent it to me.
So there is a character in The Odyssey, in the book and in the film, called Mentor, and this is where we get the word mentor, meaning, you know, someone who coaches you or whatever. And there was another tweet from imbeciles on Twitter that was like, "Christopher Nolan: 'I'll have a character who's a mentor and just call him Mentor.'"
Bill Oh, that's funny.
Joe Christopher Nolan had been, like, lazy in choosing this name for the character.
Bill Oh... That's hilarious. What's better than somebody criticizing something on historical grounds and not knowing the history?
Diane Those are the people who didn't like your tweet.
Joe And then I wondered, so the "men are pigs" thing—like the idiom, "men are pigs."
Bill Oh, they're pigs in Homer.
Bill They are pigs. Yeah. She does turn them into pigs. Uh, she is a more seductive character in Homer, which is why some parents name their children Circe, but they wouldn't after that movie. If that was the only thing they knew about Circe, they wouldn't pick that name, I don't think.
Joe Well, Cersei was also the very alluring, dastardly queen in, uh, Game of Thrones. Oh, yeah. So they took it from that, I'm sure.
Diane But also...she gives him the way home, even though she turns them into pigs when they descend on her alone. Once she finds a good man, she gives him the map to get home safely. So Circe's not all bad, people. Come on now.
Bill Fair enough.
Joe I actually liked that ambiguity, and maybe we can start finding our way into our topic today, which is: how do you separate what's AI from what's human? Like, one thing we could agree on is that human content, when well done, is often very good at acknowledging nuances and understanding sort of the subtleties of tone.
Diane Right. Mm.
Joe I thought the way in the movie they portrayed Circe was very interesting, because she doesn't have an explicitly sexual relationship with Odysseus.
Bill Right.
Joe But there's, like, an air of an understanding between the film characters that, while not sexual, is completely distinct from the relationship she has with all the other men in Odysseus' crew. And I feel like there's an undercurrent of sort of erotic tension, but it also underscores that Odysseus is kind of like this... He's a warrior, but he's also smart, and he's got a little undercurrent of this sensitive guy thing to him.
Bill I think that's right. And actually, I think this Circe scene is a good transition to the AI issue, uh, because of one line which my wife has repeated about three times since the movie—I should probably ask her why. Uh, when she says, "Go back into your disguises," when she says to the pigs, "Go back into your disguise. Be human again, hiding who you really are." It's a great line. She kind of whispers it, and it's very striking, and I feel like that's part of what we're dealing with when we're talking about AI versus human is: is there a disguise or is this a real human? You know, what is this that we're dealing with? Can we trust the words that we're reading? Can we trust the images?
Diane Oh, we're taking the AI, we're taking the little piglets, and we're dressing them in their human costumes and putting them out into the world.
Bill "Go back into the disguise from which you emerged." 'Cause a lot—I mean, that's the other thing about AI that's tricky, right? All of the language in AI came from us.
Diane Us. Exactly. That's what's funny.
Bill From humans, and now we resent that it's been processed and served back to us because it feels not quite human anymore.
Diane One of the things I've been thinking about is, like, it used to show... Like, you guys, we're writers. We're the original ghostwriters. We're the original fakes, right? We used to write things and put it in other people's names, and one of the signs of time and attention was polish. You made it nice. Now—When something's too perfect—Yeah... that's the signifier. It's like, oh, that's too clean, too neat, too perfect.
Joe Yeah. It's true that it's actually the inverse. Like something that is too polished now appears to have taken a lack of work.
Diane Right.
Joe And so this is such an interesting tension I constantly find myself brushing up against. You know, Joe Gallo, the director of communications at PayPal Ads, works across different parts of their business, and he posted on LinkedIn about using AI for pitches. He was pushing back against reporters who have said, "You know, you cannot use AI to pitch me." And he was kind of saying, "Well, why not?
Diane Don't you want to have a better pitch?"
Joe Yeah, because that's the thing: we all know—and this is, I think, where AI shines the most—what AI is extremely good at is taking all the granularities of a body of work, right? Like a corpus, be it whatever the client is working on or what a reporter has covered in the past, and turning that into a very granular, detailed pitch. So for example, let's say you wanted to pitch Patrick Coffee at The Wall Street Journal. You could use AI, especially if you're using the stronger versions, to understand everything he's written over the past year, and then say, "How does this relate to my company?" And then craft a pitch that is very tailored to what he has historically written, and therefore what he might be interested in, and that is also vetted against the rigor of what The Wall Street Journal covers to make sure it's actually worthy of their time. And that pitch, if AI is involved, is probably going to be better than whatever a human publicist comes up with.
Bill Right.
Joe And human publicists were doing horrible pitches way before the existence of AI. You know, we were all journalists, so we all received that at one point in our lives.
Bill Yes.
Joe Yeah. And so I think there really is a case for the virtues of AI and its ability to do that granular work and to pull in all that context. But as you said though, Joe, the level of polish is potentially actually an issue.
Bill I think also what the journalists really mean when they say, "Don't give me AI pitches," is, "Don't give me a pitch that I can tell was written by AI."
Joe Yeah.
Bill Like, I'm not using an AI detector, but if it comes in and it sounds like every other pitch, then I know it wasn't a human who wrote it. And that's, you know, the journalists, like the rest of us, are getting used to... We have our own internal AI detector now, and we feel insulted if we can tell it's AI. But I don't mind if I'm not sure.
Diane No. Wasn't it Joe Gallo, too, who had the typo—like there was an extra period or something, and he was like, "Oh, I was going to take it out, but then I left it in"? Maybe it was someone else, but it was like he wanted you to know that that was him.
Joe Well, I think what you said is so fascinating, Diane, because essentially—and Bill was just doubling down on it—the objection to AI is, "Did you put no work into this?" And you're right, it's the exact inverse. It used to be like, "Oh, there are typos. You put no work into this." Now it's like, "There are zero typos. It seems too good. You put zero work into this. You just went into AI and generated it."
So really the objection to AI content is in large part: are you putting effort into the thing that you are then demanding my attention to read? Like, you wrote something—that better have taken some human effort, because it's going to take human effort from me to read it. And that's where we get into all the usual stuff of: the real answer is you have to give it the proper context, you have to prompt it, you need to edit, you need to iterate on it. And if that's the case and it's really a quality product, then it really shouldn't be a problem.
Diane So I think what we're wanting from each other, both as readers and as writers and as consumers of content, is we want to know there was a human involved at some point. And we will be okay with mistakes and flawed thinking or not perfect polish. But we want it to be good, and we want to know the level of thinking or cognition that's behind it. At some point, whether it's in the initial ideation or the final pass, at some point we want to know that there's a person behind it, and a personality.
Bill I think it's interesting, like if you read an AI-generated piece and you don't realize it's AI, was anyone harmed? Like, what is the harm? I know how I feel when I see a piece that I'm certain was AI and no editing went into it.
Diane You feel offended. You're like, "Ugh, come on."
Bill I do. I get annoyed and I stop reading. Now—
Diane "It's not X, but Y," or—
Bill It's entirely possible that sometimes I'm wrong, that I'm reading something that wasn't AI-generated, but I'm sure it was because I have seen so many. Because—I forget who... Oh, Jesse Hempel recently wrote this on LinkedIn, that she has started to see in her writing some AI tics because she's read them so often. Which reminded me, I was talking to a schoolteacher years ago in New Mexico who told me that he had started making specific misspellings of words that he had never misspelled before, but because his students were routinely misspelling them in this way, he just started doing it. And that's kind of how language works, right? We're all learning from each other and not always asking...
Diane We're reading so much AI that we're starting to think like AI, right?
Bill That was what Jesse was saying, and I thought, "That's probably true." And there probably are more people writing things like, "It's not X, it's Y," because that starts to feel like good writing.
Diane So this takes me to two different directions. One is the first thing you were saying: was anyone harmed? So like Shy Girl, right? Do you know about Shy Girl?
Bill Yes.
Diane Yeah.
Diane So there's this book, and I'm dying to get my hands on it. Do you guys want to know what day my birthday is? No, I'm just kidding. Um—
Bill I know what day your birthday is. Um, it's October 28th.
Diane That's right.
Bill And Shy Girl, I think, is hard to find right now because—
Diane Exactly! That's why I'm saying. This would be a book for us to have. Yeah. So it was like this bestseller, and was it Hachette? Who was the publisher? Anyway, then this author got accused of using AI, and she's like, "No, one of the editors I worked with must have used it at the end," blah, blah, blah. Now it's taken off shelves. Like, if you want to read it, you can't.
Bill Right.
Diane I want to see what it's like, you know? I'm curious.
Bill It was a pile-on on Twitter because there were all these people taking out passages from the novel and saying, "No way. This is clearly AI." And there were so many of them that the publisher realized, "We can't publish this." Which is fascinating, 'cause why not publish it? Why not just say, "We don't know"? You know, it's funny 'cause publishers are very adamant. There was just a book—it was a memoir about being abused by a high school teacher in Texas—and it turns out there is reason to believe that some of the stories in her book actually came from a different student and were things that did not happen to her. So The New York Times did this big investigative story, and they went to the publisher, and the publisher said, "We don't fact check. So we accept that the writer says that this is what they wrote."
DianeRunning with Scissors...
Bill ...stands forever because it protects them legally from this kind of thing. That happened with the guy who wrote A Million Little Pieces, James Frey. Similar thing. Publishers say, "We don't fact check. Not our job. We accept the author." But they didn't do that with Shy Girl, Diane, and that was interesting. It's a different thing. A publisher could just as easily have said, "Hey, we don't know if it's AI or not. It's not our job. We're still putting it out," because suddenly so many people knew about it that many, like you, want to buy it.
Diane Well, it's fear of the bots. They're like, "Oh, no. Human good, AI bad," which I think is going to change. I mean, we're going to have to be more nuanced about that.
And the other thing, what you were saying my brain went to is that, yes, we're starting... Like the teacher you were saying started to write or think like his students with the same misspellings. I think that now there's kind of almost going to be... The reason I brought up Joe originally was someone we work with, Thalia, was talking about how a client objected to a certain AI construct. Verbally expressed they didn't like the aesthetic of it and didn't want it in their writing, and then in the same meeting spoke that tic.
Bill Mm-hmm.
Diane So it's almost like AI writing has become a prestige dialect, right? We all code-switch, right? Like, we'll talk and we'll drop F-bombs when we're talking to each other, but we might not talk that way with our families or with clients. So they might not want it in their LinkedIn content or their bylines, but they'll say it in a meeting 'cause it's just so much the way we think and speak now.
Joe It's like a verbal shortcut, right? It is a sort of idiom that's infecting the way we communicate and that is actually sort of disrupting the process of communication, appearing as a shorthand in actual conversation.
I thought about something as you guys were talking—something deep back from grad school. The German philosopher Martin Heidegger had this piece, Was Heißt Denken? ("What Is Called Thinking?"), and his core argument about thinking was that it's not merely computation, right? And so this would be particularly relevant to the distinction between human thinking and the thinking that might be ascribed to artificial intelligence: whereas AI is more computational, human thinking is precisely about what Heidegger calls dwelling.
He uses this verb wohnen in German, which means to live or to dwell. So, for example, this is a quote from him: "The manner in which you are and I am, the manner in which we humans are on the earth is dwelling or is living," but living in this particular way that emphasizes an instantiated being. And so this really ties to what you guys were saying about differentiating human from AI content, because essentially what we are asking when we're asking if something is human or AI-generated is: did you dwell with this? Was this produced instantaneously as a result of a form of thinking that's reducible to computation, or did you sit with this? Did the thoughts percolate? Did they evolve? And so it strikes me that this concept of dwelling is very relevant to the AI versus human classification.
Bill Was it instantaneous, or was it instantiated?
Joe It was instantiated instantaneously. Please keep talking.
Bill Yeah, exactly.
Diane No, I was just thinking, when you said "sit with it," one of the programs we use in our work is Asana, which means "seat," right? We're having to sit with these tasks. We're having to put human thought into it.
Joe Oh, okay. Mm, yeah.
Diane When you're doing a yoga pose, an asana is a position, but it means "seat."
Bill Pronouncing it asana probably wrongly, it turns out.
Diane Yeah, asana. You can say it either way.
Bill It comes from yoga.I didn't know that. That's interesting. My favorite Heidegger expression is the way that we are "thrown into the world." There's sort of this expression like "we are thrown" as a way of saying we find ourselves in this world that we suddenly have to cope with. And I think Jim Morrison picked that up for "Into this world we're thrown" in "Riders on the Storm." So there we go. Who knew we'd be talking about The Doors?
Joe I think what's so interesting about it is that another key Heideggerian concept is care, and care is sort of the fundamental relation of beings to one another, right? As humans, we are involved in the exercise of care, which might be distinguished from that of other beings (or you might problematize that hierarchy of beings). For those who don't know, Heidegger, unfortunately, was a horrible person on top of being a very interesting philosopher.
I think what is happening really is a question of care, right? And a question of time and respect. When we have an aversion to AI-generated content, it really is all about that. It's like: did you put any care into this? And by extension, are you expressing care toward me, or are you being careless in generating these things?
Diane I think our executives... like, AI will just throw anything—the kitchen sink—in. I think the human filter, we know, "Oh, my CEO would never say that." Like, we know what they wouldn't say, and that's where the care comes in.
Bill I think that's fascinating. There was a great thread the other day where this woman said that she had been chatting with a guy she met online through Tinder or whatever, and when she met him in person, he was not nearly as articulate. She realized that he had been using AI to answer her questions, and somebody said, "You got chat-fished." But it was also the question of: why is it bothersome? What is it about that? And like you said, it implies the guy was careless. He was not putting any effort into it.
When I read that, I showed my kids the movie Roxanne, the Steve Martin and Daryl Hannah update of Cyrano de Bergerac. Very charming, in which Steve Martin gives language to a very good-looking man who's totally inarticulate, and seduces her by proxy, and she's very upset. She's very upset because she thought this man had these thoughts—that these were real—but in fact, they all came from Steve Martin, the volunteer fire chief. I think that's really fascinating: the feeling of betrayal. I wonder why we feel it. When I read an AI piece and I have an aversion to it and feel revulsion, I'm like, "Oh, I'm just annoyed. Why did they just put this up here?" I guess part of it is care, and part of it is you're not telling me anything that you really believe or feel. Even though if I talk to the person, they would probably say, "No, that is what I think. It's just the AI organized it better than I would have."
Joe It's a mixture of time and authenticity. Also, Cyrano catching strays—the original marketing hack, producing counterfeit copy. But yeah, I think that's what it is. It's like: did you take the time to do this, and do you care enough about me to use your own brain and your own soul to create something for me or that is of me?
Diane We all just want to be seen.
Bill We really do.
Joe I also want you to know that in the latest Drake album, he says, "You write me heartfelt words, I find out chat's behind them."
Diane Well, chat can also mean fam, so it could be other humans creating content.
Joe Hey, I don't know how you're up on that. I don't know and I don't want to know. But yeah, once Drake sang it, it has gotten deep into the cultural bloodstream. Yes, the hoi polloi are on this, okay?
Bill For sure they are. I like Diane's... I don't know, did you coin that, Diane? That it becomes a prestige dialect to sound like AI.
Diane Well, I made ChatGPT come up with that for me.
Bill Did you really?
Diane No
Bill No, That's funny. That's the grad student Diane Anderson there.
Joe We all let our inner grad student out. That's exactly why I started this podcast: so that the three people who are still listening enjoy these allusions,
Diane as are all the rest of our friends.
Bill That is hilarious.
Diane No, chat also means when you have a group chat thread.
Joe Yeah I know that, Bill doesn’t
Bill I didn't know that use of it. So what he's saying in the song is that she's giving these words that may have been pre-approved by her friends in a chat.
Joe Right, you're right. It's polysemous, if you will.
Bill Wow
Joe Bill's like, "All right, even this is too much for me. Shut the f*** up." "Go back into the hole from which you came. Put this back into the recesses of your brain."
Diane Because you admitted you didn't know what an obelisk was, Joe, I feel comfortable asking you: what is polysemous? What is that?
Joe Oh, it just comes from the word polysemy, obviously from Greek origin—poly meaning multiple, semy meaning meaning. Polysemous is just the adjective version of polysemy, yeah.
Bill So it's a perfect PhD word, isn't it?
Joe It is. Drake drops many polysemous gems, but he doesn't get credit for that.
Diane Was that your biggest tweet ever, Joe?
Joe Oh yeah, by far. Are you kidding me? My average tweet gets, like, four likes. So when I write about the work that pays for us all to be here, three people like it. But when I write an intentionally hyperbolic phrase about The Odyssey, 2,000 racists come out of the woodwork.
Diane Well, are there any blockbusters coming out when we have our off-site? That's my question.
Bill I was thinking what you could do, Joe, is combine those two things, right? Work in an adtech reference while talking about how great The Odyssey is, and then you might bring in a whole new audience.
Joe Did we do anything but that on this podcast? This podcast is merely the representation of that idea. But to tie a bow on it, I think we are crystallizing something that relates to the lowly work of adtech marketing somewhere between Heidegger and Drake, which is that AI is introducing a new perception around an age-old question in service businesses: how much care did you put into this?
And that is the fundamental question: does the work, both in its granularity and accuracy, but also in its humanity, represent the care that the person on the other side of the work is due, whether that person is a reporter, a client, or a colleague? And so I think that is what we are in essence striving for. We're obviously not going to be Luddites and eschew AI. But as we use it, we must exercise and demonstrate proper care. And it's when you fail to meet that standard that the gap between what's AI and what's human becomes really objectionable. Is that a fair assessment?
Bill I think it is. And as you spoke, it reminded me: I mentioned William F. Buckley just before we started recording, and in his memoir, Chris Buckley told a story of being with his dad as his dad was typing up a column in the back of a taxicab, and he was finished with the column in about twenty minutes. He told his son, "Don't tell anyone that it took me twenty minutes to write that column." And Chris kind of teased his dad about it, and his dad said, "Look, it looks like twenty minutes, but it's actually my entire life that makes it possible for me to do that in twenty minutes."
Joe It's funny, I used to do that when I was starting out as a freelancer ghostwriting in our industry. I was like, "I'm not a better writer than either of you or anyone on our team, but I'm a very fast writer." And so in the age before AI, I used to be able to hop on a call with a CEO and use my human brain to write a byline in forty-five minutes. And sometimes I would just send it, but I had to know who the interlocutor was: did they trust that when I was sending a piece in forty-five minutes, it wasn't 'cause I was doing sloppy work? The work was actually very polished, but that's just a particular aptitude I have. And again, it's coming from—as all of us did—an unfathomable number of hours of writing. I had written so many arguments in my life that all of those arguments are with me when I hop on a phone with someone and write an 800-word piece in forty-five minutes from scratch.
Diane You would wait before you would send it, just so they would think...
Joe I would sometimes, yeah. It depended on the person. Sometimes I would flip it in an hour, and I would say, "We met," and then they would say, "Oh my God, this is great." And sometimes I would wait until the next morning because I didn't want them to perceive the speed as a lack of care.
Bill Exactly right. And I'm sure there are a large number of people who delay sending AI-generated responses for the exact same reason. They could send the response two minutes after they got it, but if they do, it'll be obvious that they didn't write it. So they generate it, leave the window open, and then send it in twenty minutes or an hour.
Joe And it's true that it takes some degree of time to context load, prompt, and edit, which are the three things you have to do if you want to use AI carefully to produce anything. You shouldn't instantiate it instantaneously, Bill. Um, especially if it's polysemous.
Bill I wouldn't dare.
Joe One must discern carefully over the period of at least forty-five minutes. But yeah, I learned something from you guys, and that's beautiful. I'm glad we could walk this labyrinth of illusions and arrive at the understanding that we must practice care when using AI in our work.
Bill And Joe—well, when I post this, I hope that you have "Riders on the Storm" and "Into This World We're Thrown" as the outro music.
Joe Absolutely.
Diane It's going to be great.
Joe No, it's much more Drake-esque. It's some hip-hop type vibes. So with that said, Diane, Bill, thank you, and Mike, cue the beat.
Want more insights on adtech marketing?
Subscribe to our newsletter and get every new podcast episode delivered to your inbox.
