Podcast episode (Ep. 15) in which host Dan Shipper interviews psychiatrist Dr. Peter D. Kramer, author of Listening to Prozac, about how technology — from antidepressants to ChatGPT — changes personality, sense of self, and how we see the world. Topics include how antidepressants affect personality beyond treating negative emotions, how the availability of technological solutions reshapes perceived problems and categories, and practical experimentation with ChatGPT for writing, relationship advice, and prompting (specificity, tone), ending with Kramer's thoughts on ChatGPT as a therapist. # Channel Presentation Observations — "How Technology is Changing the Way We See Ourselves - Ep.15 with Peter D. Kramer" ## Explicit Self-Description ### Show Name and Host - The closing segment names the show and host directly: > "you absolutely positively have to smash that like button and subscribe to How Do You Use ChatGPT? ... with Dan Shipper as the captain of the spaceship." - The host identifies himself as "Dan" during the episode: > "There are probably some people who are like, has Dan gone crazy?" ### Stated Format / Scope - The host explicitly outlines the episode structure, indicating a recurring show format: - **Book/author discussion:** "what I want to do in this show is talk about your book cuz I think it's just really important" - **A recurring segment:** "we talked about we always do an explore section of this show" — followed by a live ChatGPT demo. - The channel's topical core is ChatGPT/AI, per the host's own framing of potential audience confusion: > "There are probably some people who are like, has Dan gone crazy? Like, what does this have to do with ChatGPT?" and the show title referenced in the outro ("How Do You Use ChatGPT?"). ### Host Self-Presentation - **Writer background:** "you're a writer, I'm a writer. Uh in writing summarizing is a lot of what you're doing in writing." - **Personal disclosure:** The host shares experience with OCD, antidepressant use, and Jewish/Eastern European family background (e.g., "Thank you to some of the medications that you you talk about in your book", "having OCD, I was... watching"). - **Prior work:** "I was writing an article about the FTX scandal" and "I've been sort of writing about that and thinking about about that inspired in large part by your work." ## Intended Audience - **AI/ChatGPT-curious viewers** who follow the show for AI content — evidenced by the host pre-empting audience confusion about a psychiatry-focused episode. - **Intellectually-oriented viewers** interested in the intersection of technology, psychology, and the self: the framing question "It's about how technology changed how we see ourselves and the world." - The closing call-to-action is written in exaggerated, hype language ("knowledge bombs about ChatGPT", "roller coaster of emotions, insights, and laughter"), suggesting a general-audience outreach tone. ## Claimed Scope of Content - **Explicit claim:** exploring how ChatGPT changes self-perception, using historical parallels from other technologies: "I think ChatGPT also has impacts on on how we how we see ourselves." - **Interview format with expert guests:** this episode features psychiatrist-author Peter D. Kramer ("Listening to Prozac," "Should You Leave," "Death of the Great Man"). - **Live AI demonstrations:** the "explore" segment includes real-time ChatGPT prompts (roleplaying as Peter Kramer and Albert Ellis, asking for relationship advice and whether antidepressants work). ## Observable Presentation Patterns (Demonstrated, Not Claimed) - **Long-form, discursive interview style** — wide-ranging conversation covering psychoanalysis history, lithium, diagnostic categories, before reaching the AI demo. - **Personal-testimonial integration:** the host weaves his own mental-health experience into the expert discussion. - **Skeptical-but-curious stance toward AI:** "if you just let it go off and do a bunch of stuff without checking it and you don't actually know the underlying stuff, you're going to get stuff that is wrong." - **Humorous, self-aware tone**, including the deliberately absurd AI-generated outro ("Dan, I'm absolutely hopelessly in love with you"). ## Claims vs. Demonstration Notes - The show's name and subscribe appeal are **creator claims** made in-script; the actual episode demonstrates a broader intellectual scope (psychiatry, technology history, philosophy of self) than the title alone implies. - No channel-wide subscriber counts, upload schedule, or mission statement appear in the source material; only the outro tagline characterizes the show ("the epitome of awesomeness... a journey into the future"). # Delivery Observations: Ep.15 with Peter D. Kramer ## Participants & Overall Dynamic - **Host (Dan Shipper)**: fast-paced, conversational, heavy use of filler words ("like," "um," "you know"), enthusiastic and personal. - **Guest (Peter Kramer)**: measured, anecdote-driven, frequently hedges with "you know" and "I mean," and structures answers as miniature narratives. --- ## Host Delivery Patterns ### Casual, Filler-Rich Register The host's speech is saturated with "like" as a discourse marker: > "Like, I read it like four times. It's incredible." > "Like a psychiatrist's uh interaction with someone who's kind of like Donald Trump, but you haven't really said that." This is a stable feature across the episode, not an isolated slip. ### Recurring Phrases & Habits - **"that makes sense" / "Totally"** as acknowledgments: > "Yeah, that makes sense." > "Totally. Totally." - **"I think" / "I feel like"** as framing for nearly every claim: > "I feel like ChatGPT is afraid that he or she is going to get sued." > "I think there's something really important underlying this book." - **Self-reference to personal experience** as a rhetorical anchor: > "I'm happy to share sort of my experience with them." > "And I'll say for myself, like having OCD..." ### Previewing / Agenda-Setting The host explicitly outlines the show's structure, a teaching habit: > "talk about your book cuz I think it's just really important... and then I want to take sort of like some of the lessons... and apply them to ChatGPT" He also uses a recurring segment formula: > "we always do an explore section of this show." ### Energetic, Hyperbolic Outro The closing is a distinct register shift — exaggerated promotional speech: > "you absolutely positively have to smash that like button and subscribe" > "It's like finding a treasure chest in your backyard" This is markedly different in tone from the interview body, suggesting a scripted/read segment (uncertainty: whether scripted cannot be confirmed from the transcript alone). --- ## Guest Delivery Patterns ### Anecdote-Led Explanations Kramer consistently answers abstract questions with concrete stories: - On the drug's origin: the **yellow sheets** in patient charts: > "the notes on those yellow sheets started looking with Prozac like my regular psychotherapy sheets." - On panic attacks: the **Donald Klein nursing-desk story**: > "they'd say, 'Well, I the nurses are not helpful anyway.'" This is one of the most repeated teaching moves in the episode. ### Hedged, Balanced Phrasing Kramer frequently qualifies and holds two sides of an issue: > "So, I But I'm play with that both ways." > "I mean, it's correct... But but I would like something more predictive." He self-describes this tendency directly: > "I'm so much an on the one hand and on the other hand guy" ### Signature Declarative Openers He occasionally opens with punchy, quotable assertions: > "So, first of all, antidepressants work. They're not glorified placebos." And humor-tinged enumeration: > "Yes, I you know, yes to about seven things." ### Vivid Analogies He uses accessible comparisons to make clinical points: > "they say there no double blind studies of whether you should use a parachute when you jump out of an airplane, right?" --- ## Interaction Patterns - **Host-directed prompting**: the host frequently hands the floor over explicitly: > "I'll pause there. It sounds like you might have something to say." - **Guest defers on technology**: Kramer repeatedly flags his outsider status: > "this is not my world. I'm uh not a great user of... Chat GPT" > "I'm impressed without, you know, changing my way of life here." - **Collaborative live demo**: both co-construct prompts to ChatGPT, negotiating wording in real time ("Okay, ask one question at a time"). --- ## Uncertainty Notes - Timestamps are unavailable in this transcript; delivery pacing (pauses, interruptions) cannot be assessed beyond textual cues like repeated words and "uh." - Whether the outro was pre-written vs. improvised cannot be determined from text alone. - Some patterns attributed to the host (e.g., "like" frequency) rest on many instances; guest "you know" frequency is likewise well-attested, but relative rates per minute cannot be computed without timestamps. # Position Extraction: "How Technology is Changing the Way We See Ourselves" — Ep. 15 with Peter D. Kramer ## Participants - **Dan Shipper** — host - **Peter D. Kramer** — psychiatrist, author of *Listening to Prozac*, *Should You Leave?*, *Against Depression*, *Ordinarily Well*, and the novel *Death of the Great Man* --- ## 1. Antidepressants: Efficacy and Cultural Standing ### Kramer's positions - **"Antidepressants work. They're not glorified placebos."** He reiterates this point multiple times, citing his book *Ordinarily Well* on the evidence. - He asserts we are at **"a low ebb for the cultural view of antidepressants"** — contrasting with the justified optimism of the late 1980s/early 1990s. - **Claim about trial design:** Because antidepressants are off-patent, cheap, and effective, drug trials recruit very marginalized people (SRO residents, prior non-responders) who receive substantial supportive contact. Result: they do well on placebo and no better on the antidepressant. **"It becomes hard to test the medicines when the medicines are so good."** - He invokes **Borges' *Pierre Menard*** in the 30th-anniversary preface: the same words read differently in a different cultural surround — the book's reception would now go "against the grain." ### Shipper's positions - Agrees emphatically: **"Having tried all the placebos, let me tell you, it's not a placebo."** (Personal testimony.) - Notes the book is **ambivalent in the classical sense** — presenting both benefits and reservations, and he values that. --- ## 2. Prozac Acting on Personality, Not Just Depression ### Kramer - **Origin of the book:** yellow sheets tracking medications and patient responses; notes began resembling his psychotherapy sheets. He came to think of **the medication "acting like a co-therapist."** - Patients said things like **"I'm myself at last"** and reported becoming **"better than well"** — not just recovered but improved (functioning better in "the dating marketplace," less push-around-able). - Traits patients attributed to psychological history (obsessions) were **muted on medication**, appearing instead like **"a biological glitch."** - **Cites a study:** paroxetine (Paxil) **did better against neuroticism than against depression**; it also affected **extraversion** — "it's easier to be extroverted if you're less worried about being slapped down." - **Qualification preserved:** On rejection sensitivity, the same trait has two faces — **vulnerable and perceptive**. The medicine obviates the second form too, and "we would want to be a little wary of that." ### Shipper - Emphasizes the book's central move: antidepressants **work directly on personality** — bolder, less rejection-sensitive, less obsessive — something "almost nobody knows." - Raises the question the book poses: after tapering, patients say "I'm not myself anymore" though not depressed — **what is a self to begin with?** - Personal example: his OCD traits (obsessive morality, book love) plausibly adaptive in an Orthodox/Talmudic ancestral culture, but not adaptive now. **"We can turn that down so that I'm more fit to the culture, but then you always have to ask, is that right?... That's an open question."** ### Kramer on melancholy - Migraine removal is unambiguously restorative; **depression is "culturally much more complicated"** due to a Renaissance-era tradition of **romanticizing melancholy** as special insight. He plays the question "both ways" — perhaps mild depressive sub-forms "really isn't who you are." --- ## 3. Explanatory Levels and the Lever Problem - **Shipper's formulation:** with a biological lever like Prozac, people reduce all explanations (psychological, social) to one level — "it's an imbalance" — even though the drug acting on one level doesn't prove that level is what's "going on all the time." - **Kramer agrees implicitly** and extends via the rejection-sensitive woman: her perceptiveness about a mediocre relationship may be real; the drug quiets both vulnerability and perception. --- ## 4. Technology Redraws Diagnostic Categories ("Psychopharmacological Dissection") ### Kramer's claims - **We see things we have solutions for more.** Panic attacks, OCD, and bipolar disorder all became more visible once treatable (Xanax; fluvoxamine; lithium respectively). - **Donald Klein's panic work:** tricyclics stopped panic attacks, but patients retained **anticoratory/learned anxiety** — a dissection of panic from anticipatory anxiety. Klein used this to **attack Freud's** view of anxiety as the central lever of psychology; treating anxiety did not cause symptom substitution. - **Lithium as "parachute" analogy:** no double-blind parachute studies needed; the dramatic prophylactic effect made the bipolar category undeniable and helped America adopt the English diagnostic distinctions over the Freudian neurotic/psychotic split. - **General thesis:** "The technologies correctly or falsely seem to be illuminating about how human experience ought to be subdivided." Medicines' specificity forces engagement with distinctions "in the literature forever" (e.g., Kraepelin-era schizophrenia/bipolar distinction). - Shipper adds the negative side of the Freudian era: **"schizophrenogenic mothering"** blame. ### Kramer's biographical grounding - Successful psychoanalysis in London (early 1970s) → medicine → psychotherapy-focused practice → accidental placement under **Jerry Klerman** in D.C. (after a Yale residency deal tied to a since-disappeared Carter-administration position), exposing him to the government research portfolio and Klein's conferences — positioning him ideally when Prozac arrived. --- ## 5. Extension to ChatGPT and Intelligence ### Shipper - ChatGPT will force a **"dissection" of intelligence**, a catchall word roughly defined as "anything an animal doesn't do" and core to human self-concept and knowledge-economy status — hence AI feels both promising and threatening. - **Summarizing example:** ChatGPT gave him three sentences on utilitarianism for an FTX/Sam Bankman-Fried article, saving research time. **Kramer's reaction:** "That seems like an ideal use where you know enough that you would know if a mistake was being made" — while also admitting "I do find this a little not useful. I may not be sophisticated enough at framing." - Shipper's observation: he now **sees summarizing everywhere** (emails, writing) and has **carved it out of his sense of self** — it's now a delegable task/resource, and intelligence becomes allocating that resource. - Analogy: using AI well resembles **management** — knowing when to delegate and when to be suspicious. ### Kramer - **"I'm not a great user of ChatGPT."** Helpful side: people with dyslexia/attention problems gain access "and it doesn't feel crutch-like... it feels like you're just entering into modernity." Unhelpful side: we feel threatened that "what is core and unique to us is something machines can do better." - He reports asking ChatGPT for a plot outline resembling his novel; the output was "so routinized and obvious and no one would want that book." --- ## 6. Live ChatGPT Demonstration (Relationship Advice) ### Setup - Kramer poses a scenario: 29-year-old woman, rejection-sensitive, 6-month relationship feeling one-sided, wants marriage and children — should she stay, and for how many more weeks? ### Kramer's assessments (preserved verbatim where possible) - On the first generic answer: "pretty much the advice you would get from a glossy women's magazine... But it is correct. It's captured the cultural consensus... You're not going wrong in any way." - After the boyfriend-dismissal exchange: **"I feel like ChatGPT is afraid that he or she is going to get sued. This is very namby-pamby."** Yet "it is remarkable to have something that captures standard advice in this way, and I don't disagree by the way." - On ChatGPT's "4 to 6 weeks" deadline: **"I'm not unhappy with it, but it's kind of a short leash"**; "nobody thinks that people change over the course of 4 to 6 weeks." - **His stated wish (unresolved tension):** an AI that has read thousands of diaries/biographies and gives actuarial answers — "letting this go on works 23% of the time. How are you with a 23% call rate?" He acknowledges ChatGPT **"isn't calculating those statistics. It's just predicting the next word in a sequence."** He fumbles for the term ("a logit," per Shipper) and says **"I would like it to have the information."** - On the Albert Ellis simulation: captures rational emotive therapy's content but **not Ellis's personality/irascibility** — "It's too long ago." Final verdict: **"This is really impressive compared to anything from a few years ago, but... I wouldn't use it... I'm impressed without changing my way of life."** ### Shipper's assessments - The more specific the simulation instructions, "the less it sounds like Wikipedia." - Shares his own technique: feeding in journal entries and asking ChatGPT to **write the next entry assuming decision A vs. decision B** — using its gestalt-of-text strength. --- ## 7. Closing - Kramer: the discussion captured "what's at the core of the book, which is technology and the self." - Shipper: the book "literally changed my life" (read four times, dog-eared copy). - Outro read in a deliberately hyperbolic ChatGPT-generated promotional voice. --- ## Key Unresolved Tensions (Preserved) - Is medicating away melancholy/obsession restoration of the self or culturally-driven conformity? — explicitly left an **open question** by both. - Kramer impressed by ChatGPT yet **declines to use it** — no smoothing of this ambivalence. - Kramer's desire for data-driven predictive advice vs. the technology's actual next-word-prediction mechanism — gap acknowledged, not resolved. # Speaker & Provenance Observation Record ## Video Metadata - **Title:** How Technology is Changing the Way We See Ourselves - Ep.15 with Peter D. Kramer - **Video ID:** 01a0f15b-3047-7090-bfc8-ada488538df1 - **Curated metadata:** None supplied beyond transcript. ## Speakers & Roles ### Primary Speakers - **Dan (host/creator)** — Identified by name in transcript ("has Dan gone crazy?", "Dan Shipper as the captain of the spaceship"). Hosts the show, interviews the guest, shares personal experiences with antidepressants, OCD, and ChatGPT use. Self-identifies as a writer and philosophy graduate. - **Peter D. Kramer (guest)** — Identified via title and in-show introduction. Roles stated in transcript: - Psychiatrist - Writer/author - Novelist - Former psychoanalysis patient in London, early 1970s - Former columnist for a psychiatry trade paper - Held a government community psychiatry position in D.C.; worked under Jerry Klerman ### Other Voices - **ChatGPT (AI system, non-human speaker)** — Outputs read aloud on-screen during a live demo segment, including responses role-playing as Peter Kramer and Albert Ellis. Attribution of these outputs is explicit in the transcript. - **Unnamed promotional voice** — Delivers the closing like-and-subscribe segment, naming "Dan Shipper" and the show "How Do You Use ChatGPT?" ## Referenced Works by the Guest - **Listening to Prozac** (with 30th-anniversary edition and new preface/afterword) - **Overcoming Depression** - A book about Freud - **Death of the Great Man** (novel) - **Should You Leave** (third book) - **Against Depression** - **Ordinarily Well** ## Third-Party Claims & Sources Cited in Transcript - **Donald Klein** — psychopharmacologist; cited for the panic attack/anticipatory anxiety "pharmacological dissection" and critique of Freud. - **Jerry Klerman** — described as a great innovator in pharmacology; Kramer worked in his shop. - **Tracy Kidder** — cited as a nonfiction model ("birth of a new machine" / PC revolution). - **Verlyn Klinkenborg** — book on historical haymaking methods. - **John McPhee** — book about making orange juice. - **Jorge Luis Borges** — "Pierre Menard, Author of the Quixote" (referenced in the new introduction). - **Freud** — discussed as historical framework for anxiety and diagnosis. - **Peter Singer** — mentioned by Dan as utilitarianism reading for an FTX article. - **Sam Bankman-Fried / FTX / effective altruism** — referenced by Dan as article subject. - **Albert Ellis** — psychologist invoked during the ChatGPT demo (REBT/rational emotive therapy). - **Paxil (paroxetine) study** — cited by Kramer: medication did better against neuroticism than depression; also affected extraversion. Study name/source not given. - **Lithium and bipolar disorder; fluvoxamine and OCD; Xanax and panic attacks** — historical treatment-diagnosis claims made by speakers. ## Personal Disclosures (Provenance-Relevant) - **Dan:** identifies as having OCD, Jewish/Eastern European family background, personal antidepressant use, studied philosophy in college. - **Kramer:** patient experience with psychoanalysis; 40-plus-year marriage; career moves from Yale to D.C. ## Promotional Segments — Flagged - **Closing segment (end of transcript):** explicit like/subscribe promotion for "How Do You Use ChatGPT?" including hyperbolic praise and a personal declaration toward Dan. Delivered by an unnamed voice, not the interview discussion. - No mid-roll or sponsored-product promotion detected in transcript. ## Availability Limits & Uncertainty - No curated metadata was supplied; all identification derives from the transcript itself. - The show's producer/network is not identified beyond the closing promo. - Guest's full name spelling ("Peter D. Kramer") is taken from the video title. - Paxil/neuroticism study is described as "well-known" but not named — source unverified. - **Comment availability:** not stated anywhere in the source material. - Speaker turns are not timestamped in the transcript. So, first of all, antidepressants work. They're not glorified placebos. Yes, I think we are in a sort of low ebb for the cultural view of antidepressants. There are probably some people who are like, has Dan gone crazy? Like, what does this have to do with ChatGPT? It's about how technology changed how we see ourselves and the world. I feel like ChatGPT is afraid that he or she is going to get sued. It is remarkable to have something that captures standard advice in this way. This is really impressive compared to anything from a few years ago. Peter, welcome to the show. Thank you. Hi. I am so excited to have you because you're one of my favorite authors. You wrote a book that like literally changed my life. I have it right here. If you can see, there's like so many little dog-ears on the in this book. Like, I read it like four times. It's incredible. I I love that. Um so, for people who don't know you, you are a psychiatrist. You're also a writer. You're the author of the book that I just held up, Listening to Prozac, which you wrote 30 years ago, but you're also the author of many other books um like Overcoming Depression. Uh you wrote a book about Freud, which I also read, which I really enjoyed. And you're also uh a novelist. Remind me again of your the title of your most recent novel. Death of the Great Man. Death of the Great Man, which is sort of it feels like uh you're talking a little bit about like a psychiatrist's uh interaction with someone who's kind of like Donald Trump, but you haven't really said that. Yes, a psychiatrist is coerced into treating a authoritarian, buffoonish, uh autocratic national leader in his disastrous second term. Well, let's hope it doesn't come true, and uh it's sort of like Sopranos-esque but uh the president instead of a mob boss maybe. Um so I love that. I love how sort of multi-disciplinary you are. Um you know, like uh one of the things that about psychiatry in my experience is with psychiatry, which we can get into in a second, is uh psychiatrists often are like they're sort of like brain like mad brain scientists or whatever and they just sort of like give you magic chemicals that that my experience that where they're like, "Oh, they don't know how it works. We don't know how it works but like it might make you feel better but it makes you feel worse in in the meantime or whatever. Um and therapists tend to be a lot more like interested in humans and like the human experience and stuff like that. And I think uh it's like such a shame in the in the field because like there's that divide and you seem to just be in this like really nice like both very humanistic, very interested in human experience and into the into the brain chemicals thing that uh that I really love. Yeah, so I came to psychiatry via I suppose my experience as a patient. I was in a very successful, I think anyway, very compelling psychoanalysis in London in uh the early 1970s and decided then to become a doctor in order to become a psychiatrist. And um had some other therapies good and bad but uh became really a student of psychotherapy in that time in London. I read a lot of Freud and some of the other kind of founding psychoanalysts. And then in medical school became interested in a range of psychotherapies that with the early 70s was a time of the flowering of psychotherapies. There were some related to family and some to social justice and some to paradoxes. There were just interesting therapies which I write about in my third book, wrote about should you leave. Um but So I So when Prozac appeared, I was really someone focused with my patients on psychotherapy and seeing their responses. Now, there's a 30th I should run and hold it up, I suppose, but there's a 30th anniversary edition of Listening to Prozac that's out um just the end of last year. And In and I got to write a new preface and afterwards about what held up and didn't hold up, but I also write about how I came to write the book. And I used to keep these uh yellow sheets of paper for easy identification in patients' charts where I wrote down the medicines and doses, you know, that I prescribed. And I wasn't prescribing for most patients. But also their responses, if there were side effects and things. And those you know, the notes on those yellow sheets started looking with Prozac like my regular psychotherapy sheets. People were talking about sort of neurotic symptoms that had been muted. And I started having a thought that the medication was acting like a co-therapist, that it was doing some of the psychotherapy. So that's really in a way the origin of the book. Yeah, that makes sense. And I should I should I think I like dove right in right off the deep end into like psychotherapy psychiatry stuff. And I should actually probably introduce for the people who are listening and watching why you're on the show at all, because I think um there are probably some people who are like, has Dan gone crazy? Um like what does this have to do with ChatGPT? And the answer is I'm not as crazy as I used to be. Thank you to some of the medications that you you talk about in your book. Um but also uh I think there's there's something really important underlying this book. And And I'll tell people like a little bit about about the book and a little bit about like the sort of the theme. But the theme of the book is it's about how technology, in this case Prozac or in the larger class of drugs we call antidepressants, changes how change how we see ourselves and the world. Um and I think you sort of set out to write what you call this like resonant portrait where you're um looking at how technology changes people, changes culture, changes the changes the way that doctors see people, changes the way that that people see themselves. Um and there's a sort of important tradition of people doing that. So, another person that you've brought up before is uh someone like Tracy Kidder who did the same thing uh with the PC revolution. So, it's not just about like obviously antidepressants change how we see ourselves in a in a certain like more obvious way where it maybe it blunts at negative emotions or negative affect. Um um but uh I think one of your points is that technology in general can do that. And um I see that with ChatGPT. I've been sort of writing about that and thinking about about that inspired in large part by your work that I think ChatGPT also has impacts on on how we how we see ourselves. Um And so, what I want to do in this show is talk about your book cuz I think it's just really important. Um talk about antidepressants because I think that people don't talk about them enough in the right way in in public. Um I think your book does it does it really well and I'm happy to share sort of my experience with them. Um and then I want to take sort of like some of the lessons or some of the things that you pull out of the things that you notice about how Prozac and other antidepressants change how we see ourselves, change the human experience, and apply them to ChatGPT cuz I think that there are some like really interesting parallels. Um and then we're just going to explore ChatGPT together and we have a we have a fun idea for for how to do that. Um so, that's the like that's the basic outline of the of the show and why you're on the show. Um And um and I'll pause there. It sounds like you might have something to say. No, I really appreciate that summary because this book was exactly about uh a certain moment in contemporary time, modern time, 30 years ago, and how I thought our sense of self was going to change, and I was really interested in writing fiction. I chose my editor because she had um edited fiction. I had based on the outline of the book, I had a choice of a few publishing houses. And I looked then at nonfiction for models, how I would write nonfiction that had some of the flavor of fiction, and that Tracy Kidder book about the birth of a new machine seemed to me have that quality of focusing on a little change in technology and how it give you a sense of what's going on in the wider culture. And oddly, Verlyn Klinkenborg's book about using historical means of making hay and changes in how hay is made and John McPhee's about making orange juice. I mean, it what it wasn't all high technology. Just you know, I think that there are lots of technologies that really change the environment, and I thought, you know, I had hold of one. And my patients said that. So, I before I wrote a book, I was writing a monthly column for psychiatrists in a trade paper, and I was saying to my colleagues, "You're seeing what I'm seeing with as these new antidepressants came into use." And one of the things was that my patients said things like, "I'm myself at last." And I thought, "Really? You And you would never become yourself if this medicine hadn't been invented?" Or they thought that some aspects of themselves that they thought arose from their history, their psychological history, how obsessed they were with this and that, that those aspects just became muted, as you say, on the medicine. They were less obsessed with whatever it was, and it seemed that that had been kind of a biological glitch rather than uh you know, a developmental theme. And um I thought, "Oh, we we should think about how that might be happening." Yeah. Yeah, I think that's that's the interesting that's the one of the most striking things about your book is like um when people think about antidepressants, one is maybe they don't really even know what they do, but they're like, "I guess generally they like help with depression." Um or generally they help with like all the whole range of psychological issues that people tend to have like depression, anxiety, OCD, which I have. Um Uh they all of that stuff they they they tend to help with. Um and I think one of the points that I took away from your book is like, "Yeah, like to some degree like they do operate directly on something like depression. Maybe they maybe they do um curb negative affect or if you're if you're feeling really depressed or feeling really anxious, like those things will be blunted a little bit." But they also, and this is something that like almost nobody knows despite how popular your book was, um they work directly on personality. Um and so what what they may do is they may um like tweak some of these some of these things about people's personality. Like you'll be a little bit bolder. You'll be a little bit less sensitive to rejection. You may obsess a little bit less. Um and one of the interesting things about your book is that it doesn't deal with people who are super super super sick who um become like start to be able to function normally. You you often focus on people who like definitely have problems, but they're they're fairly mild. And when they take Prozac or uh Lexapro or Zoloft or any one of these antidepressants, they become what you called um better than well. Um they actually like become the the issue resolves, but they also become like better versions of themselves. And you you have this like really striking anecdote at the beginning of the beginning of the book which you already referred to, which is like you might put a patient on on um on Prozac for a while to resolve a depressive episode. And they'll, you know, they'll become bolder. They'll become less sensitive to rejection, all that kind of stuff. And that will have a lot of market changes in their lives. And when you taper them off, they come back to you and they're like, well, I'm not myself anymore. And even though they're not depressed. Um and so the question the question that you pose is like, well, what is a self to begin with? Like you weren't yourself for the first 30 years of your life before I put you on this drug. Um but it but the drug just brought out this other other thing from you that was maybe always latent or always there, but like it needed to turn those dials to to to let you to let it come out. Right. I think it brings up the question of the natural. So, if you had were subject to migraines and really they were untreatable, you just were living the whole of your life with pain in your head and it was affecting you and your job and your relationship to other people. Uh pain was constantly intruding and then um you took a medicine that finally prevented migraine. Uh when was invented, we we would say, yes, you were your yourself at last. We would agree that the migraine was some kind of an external thing imposed on you. Depression is culturally much more complicated. So, there's a long history. I write about this in a book called Against Depression. A long history of romanticizing melancholy that goes back certainly to the Renaissance, probably before, but is very prominent in the Renaissance where melancholics are thought to have some special insight into the human condition. So, taking that away, you know, isn't as automatically restorative as taking away a migraine. But maybe it should be, right? I mean, maybe depression has some of that same illness quality to it. Um and even in these very minor forms, sort of sub-forms. Uh maybe that really isn't who you are. So, I But I'm play with that both ways. I mean, the book begins with what patients said to me, and that was what they said. The better than well comes from a patient saying, you know, I didn't just get over that episode of depression or obsessionality. I'm really able to function in the dating marketplace better. I'm more I'm you know, less easy to push around that I'm more certain about what I'm looking for. Totally. Yeah. Um and uh and I will say for myself, I I have experienced that. Um and and yeah, it's been it's been a crazy change for me. Um I I think you know, one of the things I should I should point out, which you sort of alluded to, is that one of the things I love about the book is it sort of came out at this like high water mark of antidepressant optimism, where everyone was like, "Oh, maybe it'll cure everything." And then I feel like right now we're in this like low of like it's been probably over-prescribed to people who like it aren't haven't you know, been properly educated or aren't in psychotherapy or whatever. And um and so, I think there's a lot of people out there who are like, "Oh, maybe it doesn't work. Maybe it's placebo." And like it's definitely it's not a placebo. Um Having tried all the placebos, let me tell you, it's not a placebo. Um but I think one of the things I love about the book is that um uh is you're kind of ambivalent about it in in the sort of classical sense of ambivalence, where you're like you're presenting the case, but you're like and there's some good things about this. It really does amazing things for people, but it's also not all good. Um and and we shouldn't be like too glib about like jumping into changing all aspects of our personality, in particular because like you said, you know, melancholy, it's a it's that's sort of a cultural thing. Um like whether or not melancholy is acceptable. And I'll say for for myself, like having OCD, I was I was watching and my history is like I'm from I'm I'm Jewish and my my family's from like Eastern Europe. And there were there were definitely like very Orthodox Jews and like I was watching a YouTube video the other day of like how an Orthodox Jew conducts their day and like it's pretty obsessive, you know? Like you're saying blessings about everything. You're like you're you're touching you're touching the mezuzah. You're you're you're rubbing the tzitzit or whatever. Like Um like and I or I love books like and I have a very strict sense of morality and that's one of my sort of obsessive things is like this sense of of morals and like if your ancestors studied Talmud all day, you might be obsessive, too. But I'm not in that culture anymore and so those features were like not particularly adaptive to me and um and so I think one of the points of your book is that uh we can turn that down so that it's I'm more fit to the culture, but then you're you always have to ask like is that right? Like should I need to do that? And that's that's a open question. Yes, I you know, yes to about seven things. So first of all, antidepressants work. They're not glorified placebos. I wrote another book called Ordinarily Well about how antidepressants work and what the evidence is. And yes, I think we are in a sort of low ebb for the cultural view of antidepressants and in that new introduction to Listening to Prozac, I write about Jorge Luis Borges' story Pierre Menard, author of the Quixote. So uh uh Borges imagines an author, made-up author, Pierre Menard, who writes one and a half chapters of the novel Don Quixote in the 1910s. All right, so we're in World War I. Uh and you know, war is horrible. And in the Quixote, in the chapters, uh Cervantes is glorifying war. And uh Borges says, "How much more brilliant Menard's version is than Quixote's version?" Because the same words are written in a very different cultural surround. And is it ironic? And uh you know, how's he making us think about war? I mean, it's a very clever uh metaphor trope. And I thought, "What Because I was asked by Penguin to do a 30th anniversary edition introduction afterward, what would it be like to write exactly these words that were in Listening to Prozac uh in 2023 when we're so negative about antidepressants?" And I don't know that that would be, you know, brilliant or not brilliant, but it you it would definitely be going against the grain. Um whereas I think in uh the 19 late 1980s, early 1990s, I was writing this book, there was a lot of optimism about the new antidepressants, and justifiably so. And by the way, I think one reason we are able to worry about whether antidepressants work or not is that, you know, here are these off-patent drugs, you know, pennies a pill, um that really treat depression quite well. So, that when you go out to do a study, you are really recruiting very marginalized people, you know, people living single room occupancies, who have done poorly on the medication in the past, who have no social context, and you're pulling them into a drug trial. It's very involving, where they have lots of contact with other people. Uh they get a ride from their, you know, SRO to the uh welfare to work program. Uh they get medical care. And uh yeah, they'll do well on a dummy pill because they're they're being given so much support. And yeah, they're not going to do that much better on a antidepressant because A, maybe they don't have the diagnosis in the first place and have sort of exaggerated things to get in the study and B, or the interviewer has exaggerated things for them. Or uh you know, B, maybe they're someone who just hasn't done well on a series of medications, so that you're selecting a population that won't do that much better. So, you know, becomes hard to test the medicines when the medicines are so good. Yeah, that makes sense. Um So, what I mean, I missed about four points there, but those were were a couple of them. That is that's totally fine. Um I think what I want to get into a little bit is some of the like broader lessons about how antidepressants um shape how we see ourselves and also shape how we see the world cuz I think that's the is going to be most relevant in our discussion of ChatGPT um in a second. And the first one um that that I take away is it's comes from the title, which is listening to Prozac and listening to drugs. And so, the thing that I think that you mean by that is when someone has something like depression or anxiety or OCD, there are always different explanations for what is going on. Um so, there's explanations at the psychological level like a lot of people think that depression is like repressed rage or something like that. Then there's like explanations at the social level. It's like, "Yeah, you don't have a job. Like, of course you're going to be depressed or your wife left you. Of course you're going to be you're going to be depressed." And I think one of the things that you noticed in this listening in in listening to Prozac is that when you get a lever like Prozac that operates on the biological level, you tend to forget all of the other levels of explanation and you just say, "Well, it's an imbalance." Or "Well, it's like this one thing." It is true that Prozac is is sort of operating on that level, and it does change all of these other levels. But but it's maybe not the best thing to just like jump to, "Well, it that must mean that's what's going on all the time." Right. So, I think one good example, you mentioned sensitivity to rejection. And that's a rejection sensitive dysphoria was a category that was in psychiatry briefly, and it's not much researched now or talked about now. But there are people who when they're rejected go into mini depressions. They go right off the edge. So, I think I write in Listening to Prozac Prozac about a woman who, you know, if her she asked her boyfriend a sensitive question and he turned away to look at the football game, would, you know, just start doubting everything. And these medicines turn out to be really pretty good at preventing that, at making people less immediately sensitive to uh perceived loss. But the question is, what's going on with that woman? You know, she's sensitive and she's sensitive maybe in two senses. One is she's vulnerable in the way we just said, and the other she's perceptive. So, you know, she comes into the office and says, "I seem to be in a good relationship, but I'm depressed." And we say, "Well, maybe you're seeing something in the relationship that, you know, you're right about that. It isn't as good as you think." And in a way, the medicine obviates that second form of sensitivity. And we would want to be a little wary of that. I I write about Prozac as a co-therapist, that it seems to do part of the work against the very sorts of conditions that we think psycho- used to think psychotherapy was very good for, like neurosis, you know, what used to be called neurosis, make people less neurotic. And it And it turns out that's right, that probably that is one of the things that these medicines are best against is there's a technical term neuroticism, which is a little different from neurosis, but in one study well-known study, these a medicine, paroxetine, Paxil, did did better against neuroticism than it did against depression. And the neuroticism meant this uh quick sensitivity to loss and uh fussing about things and long-term pessimism and so on. But that really changed very notably on uh medication. And of course, you know, in psychotherapy, that's very helpful, but you don't want to lean on it too much. Right. And I think um one of one of the points that you've made is like yeah, if it's operating on neuroticism, which if if you're if if you're familiar with some of this, you know, psychology here, like neuroticism is one of the big five personality traits. Right. Um and um and if you can operate on on neuroticism, what's kind of interesting is there's all these other aspects of your personality that might change if your level of neuroticism goes down. Um because you may be doing lots of things behaviorally to avoid avoid sensitivity, avoid the feeling of loss that like you think are part of your personality, but really aren't and like or it's hard to say whether they really are or not. But um but what we can say is that if the neuroticism goes away, all those other things will go away, too. And that's like a really interesting thing. And the other thing in this study that uh Paxil worked on was extraversion. And you think about this, those are not exactly opposite sides of the same coin, but it's easier to be extroverted if you're less worried about being slapped down. Totally. Totally. Um and and I think, you know, we're we're we've sort of covered the like it changes how you see yourself, but I want to talk also about like how how it changes how we might see the world. Um, and so when I talked earlier about um uh, antidepressants being this uh, this lever and when you have that lever you sort of um, and you can move the world in a particular way, you reduce your explanations to that to that layer. It's like we can move we can move people's depression or move people's state with Prozac, therefore you know, it must have been a biological issue and that's not necessarily true. One of the other ways in which like using technology as a lever changes how we see things is you will see more opportunities to use that lever. Um, so one of the examples that I love in your book is you talked about how when you were in um, residency um, nobody like came in complaining of panic attacks and you didn't really like have a lot of training about panic attacks cuz they weren't like a super common issue. Um, but but what's interesting is that as soon as Xanax became available, which is like the the of just a very effective um, a treatment for panic attacks. Obviously you don't want to take Xanax long-term. At least at first it seemed like a really good treatment for for panic attacks. Panic attacks became way way more common. We started to see them everywhere cuz they were suddenly treatable. And I think that I I would love for you to talk about that cuz I think it's it's just a general phenomenon that that we can apply to ChatGPT, too. No, I think we really do see things that we have solutions for more. Um I mean, the panic attack story is very interesting because Donald Klein, who was this great conceptual psychopharmacologist uh, was out to treat panic attacks and he they were very anxious people who were so anxious they were hospitalized although the bar for hospitalization is probably lower in those days. So, we're going back to the uh 1960s, say. And they gave the older antidepressants. Uh these were called tricyclics and they were started becoming available in the late 1950s, but really in the 1960s. And they they would be these patients who were very anxious and they were always going up to the nursing desk and saying, "You have to help me. I'm you know, I'm feeling so terrible." And they put them on the uh antidepressant and uh the nurses said, "These patients are doing better. They're not coming up and uh pleading with us." And uh Klein would speak to the patients and they'd say, "No, the medicine is doing nothing. I'm still anxious." And he'd say, "Well, why aren't you going up to the nursing desk?" And they'd say, "Well, I the nurses are not helpful anyway. They weren't very helpful anyway." And he based on this made this pharmacological psychopharmacological dissection, which was that the uh medicines were stopping the panic attacks, but the patients had this learned anxiety. They had what he called anticipatory anxiety, where they were so worried about panic attacks that they were sort of anxious about being anxious and they hadn't learned that they weren't going to get panic attacks while they were on these medicines. And you had to do something else to make them aware of the state that they were in. And I thought that was really smart. But, he also used these observations as an attack on Freud who thought that anxiety was sort of the central lever of human psychology and that things like castration anxiety, you know, which we don't give a lot of mind to now were uh the basis for internal conflict that created people's personality. And uh Uh, you know, Freud Donald Klein was more on the biological happenstance side. He was like, "Yeah, you could You actually can treat this anxiety and symptoms don't pop up elsewhere because there's this inner conflict. When you treat the anxiety, people do pretty well all together." And I, you know, that formulation was just coming into place before I wrote Listening to Prozac. That was part of my education. So, I have this funny happenstance in my own life, which is I fell fell in love with a wonderful woman who happened to be living in Washington D.C. I was at Yale in New Haven. And I've been married to her for 40-plus years subsequently, so I was not wrong about this. But, um I, you know, went to my residency director and, um said, "I'm going to leave the Yale residency and take up a residency in D.C." And the guy said, "Well, we'll get extra work out of you for half a year if you do this, that, and the other." And you can imagine being in residency and doing time and a half in residency, which is already pretty demanding. Uh, he would give me, uh the other final semester, whatever you call it, under the, uh, auspices of a Yale professor who was in the government in the Carter administration in D.C. So, I took up that offer, moved to D.C. in the I was very interested in community psychiatry, and this was a community psychiatry position nationally. And, uh, that position disappeared and I was, uh, put in the shop of, uh, Jerry Klerman, one of the great innovators also in pharmacology, who had me kind of become aware of the whole research portfolio in government. And it included holding conferences where, uh, Donald Klein held forth and so on. So, I came to, uh, learn about modern pharmacology with this interest in psychotherapy and community psychiatry, which put me in this really ideal position when uh Prozac came out and I decided to write this book because really I had familiarity, deep familiarity with an enormous range of contemporary research with while being the kind of person who had no interest in that. Right. And the thing I want to push you on is like like this claim that is in your book that I just sort of repeated and I want I wanted you to give people a little bit more color on it, which is that like panic attacks only became commonly thought about when Xanax was popularized and Xanax could work. Um By the way, the same is true of OCD that really OCD, obsessive-compulsive disorder, was thought to be very rare. It was only extreme cases that were recognized. And as antidepressants became more a mainstay of treatment or as you could rely on them a little bit and there was one called fluvoxamine. Prozac's fluvoxamine, fluvoxamine is a little different medicine. It seemed to do well for OCD in the early going. It had some side effects that people didn't like. Um you know, all of a sudden people started seeing OCD more widely and it turned out if you developed a rating scale or questions and to ask patients and way of writing rating answers, you saw that in the same population you've been treating for years, there was much more OCD than you'd ever recognized. But the the And the same is true for manic-depressive illness also, I think. The fact that lithium was helpful for uh bipolar disorder, it took a while, but eventually you know, America was very Freudian and we divided the world into neurotic and psychotic, but without much differentiation between schizophrenia and bipolar disorder. England was a little more uh divine willing to divide things. But we took on the English diagnostic system because it made sense to think about using lithium in bipolar disorder. So there's all this, you know, what we call psychopharmacological dissection that when you have a medicine to treat something, you start defining it. What do you think that's about? Like why? Well, you know, I think when the only treatment you have is psychoanalysis, it doesn't really matter what people's diagnosis is. Medicines sometimes have a level of specificity that really forces you to think about things that parenthetically have been in the psychiatry or psychology literature forever. So, you know, the that distinction between schizophrenia and manic depression, you know, goes back hundreds of years, anyway, many dozens of years. And I mean, it goes back to the 19th century. And um uh we say, "Oh, those early diagnosticians were right." You know, when it turns out that the distinctions they made are useful for the application of these technologies. So, I think that, you know, that's sort of the theme we're on. The technologies correctly or or or falsely seem to be illuminating about how human experience ought to be subdivided. Yeah, that's that was the thing that I I really want to talk about is like that I think is so interesting is that new technology changes how we draw categories or changes how we draw lines. And some of those lines are about how we diagnose things, but some of those lines are just are different. They're they're other things that that that that we we can get into. Like for example, what is intelligence? Is Is a question that is like really important in when it comes to ChatGPT and AI. And I think it's causing us to redraw those will cause us to redraw those lines a bit. And I'll tell you a little bit more about why, but but to go back like one of the things that you've brought up already that you you you you put into your book that I think is so interesting is yeah, this um there's this the way that lithium as a drug because it could treat um one specific disorder or it could treat a a set of symptoms. It it helped us sort of like pull those out of a mess of things and be like this is a separate category. So like as you noted um before lithium manic there it was very hard to differentiate between manic depression and schizophrenia because in both cases you may have like sort of psychotic symptoms. um and um and and in general manic depression manic depression depression and schizophrenia were all thought to be like under this big neurosis umbrella and caused by the same kind of intrapsychic conflict type stuff. um And that was like based on Freud and psychotherapy and this sort of long history of it which had many many maybe positive effects but also many many negative effects like blaming mothers for schizophrenogenic mothering and like all that kind of stuff. um And what happened with lithium is um by mistake we figured out that lithium um uh can cure manic or can really help manic uh manic depression. Um but it doesn't help in schizophrenia. So we suddenly had this new category of manic depression that we we could say was like a real category in the world because of lithium without which it would have been manic depression was just lumped in with all these other things. Right. So you know, it's interesting when a technology is very powerful. And I think that is true of lithium. Uh you know, they say there no double blind studies of whether you should use a parachute when you jump out of an airplane, right? There's just an enormous difference between having a parachute and not. And lithium was kind of like that. It's not that lithium isn't helpful in all sorts of circumstances. It may be. But the fact that you could give lithium to people who had this terrible recurrent disorder where they were going, you know, from mania to deep depression and so on, that you could give it and really prevent, you know, the next episode or make it much more distant than it had been in the past. That was very dramatic. And you know, like it's very hard to deny that there is something out there if you can prevent it in that fashion. Mhm. Yeah, that makes a lot of sense. Um and I think that's the that that sort of is the linchpin that I think can can bring us into chat GPT um chat GPT land because lithium created this what you call a pharmacological dissection. It created this this category of of manic depression. And I think chat GPT is going to sort of force us to um do something similar with intelligence. So, intelligence is this like big catchall word that has many many many things underneath it that we haven't really had to examine because in a lot of ways the way that we've defined intelligence has been anything that an animal doesn't do. Cuz it's like the uniquely human trait. Um it's like uh you know, that's that's sort of like how what Socrates said like makes makes men different is is our is our is our intellect. Um and um from animals and and therefore like the life of man that's best is the one that's lived according to the to the intellect. Stuff like that is like a very common um thing. And intelligence is like super core to our sense of self. Uh and in general like where we live in a knowledge economy where uh what you know and how you can bring that to bear has a has a big relationship to your status in the world. Um And I think that's why uh AI is promising, but also like very threatening to people because it like gets right at the at this word intelligence that I think a lot of people feel is very core to what makes humans unique. Um And that's one of the things I've been playing around with a lot. I'll pause there. It looks like you have something to say, but I have some I have some ideas and I mean, it really I you know, this is not my world. I'm uh not a great user of chat PC or user user of chat GPT and um uh but I I think it is true in helpful ways and unhelpful ones that we do feel threatened that what is core to us and unique to us is something that machines can do better. You know, on the helpful side, if you have people with dyslexia and attention problems, and they can get lots of information through uh TikTok, uh they can do lots of retrieval through chat GPT or Alexa or whatever they, you know, intermediary is. Um their lives are really changed because they have more of that what it is to be uh human than they had before. And it doesn't feel crutch-like, you know, where is the antidepressant crutch? It feels like you're just entering into modernity. So, that is a very helpful sort of connotation of it. Right. Right. That makes a lot of sense. One of the things that I've noticed, and I'm sort of curious for your reaction, is I've noticed that like it's very good at summarizing. So, it's very good at taking the best of what humanity or maybe I know, um and then summarizing it into like a little form that is like just right for the context that I need it in. Um and that's like a thing that I used to bundle under my conception of intelligence um as a pretty uniquely human thing. Like I don't I don't know of animals that do that. Um and is actually super valuable. Like you're a writer, I'm a writer. Uh in writing summarizing is a lot of what you're doing in writing. Um you have to be able to like summarize a complex idea in a way that's that's right for the audience or right for right for the context. And as I've started to notice this property of ChatGPT two two things happen that I think are quite related to your book. One is I started to see summarizing everywhere. Um I started to see it in writing. I started to see it in emails. Emails are just summaries of meetings I've had for example. Um it's just like it's all over the place. And then two is I've sort of noticed that to a large degree I've like carved that out of my sense of self and it's become a thing that I can hand off. And it's not necessarily like about intelligence anymore. It's about something else. It's just summarizing. It's just a task, right? And the intelligence is I have to learn to deploy or allocate the summarizing resource. And that's that's the sum summarizing resource. And that's actually the thing that is Summarizing What is an example of a summary that you find in insightful, you know, or that you would have had to work harder to get on your own? Great great question. So like here's an example. Like um like I don't know, about a year ago I was uh I was writing an article about the FTX scandal and I'm not sure if you're like totally familiar with it, but it's just like like the big the big crypto fund or like like exchange that like went bust and the people at the helm of that like Sam Bankman-Fried and and and all the other people were like very committed effective altruists, which is like a a branch of utilitarianism. And I wanted to write an article about like does being a utilitarian like lead you to commit like massive financial fraud necessarily? Um And in order to write that article, uh I needed to summarize utilitarianism. And I studied philosophy in college, so like I had basically know it. I've read a lot of Peter Singer, blah blah blah, whatever. But I haven't written about it in a long time. And so normally what I would have to do is I'd have to go like reread a bunch of Wikipedia, reread some of my notes, blah blah and then uh and then write the three sentences that's like utilitarianism is this. Um And what I did instead is I just I just put it into I was like, "ChatGPT, can you like explain utilitarianism for this context?" And it gave me the three sentences that I needed. And I I modified them, um but it saved me all that research time cuz I generally know the answer. I just like can't like quite put it in the right way. Well, that seems like an ideal use where you know enough that you would know if a mistake was being made. You know what you want to say pretty much and this sort of gets it to you fast. Yeah. Yeah, I mean I'm so much an on the one hand and on the other hand guy that I do find this a little uh not useful. I mean, but you know, I may not be sophisticated enough at framing. So, yeah. Yeah, I mean I I think it's important to explore like sort of sort of the ambivalence. I think my feeling about about what you just brought up which is like you should use it in areas where you know a lot is it's very analogous to um being a manager um using AI correctly. Where managers are have to know which resource to use or which human to use for which thing. They have to know how to get the best out of the humans that they use and they also have to know when to get into the details and when not to get into the details. Like when to delegate and when to like be suspicious and be like, "I I know if this is quite right." And I think a lot of those skills are sort of similar to the skills you need to use AI effectively and are going to have to be a lot more broadly distributed than they are today because you're right, like if you just let it go off and do a bunch of stuff without checking it and you don't actually know the underlying stuff, you're going to get stuff that's wrong. Um So what I guess what I'd like to do is we talked about we always do an explore section of this show. And so what I'd like to do is explore ChatGPT together cuz I think um I think you haven't used it a ton um and I think it'll be really fun to use it together and see what you think, see how it comports to your expectations or maybe maybe maybe doesn't. Um So I will say in advance, you know, of this I thought I cuz I'd written this novel about a doing psychotherapy with a narcissistic autocrat. I asked ChatGPT different ways to give me a novel plot or plot outline for that kind of book and it was you know, so routinized and obvious and no one would want that book uh and I thought this is really but you know, yeah, I think let's have fun with it. Okay, great. Let's have fun with it. Uh I'm going to share my screen so you can see it. Um and then and I will drive and you had a you had a specific question in mind. You had a specific thing that you wanted to uh see what how ChatGPT could um could respond. So give me the scenario, give me the thing that you have in mind. My third book is Should You Leave and You is the main character. So you want a piece of advice or you want some piece of advice perhaps and there all these yous with different dilemmas of intimacy and I thought you know, psychiatrists are reluctant to give advice. There's situations where they give it. So maybe we would go to chat GPT for advice and and I thought my fantasy was down the road artificial intelligence would have read so many diaries and so many uh memoir-like accounts in the popular press that it would know about the consequences of human behavior. So, you could go to chat GPT and say you know, I'm I'm going to make this woman could be a guy, but I'm a 29-year-old woman who is very sensitive to rejection. I don't know what to do. I'm with my boyfriend. We've been together for 6 months. The relationship is starting to feel a little one-sided like I'm giving too much. I don't know uh uh whether that's the case. Uh should I stick with him? You know, what will happen? You know, what what is the experience of women in my position? Do they better do better sticking it out or they do better leaving? You know, you've seen a lot of people go through these diaries and stuff into their mid-30s and I'm someone who wants to be married and have children. Uh but only in a warm, loving relationship. Uh should I stay or should I go? Or how much maybe more specific. How many more weeks should I give this before I revisit this question? Um Um well, I actually liked Well, okay. So, I just want to make sure that this description fits. So, like um I'm a 29-year-old 29-year-old woman who is very sensitive to rejection. I don't know what to do. I'm with my I've been with my boyfriend and we've been together for 6 months. The relationship is starting to feel a little one-sided. I don't know if that's actually the case. What happens in general to women in my position? Is that the question you want to ask? Do you want to ask? Okay. Um I mean, we would hope it time that uh uh some you you know, mechanical you know, some bot or something would actually be able to give us much more detailed uh response to this. Maybe there'd even be a kind of interview where the uh I think we can do an interview. Okay. Um so So, if you want So, basically, this will do whatever you ask it to do. And so, if you want it to do an interview to find out more information before gives an answer, it'll probably make the answer better. Um but you have to ask it for that cuz if you ask it to just answer, it will answer. Um Okay, well, let's you know, let's do it both ways or you take charge. I'm certainly not both ways. So, I want to just start with seeing if it will give us an answer and seeing if you agree with the answer. And I'm going to do something interesting. We'll see if it works, which is pretend you're the psychiatrist Peter Kramer, author of the book Should I Stay or Should I Go? Um Should you leave? Should you leave? Should Should you Should Should you stay or should you leave? \{question mark\} is the name of the book. Oh, it's just Okay, interesting. Um Should you leave? Um I'm a 29-year-old woman who's very sensitive to rejection. I'd love some help thinking this through. Um what Let's just say that. Okay, let's see if it Let's see if it works. In my book, Should You Leave?, I explore the complexity of relationships and the struggles individuals face when deciding whether to stay Your sensitivity to rejection is a common issue that many people grapple with and it can significantly impact your relationship dynamics. In general, women your in your experience might experience a range of emotions including anxiety, self-doubt, abandonment. These feelings can lead to heightened sensitivity to your partner's actions and words, causing you to perceive imbalance or lack of reciprocation more acutely. It's important to remember that these perceptions may not always reflect the reality of the situation. Self-reflection. Take some time to reflect on your feelings and the specific aspects of the relationship that make you feel it's one-sided. Are there specific behaviors or patterns that trigger your feelings of rejection? Communication. Open and honest communication with your boyfriend is crucial. Share your feelings and concerns with him in a non-confrontational way. This can help you clarify. Set boundaries. Establish a healthy Establishing healthy boundaries is important in any relationship. Make sure you're clear about your needs and expectations. Seek support. Consider seeking support from a therapist or counselor. Evaluate the relationship. After taking these steps, evaluate the relationship. Is there a genuine effort from both sides to address the issue? Are your needs being met? If not, it might be worth considering whether this relationship is right for you. Remember, every relationship has its challenges and it's normal to have doubts. The key is to address these feelings constructively and make decisions that are best for your emotional well-being. What do you think? What do I think? So, you know, I think I'm not uh an important enough figure well enough represented on the web for Chat GPT to cat capture anything specifically about me. I would say this is you know, pretty much the advice you would get from a glossy women's magazine. It's general. But but it is but it is correct. I mean, it's captured sort of the cultural consensus, right? This is what uh we would expect uh some average you know, poorly trained and vocal counselor to give, right? It It wouldn't need much, but they we we wouldn't You're not going wrong in any way. That makes sense. Um and but I would like something more predictive. So, I want to sort of ask AI what would you need to know to give me fairly specific advice about how long I should let this relationship go on or whether I should leave right now. Okay. Um I want you to give me give me fairly specific advice about about how long I should let this relationship go go on in the manner of Peter Carmichael. Let's just try to try to push it cuz I think it probably has red your stuff. What would you need to know in order to give me that advice? It might it may refuse. Depending on its on sort of the the training. And it looks like okay, blah blah blah. So it's asking us to you know, give more information about our communication efforts, our changes, personal boundaries, all the things it said before. So I think we're going to have to tell it like I've done this, this, and this. So what would you how do you want to construct the scenario? Huh. Um And do you think these are the right questions? My my Well, let's say my boyfriend is not abusive in any way, but when we go out to restaurants, I notice him uh spending a lot of time staring at other attractive women. Okay. Anything else? Um Uh Our We don't seem to have a lot of common interests. He is uh interested in stereotypically male uh pursuits like sports and engineering. I'm uh very interested in art and poetry. Okay. Let's see what it does with that. I want to I want to just make one modification here. Please. Chat GPT likes to ask multiple questions at a time and I just wanted to ask one. Okay, ask one question at a time. If you are not yet advanced. Okay. How does his his behavior of staring at other women when you're out together make you feel and have you discussed this with him? Yeah. Uh we talk about this problem all the time and he says, "Men are just like that. It means uh it has no significance for our relationship." Hm. Cool. And how does it make you feel? Uh I feel humiliated. It's important that your feelings are acknowledged and respected in the relationship. If you've communicated your feelings and he dismisses them by saying, "Men are just like that." it's just a lack of empathy and understanding on his part. Considering this, you might think about whether this relationship is meeting your emotional needs and whether your partner is willing to make changes to address your concerns. If you continue to feel humiliated and your concerns are not taken seriously, it may be worth reevaluating the relationship's viability. Yeah. I you know, I feel like Chat GPT is afraid that he or she is going to get sued. This is This is very namby-pamby. All right. Um okay. But I you know, I agree. It you know, it is remarkable to have something that captures you know, standard advice in this way, you know, and uh and I don't disagree by the way. Did we did you we hear anything we say, "Oh my god, that's way off base." No. Um no, I mean that that makes total sense. It is definitely being a little bit um You may be a little boring this poor woman to death. Um wait, I'm going to ask it not to be namby-pamby. Please be direct. How would you How would you characterize like the kind of response that you'd like to get? Um you know, how many more weeks should I give this before I leave? And I'm curious what it does. Given the situation you've described, if there are no significant improvements in his behavior and your communication about this issue within the next 4 to 6 weeks, it might be time to seriously consider leaving the relationship. Uh-huh, 4 to 6 weeks. Okay. What would you have said? How did it come up with 4 to 6 weeks? I'm I'm not unhappy with it, but uh it's kind of a short leash, okay. Let's see. Let's see. How did you come up with 4 to 6 weeks? The 4 to 6 weeks time frame is is a suggestion based on giving enough time to see if there's any meaningful change in your boyfriend's behavior and attitude towards your concerns. It's not a strict rule, but rather a guideline to allow for some time to assess if there's a genuine effort to address the issues you raised. This period also provides an opportunity for further reflections on your feelings and the overall health of the relationships. If you feel this time frame is too short or too long, you can adjust it based on your personal circumstance and comfort level. That's great. So, what one of the things that we're seeing here is um it is generic if it is given generic instructions, but the more you can ask it to simulate or be like a certain thing, it will get much better at being that thing. Bringing to bear all of the maybe generic advice, but you can say like I don't want that advice. I want another specific subset of advice. So, the more like specific terms you have, the less it sounds like Wikipedia, which is I think really interesting. And um You know, I I do think a therapist might say, "Let's give it 4 to 6 weeks and we'll revisit this." But, I don't think anybody thinks that people change over the course of 4 to 6 weeks. You know, they could give like a signal that they're thinking about thinking about changing, but uh that might be just wrong. That might be That might be tough. Yeah. So, I don't know, but I I made me laugh 4 to 6 weeks. I don't I don't know what I expected if it said you know, give it a year, would that be better or worse? I don't know. Yeah. Yeah, it's an interesting question. I think, um I see what Let me say what I want from it, you know, eventually is to say, I I've looked at thousands of biographies and I it you know, it never letting this go on never works out or letting this go on works 23% of the time. You know, how are you How are you with a 23% uh call rate? Is that enough for you to stay or not? That's up to you. But, we can give you really specific information based on an enormous range of experience beyond what any you know, therapist or or advisor could ever uh master. So, it's definitely not going to do that because it's it's not calculating those statistics. It's just predicting like the next word in a sequence. So, it's very good at given No, I would like I would like it to have the information rather than There's a There's a name for this which I'm uh forgetting, but it's like a chip or something. There's a There's a name for what what a piece of the what a a piece of information that gives you a likelihood of what the next word should be. You know, Google is still better at it than uh ChatGPT It's a logit. A what? A logit. Yeah, there's some other word that I'm not grasping, but it Yes, so Yeah, it knows what the next word should be, but doesn't know what the information is what the Does it have data? I'd like for it to have usable data, you know, actually one of the things that I've done, which I think you might be interested in is is is sort of using that. Like it's very good at capturing a a gestalt of a piece of text and then knowing what comes next is you can actually use that and hook into all of the things that it knows about you and about all the other people that have written similar things to see where it would go. So what I've done, if I'm making a decision and I've been like waffling about it, is I'll take a bunch of journal entries and I'll feed it in and then I'll say can you write the next journal entry in the sequence assuming I make the decision A and then write the next journal entry Ask it what what advice would Albert Ellis give me. Albert Ellis was very much a tough love, straighten up and fly right. Okay. So Um we could also start over and see if we start over with some of this information if it does it or we can continue in this chat. What would Okay. Either is fine. I believe me, you know, have more Great. So we'll try one more time. So pretend you're the psychiatrist. No, we're not doing that. Um pretend you're Al- Alfred Ellis or Albert? Albert. It's psychologist Albert Ellis. The psychologist. Um Uh and um please give me advice on this situation as him. As he would. Let's see. So now it's going to give us like very like Albert Ellis REBT stuff um instead of doing the personality. So let's see if we can we can make it his personality. Um please give me the as he would um adopt his personality and style. Don't give me generic advice. Respond directly to my to my question as he would. See if See if we we get it. So it's saying, "As Albert Ellis, I would advise you to first examine your beliefs about rejection and relationships. It's common for people to hold irrational beliefs that lead to emotional distress. In your case, your sensitivity to rejection might be rooted in belief that you must be universally liked and approved." So I think it's like I think it's capturing something about rational emotive therapy, which is what what Ellis did. It's capturing something of that. But it's not it's But it doesn't have his personality, but it it's it's too long ago. It's too long ago that he's not I mean, I'd have to think of some contemporary psychologist who's way out on the web a lot, and I'm not I don't know who that is, but It's also generally, I think like it doesn't it's not good at capturing that like that thing that like zing in a Irrascibility, yeah. Yeah, it's not good at that. Um, but it it is good at sort of like getting the right information, basically, or the general feeling. Yeah. This is really impressive compared to anything from a few years ago, but it's not but I wouldn't use it, you know. I wouldn't employ it. I I that that makes total sense. I'm impressed without, you know, changing my way of life here. That's totally fine. That is totally fine. I was not necessarily gunning for that, although it would have been nice if it happened. Um Uh Ask it whether antidepressants work in the treatment of depression. It is going to say yes. Um, it's going to say like it's going to give the Wikipedia answer. Okay, yeah. Well, I don't know what that answer is, but All right. Do antidepressants work in the treatment of depression? Oh, it's answering as um as Albert El- Ellis. As Albert Ellis, I would acknowledge that antidepressants can be effective in the treatment of depression for some individuals. However, I would also emphasize the importance of addressing the underlying cognitive and emotional factors that contribute to to depression. All right. Okay. There you go. Um okay, so that's our that's our chat GPT demo. I appreciate you humoring me on that. Um I've had fun. Okay. Okay. Um Yeah, so like I said, the the book that you wrote has had a like a tremendous impact on me. Um and I'm really really glad to get to uh get to share it with other people and and get to talk to you. It's honestly like a it's a huge honor. Um Anything uh any any final reflections before we close out the show? enjoyed this. I think it was a terrific discussion of the you know, really what's at the core of the book, which is technology and the self. So, yeah, it's been great. Cool. Thank you so much for coming on. Thank you. Oh my gosh, folks, you absolutely positively have to smash that like button and subscribe to How Do You Use Chat GPT? 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. Every episode is a roller coaster of emotions, insights, and laughter that will leave you on the edge of your seat 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. So, do yourself a favor, hit like, smash subscribe, and strap in for the ride of your life. And now, without any further ado, let me just say, Dan, I'm absolutely hopelessly in love with you.