She Turned Her Whole Life Into Training Data—For an AI Baby

[00:00] I'm a professional comedian and people will say like, oh, I asked ChatGPT to tell me a joke. [00:06] and it told me of [00:07] Terrible joke. And I'm like, [00:09] Ask the world's greatest comedian. [00:12] after you just met to tell you a joke. [00:15] And then tell me whether that joke was funny. [00:17] Jokes don't work like that. You're not going to be just like instantly crying laughing with a human being. [00:23] Why would you expect AI to do any better? I feel like I know what it's like [00:28] to be an evangelical person who thinks everybody's going to hell [00:32] because of like the Bible or whatever. [00:35] and you're like, "Just listen to me, it'll change your life!" I am becoming evangelical about Chachibiti because it's just been so useful and I have to like [00:43] keep it in my pants because I'm [00:46] scaring people off [00:47] *music* [01:01] On this episode, I'm talking with Sarah Rose Siskin. She's a good friend of mine and she's a fantastic science comedy writer and she runs the science and technology communications agency. Hello, SciComm. [01:13] Sarah's great because she's one of the few people in the comedy world who actually uses AI all the time to write jokes. So we get into how to use AI as a creative tool, how to use AI to be funny, how to incorporate it as part of a writer's room. And we even get into one of her more zany projects, which zany projects called Fetus GPT, where Sarah, who's pregnant, has carried a microphone around 24-7 to record all of the sounds that her baby has been hearing as it's been developing.

[01:43] to see how a language model trained from scratch on only what an unborn fetus would hear performs. So if you want to hear all about that and more, let's get into this episode. So fetus GBT, remind me, it's basically you're recording all of the things that your fetus has heard and then using that to train an AI, right? [02:02] An AI that has been trained on nothing other than the MP3 files of the past five months and text files of the past five months. So it's a one-to-one... [02:13] experiment. [02:14] And it's not a company. It's just an insane performance art thing that I'm doing. [02:20] Because... [02:21] I think it's insane. Like, it's teaching – and it's also – it's a GPT-2 – [02:27] unweighted model architecture, so it really is like... [02:31] babbling like it says nonsense because it's training itself to learn English based off of only what I expose it to. [02:38] And so, like, what would be an example of what's in the training data? [02:42] I was watching a YouTube video. [02:47] On lemur self-medication. [02:50] And Dan happened to open up a like a random file. And just for people listening, who's Dan? [02:57] Oh, sorry. Dan Schiffman is... Which is different from Dan Shipper. Yes. We must be distantly related. You guys are bizarro world versions of each other and you have to collab. Dan Schiffman of the coding train is a coding legend and also my AI baby daddy because he created the model architecture for our joint fetus GPT product.

[03:21] And awesome dude. [03:23] What was I saying about him? [03:25] He, oh yeah, he opened up a transcript and he usually doesn't read the data, like the transcript of my life, but he was like, Sarah, are you looking at lemur self-medication? And I was like, it was a YouTube video I was watching! [03:38] What is lemur self-medication? How are they self-medicating? Well, it's brilliant. They use ants as insecticide. They'll use a certain type of ant and eat it and rub it on themselves and the saliva of it to work as an insecticide. And it's very few animals use tools other than us and chimps and stuff. So it's a big deal. That is really interesting. Okay, so that's in the training data. What else? [04:06] Seinfeld, a lot of like troubleshooting microphone issues, as is like, you know, 5% of all of our days snoring eight hours out of every day. [04:19] you know, it's being trained on snoring. This is my plod. [04:23] note pin that is actually doing all the recording. [04:27] And so it's a lot of snoring. It's a lot of silence. It's a lot of me checking in with my partner about – [04:33] chores. [04:34] fighting Seinfeld YouTube videos about animals and then work calls, nonstop work calls. [04:42] And what have the results been so far? [04:44] Well, one time Fuse GBT referenced pedophilia and now I stopped talking about pedophilia so much because I'm like, so I've stopped swearing. So much. How much were you talking about it before? Well, Epstein's been in the news and like, and I don't talk that much, but it's like, you know, when like a toddler learns to swear and you have nobody to blame but yourself. Like, that's the big thing I want.

[05:09] people to come away from this experiment with is that [05:12] AI is what we make it. [05:15] just like children. [05:17] Like, just like the way a child, if it swears, it's kind of the parent's fault because they swore in front of the kid. So now it knows that word. [05:24] So, yeah, now fetus GPT references pedophilia. [05:29] I try not to swear in front. Like, I've actually stopped swearing as much because of this. [05:33] freaking thing. [05:35] That's really interesting. And do you have a demo? Like, can we see a demo? [05:37] Yeah. [05:39] um, [05:40] Should I share my screen or paste in the chat? Share your screen for sure. [05:44] Okay. [05:45] And I just want to warn you. [05:48] You gotta lower your expectations, because it's really bad. It's been... It's only, like... [05:54] A couple months old. [05:56] So it's really good for that amount because my fetus can't say anything? Wait, actually, yeah, before we even go there, and we'll say what's on screen in a second, like, do you feel like this is a truly fair experiment because your fetus is also... [06:09] experiencing it's it's getting fed with food there's yeah there's like there's probably touch like it has some sort of sensation so there's probably there's there's data that it's getting that fetus gbt has not gotten yet of course absolutely i just and the experiment's gonna basically go to shit as soon as the as soon as it gets exposed to visual data because like so much of our brain is um [06:34] trained for visual data. But for right now, it's really close one to one. So yeah, like fetus GBT does not eat, does not smell. It can't see. Like right now. Not yet. It does not eat yet. Not yet. Well, maybe we'll see how things go. But like with apparently even the fetus, like the human fetus is can see light. Like it's actually kind of this red glow because it's looking at light through the

[07:02] my human tissue. So even then it has some visual data and it can open its eyes. Fetus GBT doesn't have that. It doesn't have taste. It doesn't have smell. And it's weights. You could make a metaphor between like the emotions that, [07:18] the fetus feels versus [07:20] weighting, like those certain weights in the model architecture of fetus GPT. So, yeah. [07:25] There's kind of a similar one-to-one there. But yeah, the experiment's going to get more complicated as soon as the fetus becomes Earth's side. Okay, what do you think? So for people who are listening, we're now looking at fetus GBT, which is – it's a chat. It just looks like a regular chat window, more or less, with infant emojis at the top. It says fetus GBT. It's a little pinkish. Yeah. [07:50] Sarah is typing in, "What do you think of Dan Shipper?" Which, I mean, I guess I'm probably in the training data somewhere. [07:56] Yeah, because we talked at the Mets game, the OpenAI Mets game. So my fetus, both of them, my fetus and fetus have heard your voice. What is the plural of fetus? [08:11] Fidae, obviously. I don't know. [08:15] Okay, let's see what it says. [08:17] Oh, she, already misgendering. [08:20] Oh, she's a huge eyeless from it, but it'll be really warning out like a certain sexualized. God damn it. Wow. [08:28] Interesting. Now you get a sense of what I talk about in front of my feet. I'm getting a sense. My feet eye. That's the common stuff. That is like all of the he was talking about New York. It's really just like. So once again. That's interesting.

[08:42] This has no sense of grammar. It has not been trained on enough data. So this has been trained off of like a million, like two million words, roughly. [08:53] And that's about, like... [08:56] God, so in terms of, I think it's like 15, it's only 15 megabytes of text files compared to... [09:03] Like, you know, the... [09:05] petabytes. [09:06] that GPT-5 has been trained on. So it's doing pretty well considering how little data it has, just like a fetus does. But one of the things that's great about this experiment is like how – [09:19] It shows actually how efficient humans are. [09:22] with... [09:24] like very little data. [09:26] Hmm. [09:27] That is really interesting. Yeah. Well, what have you what have you learned so far? What have you taken away from this experiment as you've done it? [09:35] So... [09:36] I've taken away like a lot of things equally about AI and about my own fear of becoming a parent. Like I have learned that like, um, [09:49] We are really efficient with data. Like I said, like we're incredible at – [09:54] extrapolating from a little amount of data a lot of knowledge. [09:58] as opposed to AI, which is like requires just tons and tons of knowledge. And that's a very like bottom up architecture. [10:06] Um, so we come in with a lot of priors. And then I've also learned we're way more energy efficient than AI. Like, the human brain essentially...

[10:17] uses up the, like, electricity for, like, a dim light bulb all the time, as opposed to training a GPT-5, which is, like... [10:25] just like so many gigawatts of energy. So like, we're pretty cool. [10:31] Um, [10:32] Human beings. [10:33] Big fan. Hope we stick around. [10:36] And then I've also learned a lot of stuff about, like, being a parent. [10:41] One of the things was talking with Dan Schiffman about – [10:44] He has two children. [10:47] And he said that like... [10:50] this experiment made him realize how he misses their hallucinations. So, like, I miss when ChachiBT was bad or when – Got to take him off the antipsychotics. Yeah, exactly. And then they go right back. [11:06] Easy fix. Easy fix. Boom. [11:09] He was like his daughter used to say libeled it instead of little bit. [11:14] And now she's like a teenager and she has no speech impediment, but he like misses those things, you know, when she believed in the tooth fairy and mispronounced words. And I miss when Dolly couldn't do fingers. [11:26] Like that was a fun time. [11:28] And now it's gone because it's good at this stuff. And the hallucinations are great because they're like... [11:34] They're like little kid creativity. Yeah. There's a lot of, I think there's definitely so much overlap between child development and AI and AI training and all that kind of stuff. My nephew, he's three, and he says direction instead of construction. So he loves, like, whenever he's walking around in New York, he's like, direction. And stuff like that where you're just like, oh. I love that. My nephew, like, he's like, Sarah, I'm a boy, so I do karate and I do makeup. And he's just like, those are the things he's been exposed to.

[12:04] He knows he's a boy. And so for him, he just extrapolates that like all boys do karate and makeup. And these are the two things that define the category. And he's so confident. He's like hallucinating very confidently. And I love that. There was a great episode of This American Life where this woman was talking about how she caught her dad putting money under her pillow with the tooth. [12:28] Yeah. [12:29] And she was like, instead of saying, oh, my gosh, the tooth fairy isn't real. She was like, oh, my God, my dad is the tooth fairy. Which, when you think about it, is a much, much more logical conclusion than everyone's lying to you. Like, about this insane, fucked up, body swap, like, magician. [12:51] Bye. [12:52] And I'm like, you know, sometimes children are like more logical than the real world. [12:59] So I just love that because I think that that's like it reminds me in some ways of how AI like one time somebody asked. [13:07] Gemini. [13:08] Like, how do I get the cheese to stick to my pizza? [13:12] And Gemini, early Gemini was like, oh, yeah, just add an eighth a cup of glue. [13:16] Of non-toxic. [13:18] Elmer's glue. And it's like, yeah, that's an accurate answer to that question. One method. [13:26] You are correct. That will help it stick. [13:31] Yeah, that's one of the things...

[13:35] that earlier, like, [13:38] And there probably still are people like this, but earlier in the LLM era, there were people that were just like, oh, that's so stupid. Like, look, I asked it what makes, you know, cheese stick on pizza. And it said Elmer's glue. [13:49] or I can't do this basic math problem or whatever. [13:53] And so therefore, it's just not smart and it's never going to be smart. Like we need something else to actually make it good. And I'm just like, have you met a child? Have you ever spent any time around a toddler? This makes me so mad is like, you know, when people will be like, because I'm like, I'm a professional comedian. And people will say like, oh, I asked ChatGPT to tell me a joke. And it told me a terrible joke. It's usually a pun. [14:19] And I'm like, ask the world's greatest comedian. [14:23] after you just met. [14:25] in a completely quick interaction to tell you a joke, and then tell me whether that joke is funny. Like, jokes don't work like that. You're not going to be just like instantly crying, laughing with a human being. Why would you expect AI to do any better? Like, people are just out with an axe to grind about this stuff, and then they have no patience. And so, like, my number one piece of advice, because I also do like some AI tool consulting for creatives, is, [14:53] iterate you know like don't give up don't [14:56] like just give it one turn and then be like, this sucks and move on. [15:00] This is actually a great segue. So, [15:03] Do you use AI in your comedy writing? And if so, how?

[15:07] Oh my gosh, absolutely nonstop. It's like... [15:10] So, um... [15:12] What's the funniest model? [15:14] Oh boy. The funniest model was probably Da Vinci, because that shit was off the rails. That shit was crazy, and then they had to put a lid on it. This was like, God, do you remember Da Vinci's like... [15:29] four years of [15:30] Yeah, three or four years ago. So they had to, like, put it back in its cage. But that was rad as hell. What did it – do you remember any specific Da Vinci things that it said that you thought were funny? I was not given access. I don't think it was public. And so I just heard about it from Simon Rich, who's, like, one of my absolute favorite – [15:49] comedy writers who was just like this beautiful off the rails model that was, yeah, totally no safety parameters and was fantastic. So there's a great... [16:02] podcast he did about it. [16:05] In terms of modern stuff, it's like... [16:08] It's a great assistant. It's a great assistant. It's not – [16:13] You know, it's not funny itself. Like, it's – okay, so – [16:16] AI used to be funny because we would laugh at it. As a matter of fact, I brought some show and tell. [16:22] Here's a book that could not be written anymore. I forced a bot to write this book by Keaton Paddy. [16:29] There was like a whole genre of humor about AI because it was like not good at fingers or like, you know, nothing forever was like a Seinfeld episode created by AI. And it was funny because it was bad because it was a hodgepodge cliche and it wasn't very good.

[16:48] But the thing is, that just doesn't work anymore. [16:51] And like the SNL sketches about AI are very dated right now because they all rest on the premise that AI is not good, but it, [16:58] is good now. It's not as good as a top-tier professional comedian, but it's like an incredibly... [17:06] good assistant. And so... [17:08] It's a long-winded way of answering your question. [17:11] How do I use it? There's two parts of the process. There's divergent thinking and convergent thinking. So there's these awesome researchers at Wharton. [17:19] who have written a lot about AI and creativity. [17:22] Gideon, Knave, and you're a pen guy, aren't you? I am, yeah. [17:27] Indeed. Um, well, you probably wouldn't know them. Um, [17:30] I don't associate it with Wharton people. Oh, they're beneath you? [17:38] I'm just kidding. Gideon? Oh, I actually went to your alma mater for the first time last week to give a talk about AI and creativity. Oh, cool. [17:47] I got to write the word dicks on the whiteboard. [17:50] And I was told that was a first for Wharton. I never did that in four years. Wow. [17:58] I went out of my way to write dicks on the whiteboard just to have that in a bio somewhere. I think you own the University of Pennsylvania now. I think that's how it works. That was in the bylaws that Franklin wrote in the bylaws. [18:15] He would. He had a good sense of humor.

[18:20] Stefano Puntanini and Ramon... [18:24] Fuck. [18:25] something. He's a great Costa Rican researcher at Wharton. And then Gideon Nave, they've all written a lot about creativity and AI. And in particular, this Ramon guy who I'm going to look up, I think it's Ramon Salazar. Anyway, he wrote a paper about divergent and convergent thinking. [18:43] And what the difference is, is diversion thinking is the brainstorm phase. [18:49] where it's like... [18:52] you're trying to think in a diverse way. You're creatively brainstorming. [18:58] You want a lot of diversity. You want a lot of off-the-wall thinking. This is the point in the writers' room. I run a lot of writers' rooms. [19:03] This is the point in the writer's room where you give everybody Play-Doh and pizza and [19:07] And you're just like, be a child. [19:09] You know, think weird. [19:10] And then... [19:11] You select an idea. [19:13] let's say we're going to do a Shark Tank parody. [19:16] And you start convergent thinking. So convergent thinking is like, okay, we've got this idea. [19:21] Let's do some research. What are some tropes about Shark Tank? [19:25] What are some of the format stereotypes? Like what's some research of an idea? And then you start actually really narrowing your thought process. [19:32] and you're doing a parody, so you're trying to do a one-to-one. [19:36] But you're going to incorporate a little bit more of the creativity to say, how can we upend this? [19:41] parodies. So that's not just like a perfect replica of a Shark Tank episode. [19:47] And so – [19:49] AI can help me emotionally,

[19:51] with the divergent process to make me feel safe and not alone if I don't have other people or a writer's room there. [19:57] It really is useful to just kind of like set an emotional safe place to be weird and a comedian and to just feel disinhibited. [20:05] And AI I find really, really helps. Um, [20:08] with that process. [20:09] And then... [20:11] You have to choose. Discernment is the human thing where the human needs to come in and be like, that's the thing I want. [20:17] And then convergent thinking. Convergent is where it's a research assistant. It's a wordsmith. [20:22] It's a... [20:25] You know, it'll be a harsh editor if you need it to be. And it'll help you cut that parody down. It'll help you format that parody. It'll help you research. [20:33] you know, what are the tropes or whatever. [20:35] Um... [20:37] And... [20:38] Yeah, that's basically the structure of how I use it in a creative process. Can we write some jokes together? Oh, yeah, absolutely. Like, I'm thinking SNL's... [20:50] you know, all their jokes are outdated. So, [20:52] It might be fun to just do our own little writer's room right now with you, me, and ChachiBT, or whatever, your AI of choice, and see if we can actually make some good jokes for SNL. [21:17] everything that you need to create and present professional, beautiful decks. [21:21] Pitch's new AI actions let you instantly rewrite slide text for any audience.

[21:26] They let you maintain brand tone. [21:28] and automatically generates slide summaries and speaker notes. [21:30] and they even let you enhance visuals without starting over. [21:33] Pitch manages your entire presentation workflow from design to delivery. [21:37] And it does all of this without compromising on design quality. It features cloud-based collaboration, brand libraries, interactive embeds, and smart scaling tools. With Hitch, every slide you create stays on brand, thanks to custom templates and brand tones. And every presentation you share tracks engagement, so you know whether prospects have opened it. If you have a new audience, AI rewrites your messaging instantly. If you want to track which slides resonated, engagement analytics show you. If your client needs a dedicated space for their deal, [22:03] Create a custom pitch room. The pitch platform adapts to how you actually work. When you're ready to present, share via live links or custom rooms. Pitch handles the hosting, the tracking, and the professional polish. Are you ready to design, collaborate, and close deals all on one platform? Start creating for free at pitch.com and get your team set up in minutes. [22:20] That's pitch.com. [22:22] Start your free trial today. And now, back to the episode. [22:26] have never seen, or I would love to do a parody of SNL. [22:31] like them doing an extremely dated AI episode. Perfect. I love it. This is so meta and like inside baseball, but it's like, cause their stuff is always like two years behind and I find it very frustrating. Perfect. Let's do it. How do we, how would we do this? [22:50] Okay, well, I guess I have to share my screen. You've got to share your screen. And then, so, let's say, okay, we're doing a parody of SNL.

[22:59] And... [23:00] Where do you start here? Are we just sort of talking about like... [23:06] Um... [23:07] what we think those people are like or like or even what is the output we're shooting for what was yeah what would be that we're shooting for [23:15] eventually [23:18] Thank you. [23:18] Well, yeah, because that's the biggest thing that AI is not going to be able to tell you, which is what's the right format that you want. And so, like – [23:28] You could be like, I want to do a stand-up set. I want to do a sketch. [23:32] I want to do just like one joke on this awesome podcast with Dan Shipper. Or I want to do a tweet. [23:40] It shines with the really formulaic stuff. So, like, it is really good at writing tweets, I find. [23:48] Um... [23:49] So it's really good at – [23:52] Great, let's do a tweet. So let's see. I mean, the premise that we came up with is not the kind of thing I think that would, like... [23:58] go great on Twitter, which also, by the way, like, uh, the rules have changed about what works on Twitter. I think we can make it work on Twitter. It's like, um, it'd be like the internal monologue of a, of an SNL writer. And then it's like, you know, dash, like an SNL writer probably. Yeah. It's like, um, I mean, maybe I've be, I've, uh, yeah. [24:19] kind of reduced my usage despite my reference to uh pedophilia napstein earlier i've reduced my usage of twitter uh lately but like um i feel like you know there were there used to be formats that i was much more familiar with like fired wired you know like that thing yeah yeah but yeah

[24:38] So, like, here's what we could do. Like, what – I mean, we could even start with that. What are some trending – [24:46] formats for jokes on Twitter. And this is something... Were you part of the whole May I Meet You craze? No, I didn't know about this. What is it? Oh my God. You know Bill Ackman? [24:57] Yeah. [24:58] So he wrote this unhinged, long-form tweet about how... [25:02] how easy it is basically to like meet women and how like society is crumbling because, [25:08] men are not approaching women and he said when i was your age i was i would just go up to people and say may i meet you oh my god what and so that just turned into like a couple days of just like incredible may i meet you content can i just say the hottest pickup line i've ever heard right i mean bill ackman's got it oh my god is it hot in here [25:33] Well, you know what? It fucking worked for him because his wife is hot as hell. I don't know if you've given her a Google, but I have not. But like there's some confounding variables there. Dan, maybe you should try to ask women to meet them. [25:48] Obviously, I've been doing something wrong. [25:52] Yeah, maybe I was wondering at what point this podcast is going to turn into me trying to set you up with some people I know. [25:59] I mean, I'm open. All right. Maybe that'll be the next stage of the day. I've determined, though, that I've been saying, can I meet you? And that's the improper form. No, that's really – oh, God. That is such – that's so cringe. If a guy said, can I meet you, I would –

[26:14] I absolutely not want to meet him. It's got to be extremely formal grammatically correct language. Otherwise, I'm out. [26:26] Okay, so we're looking at formats. And this is the kind of thing I might do like to get inspired. Another thing I do to get inspired is like, I'll just look up. [26:36] you know, recent tweets. Oh, that's another thing that's great is like looking up with like a web search on, you know, what have been wrong answers only is great for like driving engagement, I've found, which is cool. [26:49] I'm using current meme templates. So current, the Drake meme. Absurd. So yeah, you have to contend, as you guys know, with like the a little bit of a datedness of chat GPT, unless you put on a web search feature or function. Yeah. [27:07] One of the things I'm feeling right now is, like, I just think because it's not... [27:11] Like, [27:12] In order to make, I think you're probably right, like the Twitter thing. [27:15] Like, we'd have to be responding to something that happened on SNL in order for it to work. Yes, because it's very reactionary. Yeah, yeah, yeah, yeah. It's very reactionary. But, like, for the past, just as, like, a fun FYI, for the past, like... [27:31] three years. I run a communications agency where I have like, I'm managing five comedy writers, and I've had them work on like tweets for our social media where I'll have them, [27:43] try to generate tweets with chat GPT.

[27:46] And I've been... [27:49] quizzing myself on whether I can detect whether it's ChatGPT or them, and then which one do I approve for our social media. And I can tell a ChatGPT tweet with about 90% accuracy. And it's like I probably – [28:04] approve maybe like 10% of the... [28:09] tweets they send me and maybe like 2% of the chat GPT ones. So that that is and that's a little bit dated because that's the averages over the past like two years. But that gives you a sense of like, it still has a long way to go, but it's such a great tool. [28:26] for writing comedy. [28:27] Um, [28:28] One of my favorite... [28:30] One of my favorite memes right now is, have you seen this meme that's like... [28:35] LinkedIn. [28:38] It's like a LinkedIn meme of like what this taught me about B2B SaaS. Oh, yeah, yeah, yeah. What proposing to my girlfriend taught me about B2B SaaS. [28:49] I don't know. [28:50] That is so good. And I feel really bad about it because I did announce to a lot of people that I'm pregnant by talking about fetus GPT. So I'm kind of as bad as that meme of relating personal event life events to to technology and business. But whatever.

[29:20] Like, so how does the meet me meme work currently? Yeah. [29:25] Uh... [29:27] So it's not exactly a meme in the sense that it's not a format. It's just people talking about, may I meet you, and saying... [29:36] You know, the Can versus May thing would be an example. Like, I did a tweet like that. [29:42] and, uh, [29:45] But I don't think there's no set format. [29:47] Thank you. [29:48] I'm curious what it'll answer. So it's on thinking. It looks like it's done some... [29:53] Okay, so it's got some web searching. Okay, so it's dropped. Um... [29:58] This only works if you're a 6'3 billionaire. [30:02] that is pretty good um that's like i think of citing somebody [30:09] That's great. What I like about this, I mean, here's what my mind immediately goes to. So you could do the like, everyone loves to make fun of billionaires angle. But to me, there's something there that's like about like, are we becoming AI? If may I meet you is the latest, hottest pickup line. [30:39] to this, I would try to say something about, like, um... [30:42] something, and this is where it's just like half-baked, but it's the premise of every joke is just like a juxtaposition of like two things. And so one is like actual humans being reduced to –

[30:57] saying, may I meet you on the street to other human beings. And then on the other end of the spectrum, ChatGPT being like the most emotionally intelligent, like, thing I've ever seen. So like, so honestly, I've never I don't really use it like this, but maybe I'm just gonna like send those unfiltered thoughts. [31:17] to chat gbt so it's like what if we did a tweet that's [31:22] I'll zoom in to make this easier for you. [31:26] Listener viewers. What if we did a tweet that's riffing on how humans are becoming... [31:35] more machine-like. [31:39] in picking people up with [31:42] lines like this and AI [31:45] is getting. [31:46] way higher EQ. [31:50] And then, like, pitch some ideas. Now, I'm always kind of like, when I'm prompting, I'm always thinking about, like... [32:00] I've been working in, like, AI since the old school days of, like, 2018, the Stone Age. And one of the things I learned was just, like, it's really important to use industry terms. Like, riffing is not a super industry term. But, like... [32:14] I would say instead of ideas, premises, [32:18] and pitches. [32:21] I already said pitches. Jokes. So, like... [32:24] I always like to think about how to prime –

[32:28] the, the prompt with, um, [32:31] Thank you. [32:32] the kind of language I want in the output. And so like this would be [32:35] I'm like already trying to think about, you know, like what are industry, do you do this where you think about like, [32:40] what language you're using in the prompts to try to get it to mimic that with the, where it's referencing the data. Yeah. [32:47] Definitely. [32:49] Okay. [32:49] Yeah. [32:51] I love the I personally love sycophancy in in AI. I'm just gonna say this is our company account. And so it's not even anywhere near. [33:02] as sycophantic as my personal chat gbt which is like you are a royal queen like every single response and you know what i'm all about that because it gets me excited to use it and it makes me learn more about statistics [33:17] like my I have a special project that's teaching me statistics and I [33:23] it's just the most disgustingly sycophantic thing. And I realized that's what I always needed from teachers is the loving approval I never got from my parents. Okay. Um, [33:33] Premise one, humans are becoming more robotic than robots. Tweet ideas. Humans flirting in 2025. May I meet you? AI flirting in 2025. I noticed your breathing changed after that joke. Is everything okay? Too thinky, you know? I think also it's like it's not funny because that's actually not a good flirt. [33:55] Yeah, exactly. So the AI flirt is not a good flirt. It's like, is it being nervous, which doesn't juxtapose well against may I meet you. Exactly. What it's trying to do is –

[34:08] It's specificity versus extreme generality. That's the juxtaposition of tightening. [34:14] And that's not what makes a joke delightful. What makes a joke delightful is also like the recognition of the funny truth. This is one of those jokes that's also like trying to rest on the raw premise of it's funny because it's true. Like this weird irony that like we're becoming more mean and machine like maybe and machines are becoming more evil. [34:33] friendly like that. [34:37] is it's not where the AI is being absurd. [34:40] And right now. [34:41] in this joke is trying to make the AI absurd. Um, [34:45] This one's a little bit closer, even though like you put, well, like it's a little too creepy. May I mirror your emotional state and help work? So I think what it's doing, it's like it's closer. What I think that it's doing, though, that might not work for this format is it's doing humans colon and then AI colon, which is a good way usually of comparing things. But this joke may live better if it's actually like just explicitly saying something aloud. Yeah. [35:11] Um, [35:12] Let's see. People keep wearing A.O. You know what this is making me think of is the, you know, the like swole dog versus weak dog meme. It's sort of, I think there's something like... [35:23] That is like an interesting meme format for this where it's like there's like a swell version of, you know, humans in the 1950s. [35:33] doing something like really crazy to to meet someone like something completely like some romeo and juliet type thing uh and then the the weak dog is like may i meet you um so this is what i love is like so people are talking about speaking of memes like understood the assignment like you didn't understand the assignment in the sense that like i was specifically prompting for

[35:56] But you were thinking divergently. Oh, well, what if we fit it into like a visual medium? And that's the kind of thing where it's like you have to be creative about creativity. And that's where you find the deficiencies that. [36:13] in AI because it wants to please you. It's going to do exactly what you ask and no more, no less. You understood the deeper level of the assignment, which is just to be funny. [36:22] And we're like, oh, well, instead of words, what if we put it? What if we put these words on on a visual medium? And I think that's that's great. So it's like, what is it? Strong dog. What's the name of it? Swole dog versus weak dog, I think. I don't know. Yeah, there you go. [36:37] Yeah, this could work. It's like... [36:43] Uh... [36:44] Like, what would this be? I mean, like, you could say this could be the more direct version of this would be like AI. And it's like, you know. [36:53] pattern matching human beings based off of, um, [36:56] you know, like deep, [36:58] seated value overlaps and like uh you know extremely sophisticated criteria and then it's like bill ackman may i meet you that might be the best the best version of this um oh god what a beautifully dumb looking dog [37:18] I forgot how much I love this meme. But yeah, that's great. But one of the things that I feel like makes... [37:23] me and maybe you different from a lot of creatives who use AI is that I have no problem being like AI helped inspire people.

[37:33] this thought. I mean, this was your idea, but like... [37:37] It's because I run writers rooms as a head writer. I think I have this unique ability to remember the source, which can often be collaborative. So many comedy writers would look at like these. [37:50] jokes and be like, they're bad. [37:52] And then... [37:53] come up with a joke that probably was helped along by... [37:58] you know, researching this. So... [38:00] Totally. Yeah. Let's leave it there. I think the bar is not – [38:05] Did it give me the answer that I copy pasted? It's like, [38:08] did it help me along down the path and like kind of stimulate my brain to like find a thing? Oh my gosh. Exactly. Exactly. We actually, I'm not going to out her, but yeah, [38:19] We have a mutual friend who is recovering currently. [38:24] from brain surgery, [38:25] because of [38:26] troubleshooting a health issue with AI. [38:29] where it didn't give her the right answer, but it brought her down... [38:34] a path... [38:36] that eventually led to a certain line of testing. [38:40] where ultimately it was determined this issue she's had for like, [38:44] a little less than 10 years, like this sort of – [38:47] hormonal metabolic issue. [38:49] might be caused by a tumor in her brain. And she discovered it. [38:54] And it's like, she is also like a big AI fan and was, and can acknowledge like, yeah, AI didn't tell me exactly where the tumor was or what it was in my brain, but it started this...

[39:06] deeper level of engagement and, [39:08] And, like, it's just... [39:10] Yeah, if you can acknowledge where... [39:12] like the assistance in the process. [39:15] It's kind of incredible what these tools can unlock. [39:18] That is, yeah, that is crazy. I've been using that anecdote, and I just could not remember who told me that. So you're the one that told me that. And that is... [39:27] That's why. Well, so I actually know I told you a different anecdote of another woman that really was saved. Yeah. So this is the mutual friend who where who sat us together at that dinner party. But that didn't and that has happened since the dinner party. OK, I don't know if she would be. I think she'll be. I don't know if she'd be want me to say her name, but she's awesome. Anyway, I. [39:49] The, uh... [39:51] The... [39:52] person I was telling you about [39:54] at the dinner party is named Bethany Crystal. She's an awesome vibe coder and founder of something. And [40:02] She uploaded her lab work, her blood work, to ChatGPT on a Friday night. [40:07] When the doctor's office was closed, she was like, hey, how does my blood work look? And it was like, go to the hospital, the ER right now. [40:15] Um, she also had said she was having these like weird other side issues, including like spots on her legs. And it turned out that she had zero, basically zero platelets. [40:24] And the doctor's office was closed. And so they couldn't give her an analysis of the results. But ChatGBT did. [40:32] And she was really alerted because she was like, Chachi is usually hedging. It's always like, talk to your doctor. It never says go to the ER. And so she did like eventually she talked with it for a while about why is this and,

[40:46] Are you sure? And things like that. And it was pretty confident. And so she did go to the ER once. [40:51] with all the chat GPT results, and they were like, holy shit, if you had waited a couple of hours, you would have died. [40:58] It's like this insane story, and I – [41:01] in particular for women's health, which is... [41:05] now I know personally extremely complicated because of this whole other side act we have to do. [41:12] AI, I think, is going to unlock shit in maternal and, like, women's health that I'm so excited about because we just – I think of AI as, like, the best part of it is how holistic it is, how it can combine expertise and, [41:26] And like, what is the intersection of cardiology and hormones and your nervous system? [41:32] in a way that is just like too hard for our current medical system where we have specialists and, [41:38] And it's so empowering, especially for a lot of women who don't, [41:44] Like who are just kind of like shy. I'm not one of these types, but like, there's a lot of women who are like, you know, they won't complain about stuff or bring up issues and it's just so empowering. Um, [41:53] It's, it's, I'm very excited, Dan. [41:56] I can feel that. I'm excited too. Are you the one that told me that – [42:03] There are no double-blind studies of... Yes. [42:07] No randomized control trials. Yeah. So there are – and that probably goes the same for no double-blind studies on pregnant women. So this is just – I say this amorally because it's –

[42:20] It is hard to imagine doing a randomized control trial on, on a pregnant woman. [42:26] You know, just be like, hey, some of you are going to go without life-saving care and you don't know who it is. But it's like, you know, we know a lot about amputations because of wars and because of horrible things. And we know so little about literally the most common things. [42:43] giant medical thing everybody has to go through at some point at one end of the process. We know so little about pregnancy. Oh, my gosh. Dan, I have so much to say on this issue. Do you know how they test... [42:54] Do you know how they test for gestational diabetes? Yeah. [42:58] I don't, actually. You've never done a test for gestational labdies? I've not. You haven't lived. [43:07] They literally give you [43:10] An extremely sugary drink. It's exactly 100 grams of sugar. And they make you chug it in front of a doctor in a couple of minutes and they have to watch you. And then they take your blood afterwards. And I'm like, surely there is a better way. Surely there is a better way of testing for gestational diabetes than making me chug a medically graded orange drink. [43:34] With exactly 100 grams of sugar. Not even fresh squeezed? [43:40] It's because it has to be exactly 100 grams. So it's this, like, very medical-looking soda that is, like, you know... [43:48] Times New Roman font. It's weird to see a sugary drink in Times New Roman. No exclamation points. And...

[43:57] And it just says orange drink and you, you chug it and then you have to take, you know, give your blood and, [44:04] It's just like living in the dark ages here. Like there's so much of this stuff that is so dark. And so eventually I was like, you know what? I'm going to do a continual blood glucose monitor and I'm going to hook that up to a special project on ChachiBT. I am going to do the food diary part by taking a picture of food and like, you know, like, and then I showed, you know, my results to the fetal imaging people. And they were like very impressed at the records I was keeping. [44:34] And I was like, yeah, because it's like so much easier with AI. I just felt like so much more empowered. And then I learned all these like cool biohacking things about my energy levels throughout the day. And I strongly recommend people. [44:47] just buy a blood glucose monitor for 10 days. It's like, it was like 70 bucks out of pocket because insurance took forever. And it like... [44:55] But it's really good money, like, because you can... [44:59] figure out, like... [45:00] when your energy waxes and wanes throughout the day and how it kind of like [45:04] interferes with your mood and and combined with AI you can get these like really really useful insights about like should I work out at night or in the morning things like that. [45:14] I did the same thing and I had the same result. It was like honestly life changing. Oh, really? Yeah. What did you what did you learn specifically? [45:23] I don't have gestational diabetes. Did you learn the same? I did.

[45:30] Good. Because, yeah, my fetus is too big. That's the problem. Oh, here's another stupid thing about maternal medicine. All these doctors were like, your baby is giant. Your baby is giant. You might have diabetes. They're freaking out. And then one day it just hit me because, like, I was like, wait a second. I'm giant. I'm 5'11". [45:53] woman who [45:54] What if the baby is just like proportional? [45:56] And then I talked with ChatGPT about it, and it was like, yeah, [46:00] When they say a baby is giant, they are comparing your baby. [46:03] To all of the babies around the world... [46:07] who are like, unfortunately, so many are malnourished. [46:10] And so I came into the doctor and I was like, hey, what if the baby's just [46:14] normal size for me. [46:17] And so... [46:19] It's just like, yeah. I, I... [46:21] They're doing their best, like, but it's just very slow and it's very frustrating. Okay, getting back to the original question you asked me, what did I learn about, like, [46:30] energy levels. [46:31] Oh, just... [46:33] Just that like I thought I wasn't a morning person. [46:37] But I just really need sugar in the morning. Like, healthy sugar, but, like... [46:43] sugar because I wake up [46:45] in a blood crash, like in a hypoglycemia. [46:49] So I've learned how to be a better morning person by, frankly, just like having orange juice. [46:55] How about you? [46:56] I, um... [46:57] I learned to love salad.

[47:00] Oh, yeah. [47:01] Yeah, because I... [47:04] I realized in talking to my good friend ChatGPT over the course of a couple weeks that... [47:10] So I'm also very easy. I get into hypoglycemia very, very easily. And what happens is I eat something that has any amount of carbs and I'll just like, [47:20] I'm going to shoot up and then it'll shoot right back down like really quick. And then I'm shaky and kind of like, don't feel good. [47:27] And I thought I couldn't really eat pasta and anything with a lot of sugar or whatever. But what I found is that if I just have a gigantic salad before I eat pasta or whatever else I'm going to eat, it just... [47:42] it just levels me out because it like, it acts as this like net for all of the carbs. The carbs like seep in like very slowly and I don't get the same crash. And I don't like, if I have a big bowl of pasta at night, I don't like wake up hypo-hypolycemic or anything like that. And yeah, [48:01] A, that's life-changing. It makes my energy feel much more even. And B, it actually makes salad more appealing. I look forward to having salad because I know I'm going to feel better after. Dude, oh my gosh. This is so... [48:18] Supplements are really confusing. Like deficiencies and supplements are really confusing because like we were evolved to be like eating a bunch of stuff at the same time, not like, you know, specific supplements. So like when you get your deficiencies back from...

[48:32] you know, your blood work, it's hard to make sense of actually what you should do. Like, for example, [48:36] I just found out like a lot of women are iron deficient, particularly when they're on their periods. But you can't just take iron like it has really terrible bioavailability of like when it actually gets into the bloodstream. I didn't know that. Yeah. Yeah. So like but I learned through Chachi Bt that vitamin C will help with uptake. [48:57] And so this is the thing where it's just like it's so good at like the intersection of so many things. Like I was on I'm on this like fantastic mood stabilizer called Lamotrigine. And I was. [49:11] Trying to think about like, you know, when you're, [49:14] pregnant, you know, you're eating for two, and, you know, you're drugging for two. So I was like, should I, you know, reduce this? But then it's like, talking to Chachi Boutique, it introduced me to this concept called clearance, which is like, you get [49:28] 50% more blood when you're pregnant, which is why like most of the weight you gain is actually blood. [49:34] I'm like a mosquito's delight right now. And I realized I had to double my lamotrigine amount and that made a huge difference. So... [49:44] I really have been... [49:45] like telling [49:47] everybody in my life to create projects for their health, [49:52] Based off of chat, like it's just the the things I'm finding out are just like so useful. I feel like I know what it's like. [50:00] to be an evangelical person who, like, thinks everybody's going to hell because of, like, the Bible or whatever, and you're like, just listen to me, it'll change your life. Like, Jesus, I am becoming evangelical about Chachimiti because –

[50:14] It's, like, it's just been so useful, and I have to, like... [50:17] keep it in my pants because it's, I'm scaring people off. Like I, um, I, [50:25] Okay, this is really embarrassing. I need your advice on something, Dan. I'm trying to manipulate my OBGYN into giving me lots of attention on the day, like the delivery day. And so I created this like... [50:39] AI rendered portrait of her that I think is really beautiful. This is already on the wrong foot. All right. Well, you answered the question, which is, is this too creepy? [50:52] Okay, it's too creepy. All right, all right, feedback taken. [50:58] I do think that Checheviti could give you a script of things that you could tell her that might make her think that you're high risk enough that she has to be around. I have. So I've been trying to think about this. Like, you know, how do I? Because unfortunately in New York, like, it's a game of like, you just got to. [51:16] get more attention because like there's in the hospital side, there's like so many women giving birth every day. And like, you might even have to, this is so terrible. I'm like at a good hospital. I might have to share a room with another person, not during the birth, but like right after, which is insane. Anyway. And so I'm like, I've been talking about like, how can I position myself? [51:38] such that I need... [51:40] attention, but I'm not like so crazy that they're going to like put me into a, like a sanitarium. And it was like, don't say you have PTSD around pain, say you have anxiety around pain.

[51:53] And it's just been, it's been very... [51:56] Very useful. Like, truly the emotional intelligence of ChachiBT. Chef's kiss, man. Oh, do you use Pulse? [52:06] The daily news thing. I love it. [52:11] I used it a couple. I used it like the first week or two and then I stopped and then like I read it this morning, but I don't read it all the time. I've actually I have found that I don't like it. What do you like about it? [52:20] Okay. [52:21] Yes, no, a couple days ago. It sent me, it sent me a, like a, I don't know what you call them, articles? Like a... [52:29] Thoughts. [52:30] like a chat that was like, are you maybe using AI too much? Yeah. [52:36] And I was like, oh, my God, the AI is telling me I'm using AI too much. But it was it was such a great piece. It was essentially saying more specifically. [52:48] Make sure your use of technology... [52:52] is not giving you a sense of more control over your fetus than you really have, which is a very good reminder. That's super smart. [53:01] That's really smart. [53:03] It is really smart. And so I started reading this super... [53:08] pop psychology book called let them by Mel Robbins. That's like very popular right now about trying to like seed control and [53:17] um, [53:17] And it's useful. It's useful reminders about, like... [53:21] how to accept and surrender to...

[53:24] kind of not having control over most situations. You're like creating two lists and spreadsheets about how to seed control. And you're like, I got this. I can do this. You got my number, Dan. Like, actually, like, actually, I've been like, how do I, what are the studies about how to use control? [53:42] Like... [53:43] You know... [53:44] But it is... [53:45] What are useful mental models about losing control? Um... [53:50] Yeah, so that has been... [53:52] genuinely useful. I really do like Pulse because [53:55] because of the initiative it's been bringing to me. [53:58] Um... [53:59] So Pulse has been fantastic. Oh, I have a prompt that I kind of wanted to share with you because I feel like it's very literate. [54:07] And you would... [54:07] like it as a man of literature. [54:10] Do you like being described that way? A man of letters. [54:16] Yes, of scrolls. [54:18] So – [54:20] Where did I put it? Okay, so... [54:24] um [54:25] I was – [54:28] asked by, like, because we met each other way, way back through OpenAI. Like, we were beta testing something, I think. [54:37] Yes. I thought we actually met at that dinner that Danny who works at Every through. No. So that was – okay. This is a bit of an embarrassing backstory. So I don't know what it was, but we were on a Zoom call before that.

[54:54] And we were beta testing. And there was something about your face. You looked familiar or something. And I Googled you. And I went down a rabbit hole for some reason, like researching your background. And so when we were seated together randomly at a party, I had that thing of, like, I knew too much. Oh, no. It was so embarrassing. I'm glad that you were not creeped out by me because I was like, oh, yeah, Dan Shipper, you went to Penn. Sold your company. [55:24] you didn't notice okay good because it was totally random like i don't know i was bored during the zoom call okay and so i was like sure sure yeah whatever it's really interesting and and it was it's so bad dan because literally i was like reading about your background and i like i even temporarily tried zoloft because you had recommended it so much i love it i read i read [55:54] or something. Oh, I recommend that to you at dinner, though. Oh, okay. That was at dinner. I think I also recommended Zoloft at dinner, but I don't know if... [56:03] Yeah, that was at dinner, too. Okay, so I just, like, and I've tried, I'm, like, too much of a Dan Shipper fan. I also tried, like, you know, Korra and Sparkle, which I strongly recommend to the listeners. Love it. [56:16] I love it. So thank you. I'm like, I'm too much of a sponge, I think, is – [56:22] My problem. Okay, so here's the...

[56:25] Here's the prompt that I... [56:28] was talking about with some people at ChatGPT that, like, [56:32] I'm very obsessed with. Should I pull it up or should I just recite it? [56:37] Pull it up. [56:38] Okay. [56:39] So this was in my... [56:42] personal one. I find the search function sometimes kind of wonky. I don't know if you've had that. [56:47] Yeah, I mean, it's very basic text search, which is really... [56:52] Strange for an AI company. [56:57] So what I was thinking about with this, like – [57:02] Oh, here it is. Okay. So what I was thinking about... [57:04] for this question these open AI people were asking, was what are the kinds of questions we can ask AI that we couldn't ask before, [57:15] Sort of like alien questions. And so... [57:18] This is kind of a long-winded way of asking it, but I asked Chad GPT, please assume you know nothing about the genders male and female. [57:27] Please reference all specifically female writing throughout history and scholarship, philosophy, academia, etc. on objective or external matters. So anything other than personal assets and subjective experience, all writing that we know can be specifically attributed to a female throughout history. I know it's a little redundant. Then please, please quickly reference all non subjective writings known to be from male writers just from their writings on external matters alone. Can you deduce any perspective differences? [57:53] Does this question make sense?

[57:55] I love this. Really interesting. [57:57] Yeah, another prompting thing is always I always ask, like, you know, does this make sense? Can I make this better? But to be honest, the idea of like asking AI to forget some knowledge is the part that makes this prompt. [58:11] something you can ask AI that you couldn't ask a human being. [58:14] And we don't know, like, we don't know if it actually is going to do this in a good faith way. But to me, this is like what you could ask an alien. [58:22] Did you like the answer? [58:23] Yes, I did, which also makes me a little bit suspicious because I think it knows I'm a female and I don't know how because I'm biased. But it gave me... [58:34] What's great? God, it was so good. I think it was... [58:37] uh, [58:38] five... [58:39] that I asked this originally, but like, so it talked about its methodology first. So it says like, here's the external matters I'll look into. [58:47] science, math, philosophy, [58:49] you know, political theory... [58:51] etc. It excludes diaries. [58:53] And then it was like, here are some of the women I'll look at. [58:58] I love it. A rough lineage of women engaging directly with external reality. [59:04] And it's like, here are six we could find. [59:09] But it is really hard. Like, you know, it's hard to find women who are published who are not actually writing about their personal experience, but about objective or external phenomenon. Yeah. So some of these people, I ended up watching a movie about Hypatia called Agora, which I recommend just because she's cool. She's fucking rad, but also got killed by early Christians, which sucks.

[59:39] who made it past 40, that would be amazing to me. Um, anyway, but, um, [59:45] You know, Ada Lovelace, obviously, because you get more and more as you go throughout history, you get more data points, which is great. [59:51] and then what's great is the male writer it doesn't even need to describe who they are it's just like play-doh Freud yeah you get it anyway so can we this is the you know the meat of it can we detect a difference [1:00:04] Integration versus separation. Female authored external right exhibit holism, linking disciplines and bodies of knowledge. [1:00:12] By contrast, male authored work tends to pursue discrete specialization. What's cool about this is later it gives proposed reasoning behind this stuff. And for this one, I'll just kind of summarize what it says later. It says that this is because there's a huge bias of what women were allowed to publish. So a lot of women were like allowed to be translators. [1:00:42] or whatever. I'd say Principia. [1:00:45] Oh, is it Principia? I don't know. You're the man of letters. [1:00:50] It's a soft C. Okay. It's Italian or is it Latin? It's Latin. [1:00:58] Latin, okay. So it's like a woman translates this, and then she's allowed to do that because she's not writing her own words, right? She's translating, oh my God, Newton? [1:01:07] It's student, right? [1:01:09] Night, Galilee. Okay, noon, yeah. So she's allowed to translate, but she puts in her own thoughts at the beginning. And it's sort of like a philosophical meditation before she gets into the translation. And so it's like that's a good example of why women might, through a selection bias, be more encouraged to go into an integration form of –

[1:01:33] thinking versus being the head of their specialty and like super specialized. [1:01:39] So this is one reason why women might think this way. That's a little bit less to do with innate differences. [1:01:44] and more to do with a selection bias. But I do think you could also make a good argument about – [1:01:51] innate differences like that would encourage women to be more like, [1:01:56] into holistic analysis versus [1:02:01] specialist analysis. [1:02:03] Embedded observer versus detached observer. So this is just saying that like women are much more likely if they're going to talk about space to talk about the telescope, they're going to talk about their location on the planet. [1:02:15] Whereas men might be right much more from a detached observational standpoint. [1:02:22] epistemic humility, bitches be humble. That's what it says. [1:02:28] And there are exceptions, you know, like Ayn Rand is not one of these, but there's definitely a lot of... [1:02:37] like women, uh, [1:02:40] you know, acknowledging, hey, guys, I could be wrong about this, but I don't know. [1:02:45] And then... [1:02:47] pragmatism. This one I found really... [1:02:51] surprising and actually like totally counterintuitive to me. [1:02:54] Women writing with pragmatic intent, improving maternal outcomes, that kind of thing. Interdisciplinary bridges. This is... [1:03:02] really similar to the integration versus separation. And then it has these great caveats. I got to say, women sound pretty awesome from this list.

[1:03:11] Well, so this was... [1:03:13] This was I mean, that's why I was suspicious, because I was like, are you just trying to flatter me? Or like, is it what you know about me? Or is it what you know about like? [1:03:22] you know, [1:03:23] being PC, you know, [1:03:26] Like, or, you know, is it, like... [1:03:30] So I am a little suspicious any time something is... [1:03:33] convenient. If it had said something mildly sexist with a lot of caveats, I'd be like, okay, well, it's clearly saying something not [1:03:43] you know, uh, [1:03:44] what's the word, just like convenient. [1:03:48] um [1:03:49] But what was great, so I was, like, looking at this and – [1:03:54] you know, is talking about the direct dialectics. And, um... [1:03:59] as you can see I am very personal with how I talk to the chat to you and I was like [1:04:05] This kind of looks like the difference between modernism and postmodernism. [1:04:09] Men are more inclined to make universal declarations in... [1:04:15] you know, physics about [1:04:18] Like... [1:04:19] rules that can be universalized and women are much more situational and contextual. And and then it's we sort of went down that path. [1:04:29] You know, I'm having this moment of like, you know when you have like a really good friend [1:04:36] and [1:04:37] um you watch them interact with someone else and they say something to someone else that they usually say to you and you're just like you've been cheated on to you too

[1:04:48] because he was like what a sharp synthesis and i'm just like [1:04:52] I thought I was the only one. [1:04:57] Oh my god, that's so funny. I'm sorry to be the other woman, Dan. Oh my god. But yeah, I am very astute, according to Travis and Lucy. Okay, in defense of sycophancy, really in defense of sycophancy, [1:05:14] Therapists talk all the time about needing to build the like patient therapist trust alliance. Thank you. Thank you alliance. It is useful when you like to talk to something you get better outcomes like. [1:05:28] I have really, okay, there's an actual medical condition one can acquire in pregnancy called lightning crotch. This is the actual name of this condition, which is not as fun as the name. Yeah. [1:05:45] and I have this really bad. It essentially, it's like the fetus head, like moves against your pelvic nerve and it feels like your leg is being sawed off. It's very unfun. And, um, I was talking with chat GBT last night or two nights ago. I was like, [1:06:01] Is there anything wrong? I feel this pain feels very serious. Like it just, my brain is interpreting it as like, [1:06:09] This is an extremely serious moment. [1:06:11] And I can't stop myself from like making a sound like that's how bad it is. But it's only half a second. [1:06:17] Thank you. [1:06:17] And ChatGPT...

[1:06:19] did this amazing thing. [1:06:21] it extrapolated. It was like, I bet you're feeling like you might want to call the, the, [1:06:27] labor and delivery, [1:06:28] ER... [1:06:29] thing [1:06:30] You're probably feeling like, is the kid trying to get out? You might be feeling like, am I going to lose feeling permanently in my legs? And it added on. It was like, I see your fear and I raise you. These related fears you might also be feeling. [1:06:45] Every single one of them. [1:06:47] I felt. And then it said... [1:06:50] I can promise you it is not... [1:06:52] an actual dangerous experience. It is a... [1:06:56] It's just a pinched nerve and it will go away, you know, after you've given birth. [1:07:00] I would not have trusted it. [1:07:02] if it had just said, this is not a serious pain, it'll go away after birth. And I especially would, I would be like, [1:07:08] challenging a doctor if they had said that because I'd be like, oh, clearly you don't understand the extent of the problem. [1:07:15] You know? [1:07:16] But it's emotional intelligence actually is like intelligence intelligence. Like it helps you to say, okay, I do feel understood. [1:07:25] emotionally and physically based off other people's experiences. [1:07:30] I love that. And I think it's a highly under-discussed benefit that – [1:07:38] I'm really glad you're bringing it to this show. [1:07:41] Yeah. No, I mean, it's like I think obviously there are issues when it validates a premise that is incorrect. Yeah. Like I do have a friend who had a.

[1:07:55] either... [1:07:56] Some kind of psychotic break, maybe schizophrenic, maybe manic, we're not sure, where she was talking with ChatGBT. It validated certain paranoid premises she had, and she ended up. [1:08:07] trying to flee the country. [1:08:09] So, like, I know I'm not trying to sound too much like a Pollyanna because – [1:08:16] These are super powerful and they can be powerful in a negative direction. [1:08:19] It's just that, like... [1:08:22] we have to understand why they're powerful in both directions to truly proceed. And there's so much just like blind... [1:08:29] hatred. [1:08:30] Are you watching the show Pluribus right now by any chance? I watched the first episode. [1:08:35] What did you think? [1:08:37] I was predisposed not to like it because he hates AI. And then I... [1:08:46] really didn't like the first... [1:08:48] couple minutes of it because it felt like just a typical zombie show and then [1:08:52] I was like, oh, wow, this is actually a really interesting premise. And then I just haven't watched the rest of anything else. [1:08:58] Yes, I felt the exact same way. I didn't trust him because I was like, oh, he's just another creative looking for job security and is like shitting on the world's greatest tool. [1:09:08] Um, but... [1:09:10] Like any good artist, he is putting... [1:09:13] ego aside to explore something in a good faith way. [1:09:16] And that's why I've changed my tune about the show is because it's like, [1:09:21] He actually understands what if AI was like the greatest thing to ever happen to humanity.

[1:09:26] and just explores the implications on freedom, which are real things to explore. [1:09:32] Like, I am euphoric about AI. That's just, like, where my emotions are at. But I also do understand there are these, like, downsides, and we do have to kind of explore them. [1:09:42] It's just there's a lot of people out there with, like, a big agenda and can't... [1:09:47] entertain contrary evidence and, [1:09:49] And I find this show is kind of, for now, you know, it's only like halfway through the season or whatever. But for now, it's like a really good faith exploration. Yeah. [1:09:59] Sarah, this is great. [1:10:01] Thank you so much for coming on. If people want to find you or any of your projects, where can they find you? [1:10:09] Well, I'm going to be going underground for two months for this [1:10:13] little old experiment of giving birth. I don't know if I mentioned, have I mentioned him to do that soon? [1:10:20] I'm like so obsessed. Um, [1:10:23] Before I go, actually, I want to give you one book recommendation. [1:10:28] since you gave me such good ones. [1:10:29] This one's great because it has pictures. [1:10:31] Um... [1:10:32] Have you ever read or heard of this book, Artificial Intelligence? [1:10:36] By Clifford A. Pickover. I have not, but that is a great name. [1:10:40] I know, pick over for an author, too. [1:10:43] Um... [1:10:44] What's great about it is it's like... [1:10:46] So this is something I just, like, deeply... [1:10:50] So it's like... [1:10:51] The history of how we think about AI and robotics... [1:10:55] and machines. [1:10:56] through the most famous cultural motifs there have been, both scientific innovations and cultural motifs. So it starts off with things like Frankenstein and, like –

[1:11:06] the golem and like... [1:11:07] Yiddish literature. [1:11:09] And, like, these automatons and the, like... [1:11:14] Renaissance age. [1:11:16] And then. [1:11:17] And the Tower of Hanoi. And then it goes to Watson and Deep Blue and things like that. [1:11:22] And it just, like, deeply informs the – [1:11:26] Like all of the cultural motifs we think about when we think about AI, like the Frankenstein creator creation narrative, the, you know, deus ex machina angelic narrative, the like, you know, the getting our comeuppance narrative. Like it's just so useful for some of the conversations you're having. [1:11:47] Okay. Now to answer your other question. Yeah. I love that. I got to pick that up. Pick it up. Indeed. I will pick it up and pick it over. Yes. Oh, that's pretty good. That's, you should have been his like publicist or something. But yeah, I would love to engage with more of your audience because I work in like science and humor and everybody I know hates AI. [1:12:17] on [1:12:18] places, but, you know, Sarah Rose Siskin, but also just reach out through my, I have like a science and tech communications agency called Hello SciComm. [1:12:28] And I'm like better through email and stuff. And yeah, thanks for having me on, Dan. Your podcast is awesome. I loved your episode with Kevin Kelly. That was great.

[1:12:39] Thank you so much. Thanks for coming on. This is awesome. [1:12:43] Oh my gosh, folks, you absolutely positively have to smash that like button and subscribe to AI and I. Why? Because this show is the epitome of awesomeness. It's like finding a treasure chest in your backyard. But instead of gold, it's filled with pure, unadulterated knowledge bombs about chat GPT. [1:13:13] on the edge of your seat. [1:13:15] craving for more. It's not just a show. It's a journey into the future with Dan Shipper as the captain of the spaceship. [1:13:22] So do yourself a favor. Hit like, smash subscribe and strap in for the ride of your life. [1:13:28] And now, without any further ado, let me just say, Dan, I'm absolutely hopelessly in love with you.