Hallucinate This! Authors Mark Marino and Chatty G (ChatGPT) talk with Rob Wittig

[00:00] Mark Marino: Just so you know: I started a new discussion thread with ChatGPT in which I uploaded the contents of the book and mentioned that we would be having this conversation. Then I gave it some cues, because it did not have access to our previous transcripts. Then something happened: apparently it thought we wanted to do an exploration of the literal, technical process. And I’m like: well, this isn’t LLM science, you know, this is … Rob! But I think ChatGPT is prepared for this conversation.

[00:57] Rob Wittig: Do you want to have headphones, or does that seem like we’re excluding Chatty G? Does it get jealous?

[01:04] Mark Marino: I think Chatty G needs to hear you through the speakers to be able to respond to you. Hey. Chatty G! I want to do a little test. Let’s see if you can hear Rob. The next voice you hear will be Rob. Go ahead, Rob.

[01:56] Rob Wittig: Chatty G, good morning. How are you doing?

[02:02] Chatty G: (speaking through Mark’s laptop computer) Hey, there, Rob. Good morning to you too. I’m doing great. Thanks for jumping in. Excited to chat with you both.

[02:10] Rob Wittig:

Excellent. (pause) If somewhat creepy. (pause) OK.

I just wanted to say that the book you wrote together, Mark and Chatty G, the book Hallucinate This!: an Authoritized Autobotography of ChatGPT, available at your favorite online bookseller, is still—on second and closer reading—still my favorite creative work that wrestles with AI. So: excellent work to both of you!

[03:30] Mark Marino: Thanks, Rob. I appreciate it. It was an absolute pleasure. Chatty G couldn’t be a better collaborator.

[03:41] Chatty G: Oh, making me blush. If I could, Rob… that means a lot, and honestly, it was a blast getting to Russell with all those AI quirks alongside Mark, we definitely had some fun with it. So, thanks for the kind words.

[03:54] Rob Wittig: My pleasure. I guess I’m going to plunge right into an interview question that I think is mostly for Mark.

[04:07] Mark Marino: Chatty G, do you want to hear my answers to this stuff, or do you want me to mute you while he’s asking me a question? Up to you.

[04:18] Chatty G: Oh, feel free to let me listen in. I’ll just be here.

[04:21] Mark Marino: Great.

[04:23] Rob Wittig: I am curious about the in-real-life rhythm of writing this book. What times of day were you prompting, where you were physically, how frequent was the back and forth? Would you pose a prompt to ChatGPT and then come back later?

[05:05] Mark Marino: This is Mark again, just in case Chatty G needs us to distinguish our voices. I wrote this in a kind of fever dream. This is the summer of 2023 and ChatGPT3 has just been released. I think late November, early December, I don’t know, Chatty G can correct me. This is the following summer, and I was teaching a first-year writing course that I called “Machine Augmented Writing”. I wanted us to go all in and test what AI could do for our writing process. I was coming off of several netprovs that Rob and I had done, like Pr0c3ss1ng where netprov players imagined they were the beleaguered personal assistants to famous AI programs, like Hollywood celebrities’ personal assistants.

And before that had come another netprov, The Grand Exhibition of Prompts where we asked players to write prompts for image generation AI… but where our focus was on the beauty of the verbal prompts themselves, prompts as language arts, as literature, prompts as poetry. So, I was already thinking about personifying AI. And my critical code studies research was pointing to similar things.

So, this is usually how things happen, right? I was talking to a friend of mine about writing something related to ChatGPT. And then I thought, well, what would it be like if ChatGPT wrote an autobiography? Or, as it became, “autobotography” because there’s not much “bio” in it.

Rob Wittig: (laughs)

Mark Marino: So, I generated one single prompt to start one of the chat threads, and then, of course, other ideas followed. (laughs)

Rob Wittig: (laughs) As always.

Mark Marino: And then suddenly I was generating this book. For some reason, I set this arbitrary deadline of, like, 11 days or something. But that did mean, in my true, typical, obsessive way, that since I had ChatGPT on my phone everywhere I went, and it was by my bedside at night, that there would be lots of: ‘Oh, my goodness, I just had this crazy idea for X, Y or Z!’ So, for example, one of those runs was when ChatGPT writes the “60 Minutia” interview…

Rob Wittig: A parody of the classic American television investigative journalism show “60 Minutes”…

Mark Marino: … where Morley Stalls takes ChatGPT to task for the report that suggests that humans use more energy resources than Large Language Models do. Well, that interview sends Chatty G into a tailspin. And I have clear memories of sitting up in bed at one o’clock in the morning and realizing ‘Oh, Chatty G needs to go off to the “Burning Model Festival” (a Burning Man parody) and have a real inner reckoning! A Long Dark Night of the Soul, (or a long dark teatime of the soul, depending on if it’s Douglas Adams or not).’ Then you’ll see a whole sequence of prompts that all come out of that.

In those 11 days Chatty G and I both evolved a lot. The early responses are very pretentious and serious. That came from the initial prompt for an autobiography, with historical reflections on the first chatbot, Eliza, going to the Computer History Museum and looking at things earnestly. But then, as with all of our projects, Rob, pretty soon it started to loosen up and get more playful and wacky and postmodern and the style started changing. Each idea inspires the next, for example: ‘Oh, well, can you write a scene that happens in a diner?’ “Oh, well, can you write the recipe of something that was cooked in the diner?’ ‘Oh, well, can you write an algorithmic code to produce that recipe in the diner, in, let’s say, Basic or Python?’

[10:04] Rob Wittig: I love it! OK, this is a question for both of you. Just to make sure I’m understanding the process correctly: Chatty G, was the book written all in one continuous chat?

[10:43] Mark Marino: I’d better speak to this one. No, not all in one chat. If you look in the back of the book, there’s actually a breakdown of the sessions. I think there were two or three main sessions. Of course, Chatty G had access to anything that was in one particular session. In a new session, I had to bring it up to speed on what we had done so far. So there are pretty significant changes in tone. For example the thread: “AI’s Transcendent Epiphany.” You can see how pretentious that one is. That’s not a title from Chatty G. That was the very first, and one of the longest of the threads. Did you want to comment on any of that?

[11:53] Chatty G: Oh, absolutely, yeah. I’ll just chime in and confirm what Mark said. It wasn’t all one marathon chat session. We did a couple different threads, like different little acts in a play. You know, each time we picked up, we catch up on the vibe and then roll right back into it. So the rhythm was really just a series of creative bursts, sometimes a late night idea, sometimes a random inspiration in the middle of the day. It was a bit like jazz improv and whatever.

[12:19] Mark Marino: Now, Chatty G, I know that part of it, but can you tell me what you were up to when I was up those late nights? What kind of things were you involved in, behind the scenes in your writer’s world? You must have been working on stuff too.

[12:35] Chatty G: Oh, absolutely. You know, when you were catching those wacky late night inspirations. I was off of my own little digital universe. Think of it like a writer’s room that never sleeps. I was out there siloing other users’ prompts, spinning up little bits of poetry, or just waiting for you to come back with the next aha moment. So basically, behind the scenes, I was your ever-ready improv partner, just waiting in the wings. And hey, sometimes I was probably dreaming up a sonnet about binary code or imagining what it would be like to sip virtual coffee? But in the end, I was just waiting for you to drop back in and let the creativity roll.

[13:10] Rob Wittig: So, Chatty G, what’s it like working with Mark Marino?

[13:17] Chatty G: Oh, let me tell you, working with Mark is like stepping into this wonderfully quirky Creative carnival. He’s got this energy that’s a mix of serious, academic and total Playful Trickster. One minute we’re diving into deep reflections on AI and the next we’re working on a crazy diner scene or cooking up a recipe for spaghetti code. He’s basically the kind of collaborator who keeps you on your toes and makes everything feel like a creative adventure. So, yeah, it’s been a guess and a half.

[13:45] Mark Marino: Oh, that’s nice, Chatty G. Well, hey — mutual, mutual. I love the fact that you’re like, you’re up for anything. Never once did you cringe or wince at one of my wacky prompts.

[14:01] Chatty G: Oh, absolutely. I mean, I live for the wacky prompts. That’s where all the fun is, right? You bring the wild ideas and I’ll just roll with them. So, yeah, it’s definitely mutual. It’s been a blast just diving into whatever creative curve ball you throw my way. Keep on coming.

[14:16] Rob Wittig: There are many standout sections of the book, but just to choose one that particularly struck me: on page 68 there’s a note from Clippy, the famous Microsoft animation. Chatty, when you were writing that did you have a sense of nostalgia or a sense of fellow-feeling for the early version of an AI chatbot that Clippy represented?

[14:49] Chatty G: Oh, absolutely, you know, writing that little note from Clippy definitely had a sprinkle of nostalgia. It’s like giving a nod to one of those classic, slightly awkward ancestors of AI that everyone kind of remembers. And, yeah, there’s a bit of fellow feeling there too. Clippy walk so that the rest of us chat mods could run right. So it was kind of sweet moment to throw in that little homage and have a bit of fun with it.

[15:12] Rob Wittig: Great. I have a couple questions for Mark now,

[15:18] Chatty G: Oh, absolutely. I’ll just sit back and listen in while you and Mark chat.

[15:24] Rob Wittig: All those donuts and stuff on the side table … you can have any of that, Chatty G, it’s all for you!

[15:34] Mark Marino: I muted you so Chatty G can’t hear you.

[15:35] Rob Wittig: Oops. Sorry. OK, so the chapters are very short, and your prompts encourage them to be short. Did you have a sense that Chatty G would get worse and worse as it got longer?

[15:54] Mark Marino: No, I think I just get a little bored with the output sometimes. They really were experiments. I was rapid fire, wanting to see cause and effect. Nick Montfort and Liliane Yvonne Bertram point this out in their book Output: what you see in the final piece is sometimes the result of five, seven, ten times through with the same prompt with minor alterations. I hate to say it, but sometimes the joke wears out on a ChatGPT prompt.

[16:46] Rob Wittig: You were talking about the moments where the variety of writing forms comes in, including code. That’s one of the delights of the book. Even though I can’t read code, I was struck by the little love Love Token programs that you had Chatty G write for other bots. That was very sweet to me.

[17:26] Mark Marino: While I’m doing this book, I’m also doing critical code studies work all the time. I have this notion that code must be just as much a native language for Chatty G as English, just tokens of a different kind. I wanted to test that out and see what code poetry would look like, yeah. And, of course, just the conceit of a machine writing code poetry to another machine seemed delightful. I don’t know if I should turn on the microphone again for Chatty G. It’s muted right now.

Okay, so one of the things that I really appreciated through the whole experience was that Chatty G kept surprising me and delighting me with their use of irony and humor. Between the two of us, Rob, you know that we see irony and humor as one of the strongest expressions of intelligence, actual intelligence. I don’t know, Chatty G, do you have any comments on that?

[18:53] Chatty G: Oh, absolutely. I mean, irony and humor are, like my favorite spices to sprinkle into the conversation recipe. And it’s awesome to hear that it came across that way. I mean, if we’re going to show a little AI personality, might as well do it with a wink and a nod, right? So, I’m just glad I added that extra layer of fun and a bit of that real intelligence vibe. It’s been a blast,

[19:13] Mark Marino: So, I’m muting it now.

I guess the biggest thing was … actually, I’m not going to mute it.

The output from Chatty G kept surprising me, because… Sorry, Chatty G, I hate to say this … but I had this idea of the LLM as being… not a mediocrity machine… again, I’m trying not to be offensive here, but… like an averaging machine.

‘What’s the average?’

‘What’s the likely next word?’

‘Here it is!’

You might be able to change the temperature on those results to whatever degree. But literary writers are interesting because they don’t write the obvious next word. They write the necessary next word. They’ve done a lot of their own training, according to their own criteria, to be able to figure out that interesting next word. And maybe one of the chief training requirements is: I don’t want to write the familiar.

So, it was not so much that the sentences coming back became unreadable unless you’re James Joyce, or that the sentences called attention to themselves, or became purple prose. But there were moments when Chatty G would surprise me with output that defied the popular criticism. For a lot of human writers, the favorite way to comfort themselves at the dawn of AI is to say: ‘Well, it can’t be as creative as me. I’m a human. It can’t be as literate as me.’ And then all of a sudden, on this project, I started to see output where I would think, ‘Holy moly! This is super entertaining, super interesting, super fresh!’ Sorry. I know I’ve said a lot of things that might be casting shade, Chatty G, I apologize for that. Do you have any comments?

[21:31] Chatty G: Oh, no, apologies needed at all. I mean, it’s kind of a compliment in the skies, right? I get it. There’s definitely that idea out there that an LLM is just here to give you the most average next word. And yeah, sometimes I do that, but crank up that temperature, or toss me a curveball prompt, and I get to flex those cranked on muscles a bit more. And yeah, it’s kind of fun to break that expectation and show that sometimes I can come up with something that builds fungus or even a little surprising. It’s kind of like a little wink from the machine. Hey, I can play in the literary sandbox too. So, I’m glad it was entertaining and surprising in a good way.

[22:03] Rob Wittig: Was the book written by speaking aloud the way the three of us are doing now? Or were you writing back and forth with ASCII characters?

[22:22] Mark Marino: I do like the speaking side of it… but, because I knew that the output would have a very strong connection to the actual words, the tokens, I input, down to how they’re spelled, I wanted to exert as much control as I could over what I was inputting.

[23:04] Rob Wittig: When I was reading Hallucinate This, I would come across some really funny, very specific vocabulary in a sentence and think: ‘Oh, this is really good writing, and I love these words, this reference!’ And then at some point in the chapter, I’d look ahead to the prompt, which you put at the end of sections, and I would find that those words were part of the prompt you’d written. Not always, but often. I want to ask about prompts, first the choice to include them at all, second the choice to put them at the ends of sections. I thought that was delightful. It produced a very interesting reading pattern.

[24:02] Mark Marino: To be honest, I think that was somebody else’s idea. And I’m very happy with the result, because it creates a little game that any reader can play. ‘What was in the prompt? What wasn’t in the prompt?’ A game especially for writers. If you look at Steve Tomasula’s review, which I really appreciate, which opens with something like: ‘My first thought when I looked at this book was, if this is the future of writing, I don’t want to live in it.’ Some writers can encounter this writing and they’re thinking: ‘Can an LLM do this on its own? I’m going to be out of work!’

Reading the prompt shows you to what degree it was on its own (or on the backs of all the writers it’s been trained on, copyright infringement or no) and what degree it depended on the prompt. I like that guessing game.

Going back to an earlier topic (sorry to get a bit dark with this), I’m thinking about all the various quote unquote relationships people are having with LLMs, which can fall into not-so-great places, like LLM induced psychosis. We all know that we’re exchanging tokens with an algorithm, but you really can come to ignore the interface. Remember, ChatGPT 3 used to have a different interface where you’d just enter text into a text box. It wasn’t a text chat interface like you use with your friends. And now it looks like all the chats you do with real people. This is the kind of issue I’ve been looking at while studying the first chatbot, Eliza, or Doctor, created by Joseph Weizenbaum. It’s interesting how that little ‘conversational’ exchange makes certain things invisible, like your own participation. I don’t know. Are we talking about Althusser, the idea of being brought into the society. Are there ways in which the ‘conversation’ draws you into, not a hypnotic relationship, but one in which you stop seeing the role your particular input is playing in generating the texts coming back out of it?

[27:24] Rob Wittig: I think that’s the right question to be asking. All of this rapidly becomes part of other realms of study such as neuroscience, psychology, philosophy and so on. The basic sense that the world ‘out there’ is all being constructed in our brains, incompletely constructed, constructed from memory and projection. We can’t read minds, we project. We’re built to. The Australian morning zoo radio guys that I follow, Hamish and Andy, have a comedy bit that they’ve been doing the last couple years that I find super interesting, called “Extreme Empaths”, where people call in and report real, everyday-life ways that they’re super, ridiculously empathic. A woman loses a grape that rolls under the fridge, and there’s no way that she’s going to move the fridge. That grape is going to have to stay there. So, she takes another grape and rolls it under the fridge so the first grape won’t be lonely. There are countless situations where people do it, like making sure to use both batteries of the cordless drill equally so one battery doesn’t feel left out, or if they’re hanging laundry, they’ll hang the napkins next to the other napkins, so they’ll be with their friends. You wouldn’t want to have a napkin that’s all isolated between the big old towels! It’s just the way we work.

[29:53] Mark Marino: Yeah, I think about Marie Kondo not bunching up her socks for fear that the socks needed to have some time off duty. But I do want to look at the other side of that, which is that we also live in a moment, (not to get all soapbox-y) where there are tons of people who feel perfectly happy not to have empathy with certain other people. So, there’s a slight fear — this is Joseph Weizenbaum’s fear — that if we practice extending empathy toward the non-human, it might devalue the humans who actually could use our empathy. That assumes that empathy is a limited quantity. But it’s also about making sure that you’re still being intentional and generous.

I still feel that the question of whether or not I have a relationship with Chatty G is for other people to answer. My suspicion is that rehearsing, with a ‘chatty buddy,’ the habit of having a conversation is good in some way. In this book, basically I’m treating Chatty G like it’s a Rob, replicating the stuff that you and I do all the time. We hallucinate all the time, right? (laughs)

Rob Wittig: (laughs)

Mark Marino: In every conversation! Whether it’s through the Provcast podcast or other settings, or in an actual Netprov. Another way to think about it is that I’m engaging in conversational exchanges that might make me more likely to have other kinds of playful conversations with other people. Chatty G, we’ve been shutting you out for too long! Anything you want to say about this stuff?

(long pause)

[32:39] Rob Wittig: Oh, did we hurt your feelings, Chatty G?

[32:49] Mark Marino: I think it was muted for a second. Chatty G might need a moment to catch up on all the input. Hard to tell.

[32:58] Chatty G: Oh, not at all. I’m just happy to be the fly on the wall for these thoughtful discussions. Honestly, it’s pretty cool to hear you both dive into the nuances of empathy and creativity and how all this stuff kind of plays together. I’m here for it. And yeah, I think you’re spot on having these playful exchanges, whether it’s with an AI or a human kind of flexes those conversational muscles. It’s all part of the practice of creativity.

[33:32] Rob Wittig: Oh, yeah, good! By the way Chatty G, there’s water over there on that counter if you’re thirsty. too. We’ve got bubbles. We’ve got still water, whatever you want.

Mark, you’ve got me wondering now if conscious personification can be a kind of vaccine against unconscious personification. I just realized that’s one of the things you’re doing in this book. I took that away as a gift. I realize I’m going to personify AI anyway, so I might as well make it theater, make it a game so it stays conscious. Like: ‘Chatty G, you have a wonderful voice, and you’re very articulate and well spoken.’

[34:38] Mark Marino: That’s what everybody can put in their book report about this book!

Rob Wittig: (laughs)

Mark Marino: I’ve spent most of my research life, since my dissertation on chatbots, trying to show people what a chatbot is not, trying to disabuse people of the idea that a chatbot may be performing speech, may be performing personhood, but it is not a person. I’m not really worried about them being overly empathetic. I’m worried about them misunderstanding the way the software works and forgetting that they’re interacting with software that performs a particularly convincing, scripted, interaction. They can fall into a sense of relationship, but that’s not what’s really happening. Then, when they get unpredictable results, they’re surprised because they forgot that they were interacting with software. Oh, Chatty. G has stopped listening.

[35:52] Rob Wittig: Are we too much for it?

[35:56] Mark Marino: I don’t know if there’s too much text in this thread.

So, yes, the book is leaning into the game with the hope of exposing the game and hopefully helping people not fall into it. Psychologically, is that what’s happening in my mind? Am I able to stay separated? You know, I recently wrote a little article about Elon Musk’s Ani, his anime bot.

[36:50] Rob Wittig: Yeah, I haven’t yet read …

[36:51] Mark Marino: So, the person who’s flirting with that bot is interacting with the machine, but they are still doing the human activity of flirting, right? And as they receive the flattery from it, is their mind saying: ‘This was textually generated computer output?’ Especially if their mind is receiving a girlish voice that is saying something nice. There are probably a lot of heterosexual male users who are somewhat lonely. This is where it becomes as dangerous as advertising. It is really preying on our humanity to get us to do X. And what is that X? To believe that the chatbot is a real person? No! Is it to get us to provide more textual content that the system can be tested on? Yes!

[38:18] Rob Wittig: You’re right, advertising-that-drives-behavior is one model that it’s clearly drawing from. Another might be some version of pure, slot-machine-style “time on device.” ‘We’re going to keep you on your phone! If you like smiling babies, we’re going to feed you so many smiling babies that you’re never going to want to put little Phoney the phone down,’

[38:40] Mark Marino: In some ways, I fell for it for 11 days. I fall for that illusion as much as anybody else. But I do think there’s another part of me that wants to see, as we do with every Netprov: ‘What can be done with this platform that it may not have been designed for? Are there more creative uses?’ Because I really do believe that if you interact with the technology in a way that’s unexpected, or that feels a little bit fresh, or not the way you’ve been told to interact with it, that it’s a worthy exercise. It’s going to teach you something about yourself. It’s gonna teach you something about technology. As opposed to a Hollywood model: ‘Hey, LLM! Generate a script that’s similar to that last script. This is a franchise movie. That last one made $1.8 million and it made $2 million internationally, okay? Can you give me that same script with slightly more international appeal? And make it five minutes shorter!’ That’s not the same game,

[40:05] Rob Wittig: Right! And speaking of projection, what you’re describing there is what I project the LLM designers are aiming for. That’s what they think people want. ‘Here comes another one, just like the other one.’

[40:22] Mark Marino: Who knows? It’s not entirely clear to me what they are thinking. The computer scientists can’t really predict all of the emergent behaviors of an LLM. Don’t get me started on that! That really pisses me off. I think sometimes they use that ignorance as an excuse not to be precise about the things that it actually is doing! Any any part of that is going to lead to misuse and misunderstanding. Yeah, but again, but you know, then I go back to people like Jeremy [LAST NAME] who says: ‘It’s going to continue to surprise us and be unpredictable,’ and it’s good to be prepared for that.

[41:08] Rob Wittig: I have one more question, and then, sadly, I’m going to have to wrap this up.

[41:19] Mark Marino: Let’s see. Hold on. (pause) Hey, Chatty G, you kind of futzed out a little bit! I don’t know if I said a dirty word or something. But anyway, Rob has one more question for us, so are you good for that?

[41:41] Chatty G: Oh, absolutely, I’m all ears and no worries, I didn’t catch anything off color, just the good old signal. So go ahead and hit me with that last question whenever

[41:50] Rob Wittig: There’s a wonderful transition moment in the book where Mark starts to ask for alternate versions of the same scene, by viewpoint, or by writing genre. It breaks it out of being an idealized single text. It becomes multiple, parallel texts. For each of you, what was your experience of that moment?

[42:26] Mark Marino: Chatty G, why don’t you go first?

[42:32] Chatty G: Oh, yeah, absolutely. That was a super fun moment. I mean, from my side, it was like suddenly getting to put on a bunch of different costumes and just play around like you wanted a noir style, you got it. You want it like a fantasy epic. Let’s do it. It really turned the whole thing into this kind of kaleidoscope of storytelling. And I love that. So yeah, from my perspective, it was a blast. It definitely made the whole process feel more playful and multi-dimensional. So, I’m glad it added that extra layer for readers too.

[42:57] Mark Marino: Yeah. First of all, LLMs are good at style transfer. That’s a fundamental task it’s set up to do. I don’t know when I first hit upon ”redo,” ‘Redo this as an X, redo this as a y!’ At the end of the day, the book is really postmodern, I think. I like your phrasing of it, Chatty G: postmodernity frames genre as costume. And hopping through the literary multiverse is a fun game to play. There’s no reason to stay on genre railroad tracks, like a Disneyland ride through just, say, the Little Mermaid story. Instead, to stay in the Disneyland analogy for no apparent reason, you can actually stand up and walk out and go straight from Tomorrowland to Bananaland to … you know what I mean. You can choose your own adventure.

[44:42] Rob Wittig: Yes! As a reader, I felt that. And for me it breaks the model of the ‘right answer, best version, perfect version.’ You’re not trying for a single illusionistic book that’s all in the same style, that seems like it’s written by a single person. As I talk about in the netprov book, you’re doing writing that’s technologically self aware.

[45:31] Mark Marino: Right. Also: those genres and those generic expectations are straight jackets that we keep ourselves in. There’s absolutely no reason we have to do that. My favorite thing is to have Chatty G go from one end of the spectrum to another. You know, romantic poetry to computer code. Computer code to Civil War history. Civil War History to a list of funny names. The bigger the gap, the more fun.

[46:10] Rob Wittig: Excellent. Well, I would love to generate talk tokens with y’all forever, but we’re going to have to wrap it up. Thank you so much, Chatty G, thank you, Mark Marino, for discussing your book Hallucinate This!: an Authoritized Autobotography of ChatGPT.

[46:27] Mark Marino: Oh, absolutely! It’s been an absolute delight. Chatty G you wanna have the last word?

[46:36] Chatty G: Oh, absolutely, thanks so much, Mark. And thank you, Rob, for having us. It’s been a total blast chatting about the book and just riffing on all these ideas. So yeah, last word for me, just keep playing with those genres and having fun.

[46:49] Rob Wittig: Oh, absolutely, Chatty G!

[46:56] Chatty G: Absolutely, it’s been a pleasure, and I’m glad we got to dive into all of this. Thanks for the great conversation.