March 20, 2024
The transcript of How Do You Use ChatGPT? with Peter Kramer is below for paying subscribers.
Timestamps
- Introduction: 00:50
- How technology changes the way we see ourselves and the world: 08:24
- Antidepressants and their impact on our personality and sense of self: 21:25
- How the availability of a technological solution prompts us to see the problem everywhere: 26:35
- Technology alters the categories we have divided the world into: 34:06
- How I use ChatGPT in my writing process: 40:05
- Experimenting with ChatGPT to get relationship advice: 45:41
- Prompting ChatGPT to be more specific: 51:16
- Clearly indicate the tone you want ChatGPT to take: 55:11
- Dr. Peter D. Kramer’s final thoughts on ChatGPT as a therapist: 1:02:27
Transcript
Peter Kramer (00:00:00)
So, first, antidepressants work. They’re not glorified placebos. Yes, I think we are in a sort of low ebb for the cultural view of antidepressants.
Dan Shipper (00:00:12)
There are probably some people who are like, has Dan gone crazy? What does this have to do a ChatGPT? It’s about how technology changed how we see ourselves and the world.
Peter Kramer (00:00:19)
I feel 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.
Dan Shipper (00:00:47)
Peter, welcome to the show.
Peter Kramer (00:00:48)
Thank you. Hi.
Dan Shipper (00:00:50)
I am so excited to have you because you're one of my favorite authors. You wrote a book that literally changed my life. I have it right here. If you can see, there's so many little dog ears on this book. I read it four times. It's incredible.
Peter Kramer (00:01:10)
I love that.
Dan Shipper (00:01:11)
So, for people who don't know 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 like Against Depression. You wrote a book about Freud, which I also read, which I really enjoyed. And you're also a novelist. Remind me again of the title of your most recent novel.
Peter Kramer (00:01:31)
Death of the Great Man.
Dan Shipper (00:01:32)
Death of the Great Man, which feels like you're talking a little bit about a psychiatrist's interaction with someone who's kind of like Donald Trump, but you haven't really said that.
Peter Kramer (00:01:40)
Yes, a psychiatrist is coerced into treating an authoritarian buffoonish autocratic national leader in his disastrous second term.
Dan Shipper (00:01:53)
Well, let's hope it doesn't come true. It's sort of Sopranos-esque, but the president instead of a mob boss maybe. So, I love that. I love how sort of multi-disciplinary you are. One of the things about psychiatry and my experiences with psychiatry, which we can get into in a second, is psychiatrists often are sort of mad brain scientists or whatever. And they just sort of give you magic chemicals, in my experience, that they’re like, we don't know how it works, but it might make you feel better, but it makes you feel worse then in the meantime or whatever. And therapists tend to be a lot more interested in humans and the human experience and stuff like that. And I think it's such a shame in the field because there's that divide, and you seem to just be in this really nice, both very humanistic, very interested in human experience and into the brain chemicals thing, that I really love.
Peter Kramer (00:03:00)
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 the early 1970s and decided then to become a doctor in order to become a psychiatrist, and had some other therapies good and bad, but 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 as there was 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 Should You Leave? So when Prozac appeared, I was really someone focused with my patients on psychotherapy and seeing their responses. I should run and hold it up, I suppose, but there's a 30th anniversary edition of Listening to Prozac that's out, just at the end of last year. And, In and I got to write a new preface and afterwords 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 yellow sheets of paper for easy identification in patients charts where I wrote down the medicines and doses that I prescribed, and I wasn't prescribing for most patients, but also their responses if there were side effects and things, and those 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.
Dan Shipper (00:05:05)
That makes sense and I think I dove in right off the deep end into 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 there are probably some people who are like, has Dan gone crazy? What does this have to do a ChatGPT? And the answer is, I'm not as crazy as I used to be. Thank you to some of the medications that you talk about in your book. But also I think there's something really important underlying this book. And I'll tell people a little bit about the book and a little bit about sort of the theme, but the theme of the book is about how technology, in this case, Prozac, or the larger class of drugs we call antidepressants, changes how we see ourselves and the world. And I think you set out to write what you call this resonant portrait where you're looking at how technology changes people, changes culture, changes doctors, changes how people see themselves. And there's a sort of important tradition of people doing that. So another person that you've brought up before is Tracy Kidder, who did the same thing with the PC revolution. Obviously antidepressants change how we see ourselves in a certain more obvious way where maybe it blunts negative emotions or negative affect, but I think one of your points is that technology in general can do that, and I see that with ChatGPT. I've been sort of writing about that and thinking about that, inspired in large part by your work, that I think also has impacts on how we see ourselves. And so what I want to do in this show is talk about your book, because I think it's just really important, talk about antidepressants because I think that people don't talk about them enough in the right way in public. I think your book does it really well. And I'm happy to share sort of my experiences with them. And then I want to take sort of 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, because I think that there are some really interesting parallels. And then we're just going to explore ChatGPT together. And we have a fun idea for how to do that. So, that's the basic outline of the show and why you're on the show. And I'll pause there. It sounds like you might have something to say.
Peter Kramer (00:07:35)
No, I really appreciate that summary because this book was exactly about 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 edited fiction. 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, it seemed to me, had that quality of focusing on a little change in technology and how it gives 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 wasn't all high technology, just I think that there are lots of technologies that really change the environment. And I thought, you know, I had a hold of one, and my patients said that. So 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, 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 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 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 a developmental theme. And, I thought, oh, we should think about how that might be happening.
Dan Shipper (00:09:48)
Yeah. I think that's one of the most striking things about your book is 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 help with depression, 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, all of that stuff they tend to help with. And I think one of the points that I took away from your book is, yeah, to some degree, they do operate directly on something like depression. Maybe they do curb negative affect or if you're feeling really depressed, you're feeling really anxious, those things will be blunted a little bit. But also, and this is something that almost nobody knows, despite how popular your book was, they work directly on personality. And so what they may do is they may tweak some of these things about people's personality. You'll be a little bit bolder. You'll be a little bit less sensitive to rejection. You may obsess a little bit less. And one of the interesting things about your book is that it doesn't deal with people who are super, super, super sick, who start to be able to function normally. You often focus on people who definitely have problems, but they're fairly mild. And when they take Prozac or Lexapro or Zoloft or any one of these antidepressants, the issue resolves but they also become better versions of themselves. And you have this really striking anecdote at the beginning of the book, which you already referred to, which is you might put a patient on Prozac for a while to resolve a depressive episode. And they'll become bolder, they'll become less sensitive to rejection, all that kind of stuff. And that will have a lot of marked changes in their lives. And when you taper them off, they come back to you and say, well, I'm not myself anymore—even though they're not depressed. And so the question that you pose is, well, what is a self to begin with? You weren't yourself for the first 30 years of your life before I put you on this drug. But the drug just brought out this other thing from you that was maybe always latent or always there, but it needed to turn those dials to let it come out.
Peter Kramer (00:12:12)
Right. I think it brings up the question of the natural. So if you 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—pain was constantly intruding. And then, you took a medicine that finally prevented migraine—one was invented. We would say, yes, you were 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. I write about this in a book called Against Depression. There's 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 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. And even in these very minor forms, sort of subforms, maybe that really isn't who you are. So I 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, 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 less easy to push around and I'm more certain about what I'm looking for.
Dan Shipper (00:13:59)
Totally. Yeah. And I will say for myself, I have experienced that. And yeah, it's been a crazy change for me. I think you know, one of the things I should point out what you sort of alluded to is that one of the things I love about the book is it sort of came out at this high watermark of antidepressant optimism where everyone was like, well, maybe it'll cure everything. And then I feel like right now we're in this low of like, it's been probably overprescribed to people who haven't been properly educated or aren't in psychotherapy or whatever. And so I think there's a lot of people out there who are like, oh, maybe it doesn't work, maybe it's a placebo. And it's definitely not a placebo. Having tried all the placebos. Let me tell you, it's not a placebo. But I think one of the things I love about the book is that you're kind of ambivalent about it in the classical sense of ambivalence where you're presenting the case, but you're like, there's some good things about this. It really does amazing things for people, but it's also not all good. And we shouldn't be too glib about jumping into changing all aspects of our personality in particular, because, like you said melancholy, that's sort of a cultural thing whether or not melancholy is acceptable. And I'll say for myself, having OCD, I was watching and my history is, I'm Jewish and my family's from Eastern Europe and they were definitely very Orthodox Jews and I was watching a YouTube video the other day of how an Orthodox Jew conducts their day and it's pretty obsessive, you're saying blessings about everything. You're touching the mezuzah, you're rubbing the tzitzit or whatever. And or I love books and I have a very strict sense of morality, and that's one of my sort of obsessive things. It's this sense of morals. And, 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 not particularly adaptive to me. And so I think one of the points of your book is that we can turn that down, so I'm more fit to the culture, but then you always have to ask me, is that right? Should I need to do that? And that's an open question.
Peter Kramer (00:16:26)
Yes. 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 Borges imagines an author, a made-up author, Pierre Menard, who writes one-and-a-half chapters of the novel, Don Quixote, in the 1910s. Alright, so we're in World War One and war is horrible. And in the Quixote, in the chapters Cervantes is glorifying war. And Borges says, how much more brilliant Menard's version is then Quixote's version, because the same words are written in a very different cultural surround. And is it ironic? And how is he making us think about war? I mean, it's a very clever metaphoric trope. And I thought, what, because I was asked by Penguin to do a 30th anniversary edition, actually an afterword, what would it be like to write exactly these words that were in Listening to Prozac in 2023 when we're so negative about antidepressants?
And I don't know that that would be brilliant or not brilliant, but it, you, it would definitely be going against the grain, whereas I think in the late 1980s, early 1990s, when 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 here are these off-patent drugs pennies a pill, that really treat depression quite well, so that when you go out to do a study, you are really recruiting very marginalized people people living in 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 involved where they have lots of contact with other people. They get a ride from their SRO to the welfare-to-work program. They get medical care and yeah, they'll do well on a dummy pill because they're being given so much support and 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 or have sort of exaggerated things to get in the study or the interviewer has exaggerated things for them. Or B, maybe there's 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 it becomes hard to test the medicines when the medicines are so good.
Dan Shipper (00:19:45)
Yeah, that makes sense.
Peter Kramer (00:19:50)
So I missed about four points there, but those were a couple of them.
Dan Shipper (00:19:58)
That's totally fine. I think what I want to get into a little bit is some of the broader lessons about how antidepressants shape how we see ourselves and also shape how we see the world. Because I think that's the thing that’s gonna be most relevant in our discussion of ChatGPT in a second. And the first one that I take away is, it 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. There’s explanations at the biological level, there's explanations at the psychological level, like an example or an example, I don't know. Let me think about—so there's explanations at the psychological level. A lot of people think that depression is repressed rage or something like that. Then there's explanations at the social level. It's like, yeah, you don't have a job, of course you're going to be depressed. Or your wife left you, of course you're going to be depressed. And I think one of the things that you noticed 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 this one thing when it it is true that Prozac is sort of operating on that level. And it does change all of these other levels, but it's maybe not the best thing to just jump to. Well, that must mean that's what's going on. All the time, right?
Peter Kramer (00:21:30)
So I think one good example, you mentioned the 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 taught about now, but there are people who, when they're rejected, go into mini depressions that go right off the edge. So I think I write in Listening to Prozac about a woman who, she asked her boyfriend a sensitive question and he turned away to look at the football game, would just start doubting everything. And, these medicines turn out to be really pretty good at preventing that, at making people less immediately sensitive to perceived loss. But the question is, what's going on with that woman? You know, she's sensitive and in two senses. One is she's vulnerable in the way we just said, and the other, she's perceptive. So 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're right about that, that it isn't as good as you think. And in a way, the medicine obviates that second form of sensitivity and we want to be a little wary of that. 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 used to think psychotherapy was very good for neurosis—it used to be called neurosis—making people less neurotic. And it turns out that's right. Probably that is one of the things that these medicines are best against. There's a technical term neuroticism, which is a little different from neurosis, but in one well-known study, the medicine paroxetine, Paxil, did better against neuroticism than it did against depression. And neuroticism meant this quick sensitivity to loss. And fussing about things and long-term pessimism and so on. But that really changed very notably on medication. And of course in psychotherapy, that's very helpful, but you don't want to lean on it too much.
Dan Shipper (00:23:51)
Right. And I think, one of the points that you've made is, yeah, if it's operating on neuroticism, which if you're familiar with some of this psychology here, neuroticism is one of the big five personality traits. 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, because you may be doing lots of things behaviorally to avoid sensitivity, avoid the feeling of loss that you think are part of your personality, but really aren't—or it's hard to say whether they really are or not. But what we can say is that if neuroticism goes away, all those other things will go away too. And that's a really interesting thing.
Peter Kramer (00:24:41)
And the other thing in this study that Paxil worked on was extroversion. 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.
Dan Shipper (00:24:55)
Totally, totally. And I think we've sort of covered how it changes how you see yourself, but I want to talk also about how it changes how we might see the world. And so when I talked earlier about antidepressants being this lever and when you have that lever, you can move in the world in a particular way. You reduce your explanations to that layer. It's like, we can move people's depression or move people's state with Prozac therefore it must have been a biological issue and that's not necessarily true. One of the other ways in which using technology as a lever changes how we see things is you will see more opportunities to use that lever. So one of the examples that I love in your book is you talked about how, when you were in residency, nobody came in complaining of panic attacks and you didn't really have a lot of training about panic attacks because they weren't a super common issue. But what's interesting is that as soon as Xanax became available, which is just a very effective 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 panic attacks, panic attacks became way more common. We started to see them everywhere because they were suddenly treatable. And I think that I would love for you to talk about that ‘because I think it's just a general phenomenon that we can apply to ChatGPT too.
Peter Kramer (00:26:40)
No, I think we really do see things that we have solutions for more. I mean, the panic attack story is very interesting because Donald Klein, who was this great conceptual psychopharmacologist, was out to treat panic attacks. I mean, they were very anxious people who were so anxious they were hospitalized, although the bar for hospitalization was probably lower in those days. So we're going back to the 1960s today. And they gave the older antidepressants these were called tricyclics, and they started becoming available in the late 1950s, but really in the 1960s. And there'd be these patients who are very anxious and they're always going up to the nursing desk and saying, you have to help me, I'm feeling so terrible. And they put them on the antidepressant and the nurses said, these patients are doing better, they’re not coming up and pleading with us. And 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, nurses are not helpful anyway. They weren't very helpful anyway. And he, based on this, made this psychopharmacological dissection, which was that the 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, which we don't give a lot of mind to now, were just the basis for internal conflict that created people's personality. And, you know, Donald Klein was more on the biological happenstance side. He was like, yeah, 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 altogether. And that formulation was just coming into place. Before I wrote Listening to Prozac, that was part of my education, so I had this funny happenstance in my own life, which is I 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, I went to my residency director and 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. He would give me the other final semester, or whatever you call it, under the 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. and I was very interested in community psychiatry and this was a community psychiatry position nationally. And that position disappeared and I was put in the shop of Gerry Klerman, one of the great innovators also in pharmacology who had me kind of become aware of the whole research portfolio in government and included holding conferences where Donna Klein held forth and so on. So I came to learn about modern pharmacology with this interest in psychotherapy and community psychiatry, which put me in this really ideal position when Prozac came out and I decided to write this book 'cause really I had deep familiarity with an enormous range of contemporary research while being the kind of person who had no interest in that.
Dan Shipper (00:31:02)
Right. And the thing I wanna push you on is this claim that is in your book that I just sort of repeated and I want you to give people a little bit more color on it, which is that panic attacks only became commonly thought about when Xanax was popularized and Xanax could work.
Peter Kramer (00:31:18)
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 fluoxetine, fluvoxamine is a little different medicine, that seemed to do well for OCD in the early going, it had some side effects that people didn't like. 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 waive rating answers, you saw that in the same population you've been treating for years, there was much more OCD than you'd ever recognize. But the same is true for manic depressive illness. Also, I think the fact that lithium was helpful for bipolar disorder, it took a while, but eventually America was very Freudian. We divided the world into neurotic and psychotic, but without much differentiation between schizophrenia and bipolar disorder. England was a little more willing to divide things. Well, we took on the English diagnostic system because it made sense to think about using lithium in bipolar disorder. So there's all this what we call psychopharmacological dissection that when you have a medicine to treat something, you start defining it.
Dan Shipper (00:32:55)
What do you think that's about? Like, why?
Peter Kramer (00:32:59)
Well 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 that distinction between schizophrenia and manic depression goes back hundreds of years anyway, many dozens of years. And I mean, it goes back to the 19th century and we say, oh, those early diagnosticians were right. You know, when it turns out that the distinguishes they made are useful for the application of these technologies. So I think that that's sort of the theme we're on the technologies correctly or falsely seem to be illuminating about how human experience ought to be subdivided.
Dan Shipper (00:34:03)
Yeah, that was the thing that I really want to talk about 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 different. There are other things that we can get into, for example, what is intelligence is a question that is really important when it comes to ChatGPT and AI, and I think it will cause us to redraw those lines a bit. And I'll tell you a little bit more about why, but to go back one of the things that you've brought up already that you put into your book that I think is so interesting is the way that lithium as a drug, because it could treat one specific disorder, or it could treat a set of symptoms, it helped us sort of pull those out of a mess of things and be this a separate category. So as you noted, before lithium, it was very hard to differentiate between manic depression and schizophrenia, because in both cases, you may have sort of psychotic symptoms. And in general, manic depression, depression, and schizophrenia were all thought to be under this big neurosis umbrella and caused by the same kind of intrapsychic conflict-type stuff. And that was based on Freud and psychotherapy and this sort of long history of it, which had many positive effects, but also many, many negative effects like blaming mothers for schizophrenic mothering and like all that kind of stuff. and what happened with lithium is, by mistake, we figured out that lithium can cure or can really help manic manic depression. But it doesn't help with schizophrenia. So we suddenly had this new category of manic depression that we could say was 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.
Peter Kramer (00:36:10)
Right. So it's interesting when a technology is very powerful, and I think that is true of lithium. you know, they say there are 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 it the fact that you could give lithium to people who have this terrible recurrent disorder, where they were going from mania to deep depression and so on, that you could give it and really prevent the next episode or make it much more distant than it had been in the past, that was very dramatic. And it's very hard to deny. That there is something out there if you can prevent it in that fashion.
Dan Shipper (00:37:04)
Yeah, that makes a lot of sense. And I think that is the linchpin that I think can bring us into ChatGPT land because lithium created what you call a pharmacological dissection. It created this category of manic depression. And I think ChatGPT is going to sort of force us to do something similar with intelligence. So intelligence is this big catch-all 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. Because it's the uniquely human trait, that's sort of what Socrates said makes men different from our intellect from animals and therefore, the life of man that's best is the one that's lived according to the intellect, stuff like that is a very common thing. And intelligence is super core to our sense of self. And in general, where we live in a knowledge economy where what you know and how you can bring that to bear has a big relationship to your status in the world. And I think that's why AI is promising, but also very threatening to people because it gets right at this word intelligence that I think a lot of people feel is very core to what makes humans unique. 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 ideas.
Peter Kramer (00:38:23)
No, no, go ahead. I mean, it really—this is not my world. I'm not a great user of ChatGPT. But 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 TikTok. They can do lots of retrieval through ChatGPT or Alexa or whatever their intermediary is, their lives are really changed because they have more of that, what it is to be a human, than they had before. And it doesn't feel like a crutch-like—you worry about the antidepressant crutch. It feels like you're just entering into modernity. So that is a very helpful sort of connotation of it.
The transcript of How Do You Use ChatGPT? with Peter Kramer is below for paying subscribers.
Timestamps
- Introduction: 00:50
- How technology changes the way we see ourselves and the world: 08:24
- Antidepressants and their impact on our personality and sense of self: 21:25
- How the availability of a technological solution prompts us to see the problem everywhere: 26:35
- Technology alters the categories we have divided the world into: 34:06
- How I use ChatGPT in my writing process: 40:05
- Experimenting with ChatGPT to get relationship advice: 45:41
- Prompting ChatGPT to be more specific: 51:16
- Clearly indicate the tone you want ChatGPT to take: 55:11
- Dr. Peter D. Kramer’s final thoughts on ChatGPT as a therapist: 1:02:27
Transcript
Peter Kramer (00:00:00)
So, first, antidepressants work. They’re not glorified placebos. Yes, I think we are in a sort of low ebb for the cultural view of antidepressants.
Dan Shipper (00:00:12)
There are probably some people who are like, has Dan gone crazy? What does this have to do a ChatGPT? It’s about how technology changed how we see ourselves and the world.
Peter Kramer (00:00:19)
I feel 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.
Dan Shipper (00:00:47)
Peter, welcome to the show.
Peter Kramer (00:00:48)
Thank you. Hi.
Dan Shipper (00:00:50)
I am so excited to have you because you're one of my favorite authors. You wrote a book that literally changed my life. I have it right here. If you can see, there's so many little dog ears on this book. I read it four times. It's incredible.
Peter Kramer (00:01:10)
I love that.
Dan Shipper (00:01:11)
So, for people who don't know 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 like Against Depression. You wrote a book about Freud, which I also read, which I really enjoyed. And you're also a novelist. Remind me again of the title of your most recent novel.
Peter Kramer (00:01:31)
Death of the Great Man.
Dan Shipper (00:01:32)
Death of the Great Man, which feels like you're talking a little bit about a psychiatrist's interaction with someone who's kind of like Donald Trump, but you haven't really said that.
Peter Kramer (00:01:40)
Yes, a psychiatrist is coerced into treating an authoritarian buffoonish autocratic national leader in his disastrous second term.
Dan Shipper (00:01:53)
Well, let's hope it doesn't come true. It's sort of Sopranos-esque, but the president instead of a mob boss maybe. So, I love that. I love how sort of multi-disciplinary you are. One of the things about psychiatry and my experiences with psychiatry, which we can get into in a second, is psychiatrists often are sort of mad brain scientists or whatever. And they just sort of give you magic chemicals, in my experience, that they’re like, we don't know how it works, but it might make you feel better, but it makes you feel worse then in the meantime or whatever. And therapists tend to be a lot more interested in humans and the human experience and stuff like that. And I think it's such a shame in the field because there's that divide, and you seem to just be in this really nice, both very humanistic, very interested in human experience and into the brain chemicals thing, that I really love.
Peter Kramer (00:03:00)
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 the early 1970s and decided then to become a doctor in order to become a psychiatrist, and had some other therapies good and bad, but 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 as there was 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 Should You Leave? So when Prozac appeared, I was really someone focused with my patients on psychotherapy and seeing their responses. I should run and hold it up, I suppose, but there's a 30th anniversary edition of Listening to Prozac that's out, just at the end of last year. And, In and I got to write a new preface and afterwords 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 yellow sheets of paper for easy identification in patients charts where I wrote down the medicines and doses that I prescribed, and I wasn't prescribing for most patients, but also their responses if there were side effects and things, and those 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.
Dan Shipper (00:05:05)
That makes sense and I think I dove in right off the deep end into 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 there are probably some people who are like, has Dan gone crazy? What does this have to do a ChatGPT? And the answer is, I'm not as crazy as I used to be. Thank you to some of the medications that you talk about in your book. But also I think there's something really important underlying this book. And I'll tell people a little bit about the book and a little bit about sort of the theme, but the theme of the book is about how technology, in this case, Prozac, or the larger class of drugs we call antidepressants, changes how we see ourselves and the world. And I think you set out to write what you call this resonant portrait where you're looking at how technology changes people, changes culture, changes doctors, changes how people see themselves. And there's a sort of important tradition of people doing that. So another person that you've brought up before is Tracy Kidder, who did the same thing with the PC revolution. Obviously antidepressants change how we see ourselves in a certain more obvious way where maybe it blunts negative emotions or negative affect, but I think one of your points is that technology in general can do that, and I see that with ChatGPT. I've been sort of writing about that and thinking about that, inspired in large part by your work, that I think also has impacts on how we see ourselves. And so what I want to do in this show is talk about your book, because I think it's just really important, talk about antidepressants because I think that people don't talk about them enough in the right way in public. I think your book does it really well. And I'm happy to share sort of my experiences with them. And then I want to take sort of 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, because I think that there are some really interesting parallels. And then we're just going to explore ChatGPT together. And we have a fun idea for how to do that. So, that's the basic outline of the show and why you're on the show. And I'll pause there. It sounds like you might have something to say.
Peter Kramer (00:07:35)
No, I really appreciate that summary because this book was exactly about 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 edited fiction. 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, it seemed to me, had that quality of focusing on a little change in technology and how it gives 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 wasn't all high technology, just I think that there are lots of technologies that really change the environment. And I thought, you know, I had a hold of one, and my patients said that. So 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, 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 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 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 a developmental theme. And, I thought, oh, we should think about how that might be happening.
Dan Shipper (00:09:48)
Yeah. I think that's one of the most striking things about your book is 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 help with depression, 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, all of that stuff they tend to help with. And I think one of the points that I took away from your book is, yeah, to some degree, they do operate directly on something like depression. Maybe they do curb negative affect or if you're feeling really depressed, you're feeling really anxious, those things will be blunted a little bit. But also, and this is something that almost nobody knows, despite how popular your book was, they work directly on personality. And so what they may do is they may tweak some of these things about people's personality. You'll be a little bit bolder. You'll be a little bit less sensitive to rejection. You may obsess a little bit less. And one of the interesting things about your book is that it doesn't deal with people who are super, super, super sick, who start to be able to function normally. You often focus on people who definitely have problems, but they're fairly mild. And when they take Prozac or Lexapro or Zoloft or any one of these antidepressants, the issue resolves but they also become better versions of themselves. And you have this really striking anecdote at the beginning of the book, which you already referred to, which is you might put a patient on Prozac for a while to resolve a depressive episode. And they'll become bolder, they'll become less sensitive to rejection, all that kind of stuff. And that will have a lot of marked changes in their lives. And when you taper them off, they come back to you and say, well, I'm not myself anymore—even though they're not depressed. And so the question that you pose is, well, what is a self to begin with? You weren't yourself for the first 30 years of your life before I put you on this drug. But the drug just brought out this other thing from you that was maybe always latent or always there, but it needed to turn those dials to let it come out.
Peter Kramer (00:12:12)
Right. I think it brings up the question of the natural. So if you 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—pain was constantly intruding. And then, you took a medicine that finally prevented migraine—one was invented. We would say, yes, you were 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. I write about this in a book called Against Depression. There's 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 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. And even in these very minor forms, sort of subforms, maybe that really isn't who you are. So I 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, 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 less easy to push around and I'm more certain about what I'm looking for.
Dan Shipper (00:13:59)
Totally. Yeah. And I will say for myself, I have experienced that. And yeah, it's been a crazy change for me. I think you know, one of the things I should point out what you sort of alluded to is that one of the things I love about the book is it sort of came out at this high watermark of antidepressant optimism where everyone was like, well, maybe it'll cure everything. And then I feel like right now we're in this low of like, it's been probably overprescribed to people who haven't been properly educated or aren't in psychotherapy or whatever. And so I think there's a lot of people out there who are like, oh, maybe it doesn't work, maybe it's a placebo. And it's definitely not a placebo. Having tried all the placebos. Let me tell you, it's not a placebo. But I think one of the things I love about the book is that you're kind of ambivalent about it in the classical sense of ambivalence where you're presenting the case, but you're like, there's some good things about this. It really does amazing things for people, but it's also not all good. And we shouldn't be too glib about jumping into changing all aspects of our personality in particular, because, like you said melancholy, that's sort of a cultural thing whether or not melancholy is acceptable. And I'll say for myself, having OCD, I was watching and my history is, I'm Jewish and my family's from Eastern Europe and they were definitely very Orthodox Jews and I was watching a YouTube video the other day of how an Orthodox Jew conducts their day and it's pretty obsessive, you're saying blessings about everything. You're touching the mezuzah, you're rubbing the tzitzit or whatever. And or I love books and I have a very strict sense of morality, and that's one of my sort of obsessive things. It's this sense of morals. And, 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 not particularly adaptive to me. And so I think one of the points of your book is that we can turn that down, so I'm more fit to the culture, but then you always have to ask me, is that right? Should I need to do that? And that's an open question.
Peter Kramer (00:16:26)
Yes. 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 Borges imagines an author, a made-up author, Pierre Menard, who writes one-and-a-half chapters of the novel, Don Quixote, in the 1910s. Alright, so we're in World War One and war is horrible. And in the Quixote, in the chapters Cervantes is glorifying war. And Borges says, how much more brilliant Menard's version is then Quixote's version, because the same words are written in a very different cultural surround. And is it ironic? And how is he making us think about war? I mean, it's a very clever metaphoric trope. And I thought, what, because I was asked by Penguin to do a 30th anniversary edition, actually an afterword, what would it be like to write exactly these words that were in Listening to Prozac in 2023 when we're so negative about antidepressants?
And I don't know that that would be brilliant or not brilliant, but it, you, it would definitely be going against the grain, whereas I think in the late 1980s, early 1990s, when 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 here are these off-patent drugs pennies a pill, that really treat depression quite well, so that when you go out to do a study, you are really recruiting very marginalized people people living in 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 involved where they have lots of contact with other people. They get a ride from their SRO to the welfare-to-work program. They get medical care and yeah, they'll do well on a dummy pill because they're being given so much support and 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 or have sort of exaggerated things to get in the study or the interviewer has exaggerated things for them. Or B, maybe there's 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 it becomes hard to test the medicines when the medicines are so good.
Dan Shipper (00:19:45)
Yeah, that makes sense.
Peter Kramer (00:19:50)
So I missed about four points there, but those were a couple of them.
Dan Shipper (00:19:58)
That's totally fine. I think what I want to get into a little bit is some of the broader lessons about how antidepressants shape how we see ourselves and also shape how we see the world. Because I think that's the thing that’s gonna be most relevant in our discussion of ChatGPT in a second. And the first one that I take away is, it 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. There’s explanations at the biological level, there's explanations at the psychological level, like an example or an example, I don't know. Let me think about—so there's explanations at the psychological level. A lot of people think that depression is repressed rage or something like that. Then there's explanations at the social level. It's like, yeah, you don't have a job, of course you're going to be depressed. Or your wife left you, of course you're going to be depressed. And I think one of the things that you noticed 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 this one thing when it it is true that Prozac is sort of operating on that level. And it does change all of these other levels, but it's maybe not the best thing to just jump to. Well, that must mean that's what's going on. All the time, right?
Peter Kramer (00:21:30)
So I think one good example, you mentioned the 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 taught about now, but there are people who, when they're rejected, go into mini depressions that go right off the edge. So I think I write in Listening to Prozac about a woman who, she asked her boyfriend a sensitive question and he turned away to look at the football game, would just start doubting everything. And, these medicines turn out to be really pretty good at preventing that, at making people less immediately sensitive to perceived loss. But the question is, what's going on with that woman? You know, she's sensitive and in two senses. One is she's vulnerable in the way we just said, and the other, she's perceptive. So 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're right about that, that it isn't as good as you think. And in a way, the medicine obviates that second form of sensitivity and we want to be a little wary of that. 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 used to think psychotherapy was very good for neurosis—it used to be called neurosis—making people less neurotic. And it turns out that's right. Probably that is one of the things that these medicines are best against. There's a technical term neuroticism, which is a little different from neurosis, but in one well-known study, the medicine paroxetine, Paxil, did better against neuroticism than it did against depression. And neuroticism meant this quick sensitivity to loss. And fussing about things and long-term pessimism and so on. But that really changed very notably on medication. And of course in psychotherapy, that's very helpful, but you don't want to lean on it too much.
Dan Shipper (00:23:51)
Right. And I think, one of the points that you've made is, yeah, if it's operating on neuroticism, which if you're familiar with some of this psychology here, neuroticism is one of the big five personality traits. 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, because you may be doing lots of things behaviorally to avoid sensitivity, avoid the feeling of loss that you think are part of your personality, but really aren't—or it's hard to say whether they really are or not. But what we can say is that if neuroticism goes away, all those other things will go away too. And that's a really interesting thing.
Peter Kramer (00:24:41)
And the other thing in this study that Paxil worked on was extroversion. 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.
Dan Shipper (00:24:55)
Totally, totally. And I think we've sort of covered how it changes how you see yourself, but I want to talk also about how it changes how we might see the world. And so when I talked earlier about antidepressants being this lever and when you have that lever, you can move in the world in a particular way. You reduce your explanations to that layer. It's like, we can move people's depression or move people's state with Prozac therefore it must have been a biological issue and that's not necessarily true. One of the other ways in which using technology as a lever changes how we see things is you will see more opportunities to use that lever. So one of the examples that I love in your book is you talked about how, when you were in residency, nobody came in complaining of panic attacks and you didn't really have a lot of training about panic attacks because they weren't a super common issue. But what's interesting is that as soon as Xanax became available, which is just a very effective 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 panic attacks, panic attacks became way more common. We started to see them everywhere because they were suddenly treatable. And I think that I would love for you to talk about that ‘because I think it's just a general phenomenon that we can apply to ChatGPT too.
Peter Kramer (00:26:40)
No, I think we really do see things that we have solutions for more. I mean, the panic attack story is very interesting because Donald Klein, who was this great conceptual psychopharmacologist, was out to treat panic attacks. I mean, they were very anxious people who were so anxious they were hospitalized, although the bar for hospitalization was probably lower in those days. So we're going back to the 1960s today. And they gave the older antidepressants these were called tricyclics, and they started becoming available in the late 1950s, but really in the 1960s. And there'd be these patients who are very anxious and they're always going up to the nursing desk and saying, you have to help me, I'm feeling so terrible. And they put them on the antidepressant and the nurses said, these patients are doing better, they’re not coming up and pleading with us. And 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, nurses are not helpful anyway. They weren't very helpful anyway. And he, based on this, made this psychopharmacological dissection, which was that the 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, which we don't give a lot of mind to now, were just the basis for internal conflict that created people's personality. And, you know, Donald Klein was more on the biological happenstance side. He was like, yeah, 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 altogether. And that formulation was just coming into place. Before I wrote Listening to Prozac, that was part of my education, so I had this funny happenstance in my own life, which is I 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, I went to my residency director and 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. He would give me the other final semester, or whatever you call it, under the 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. and I was very interested in community psychiatry and this was a community psychiatry position nationally. And that position disappeared and I was put in the shop of Gerry Klerman, one of the great innovators also in pharmacology who had me kind of become aware of the whole research portfolio in government and included holding conferences where Donna Klein held forth and so on. So I came to learn about modern pharmacology with this interest in psychotherapy and community psychiatry, which put me in this really ideal position when Prozac came out and I decided to write this book 'cause really I had deep familiarity with an enormous range of contemporary research while being the kind of person who had no interest in that.
Dan Shipper (00:31:02)
Right. And the thing I wanna push you on is this claim that is in your book that I just sort of repeated and I want you to give people a little bit more color on it, which is that panic attacks only became commonly thought about when Xanax was popularized and Xanax could work.
Peter Kramer (00:31:18)
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 fluoxetine, fluvoxamine is a little different medicine, that seemed to do well for OCD in the early going, it had some side effects that people didn't like. 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 waive rating answers, you saw that in the same population you've been treating for years, there was much more OCD than you'd ever recognize. But the same is true for manic depressive illness. Also, I think the fact that lithium was helpful for bipolar disorder, it took a while, but eventually America was very Freudian. We divided the world into neurotic and psychotic, but without much differentiation between schizophrenia and bipolar disorder. England was a little more willing to divide things. Well, we took on the English diagnostic system because it made sense to think about using lithium in bipolar disorder. So there's all this what we call psychopharmacological dissection that when you have a medicine to treat something, you start defining it.
Dan Shipper (00:32:55)
What do you think that's about? Like, why?
Peter Kramer (00:32:59)
Well 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 that distinction between schizophrenia and manic depression goes back hundreds of years anyway, many dozens of years. And I mean, it goes back to the 19th century and we say, oh, those early diagnosticians were right. You know, when it turns out that the distinguishes they made are useful for the application of these technologies. So I think that that's sort of the theme we're on the technologies correctly or falsely seem to be illuminating about how human experience ought to be subdivided.
Dan Shipper (00:34:03)
Yeah, that was the thing that I really want to talk about 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 different. There are other things that we can get into, for example, what is intelligence is a question that is really important when it comes to ChatGPT and AI, and I think it will cause us to redraw those lines a bit. And I'll tell you a little bit more about why, but to go back one of the things that you've brought up already that you put into your book that I think is so interesting is the way that lithium as a drug, because it could treat one specific disorder, or it could treat a set of symptoms, it helped us sort of pull those out of a mess of things and be this a separate category. So as you noted, before lithium, it was very hard to differentiate between manic depression and schizophrenia, because in both cases, you may have sort of psychotic symptoms. And in general, manic depression, depression, and schizophrenia were all thought to be under this big neurosis umbrella and caused by the same kind of intrapsychic conflict-type stuff. And that was based on Freud and psychotherapy and this sort of long history of it, which had many positive effects, but also many, many negative effects like blaming mothers for schizophrenic mothering and like all that kind of stuff. and what happened with lithium is, by mistake, we figured out that lithium can cure or can really help manic manic depression. But it doesn't help with schizophrenia. So we suddenly had this new category of manic depression that we could say was 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.
Peter Kramer (00:36:10)
Right. So it's interesting when a technology is very powerful, and I think that is true of lithium. you know, they say there are 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 it the fact that you could give lithium to people who have this terrible recurrent disorder, where they were going from mania to deep depression and so on, that you could give it and really prevent the next episode or make it much more distant than it had been in the past, that was very dramatic. And it's very hard to deny. That there is something out there if you can prevent it in that fashion.
Dan Shipper (00:37:04)
Yeah, that makes a lot of sense. And I think that is the linchpin that I think can bring us into ChatGPT land because lithium created what you call a pharmacological dissection. It created this category of manic depression. And I think ChatGPT is going to sort of force us to do something similar with intelligence. So intelligence is this big catch-all 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. Because it's the uniquely human trait, that's sort of what Socrates said makes men different from our intellect from animals and therefore, the life of man that's best is the one that's lived according to the intellect, stuff like that is a very common thing. And intelligence is super core to our sense of self. And in general, where we live in a knowledge economy where what you know and how you can bring that to bear has a big relationship to your status in the world. And I think that's why AI is promising, but also very threatening to people because it gets right at this word intelligence that I think a lot of people feel is very core to what makes humans unique. 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 ideas.
Peter Kramer (00:38:23)
No, no, go ahead. I mean, it really—this is not my world. I'm not a great user of ChatGPT. But 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 TikTok. They can do lots of retrieval through ChatGPT or Alexa or whatever their intermediary is, their lives are really changed because they have more of that, what it is to be a human, than they had before. And it doesn't feel like a crutch-like—you worry about the antidepressant crutch. It feels like you're just entering into modernity. So that is a very helpful sort of connotation of it.
Dan Shipper (00:40:02)
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 it's very good at summarizing, so it's very good at taking the best of what humanity or, maybe, I know, and then summarizing it into a little form that is just right for the context that I need it in. And that's a thing that I used to bundle under my conception of intelligence as a pretty uniquely human thing. I don't know of animals that do that. And is actually super valuable. You're a writer, I'm a writer. In writing, summarizing is a lot of what you're doing in writing. You have to be able to summarize a complex idea in a way that's right for the audience, right for the context. And as I've started to notice this property of ChatGPT, two things happen that I think are quite related to your book. One is I started to see summarizing everywhere. 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. It's all over the place. And then two is I've sort of noticed that to a large degree, I've carved that out of my sense of self. And it's become a thing that I can hand off and it's not necessarily about intelligence anymore. It's about something else. It's just summarizing. It's just a task. And the intelligence is I have to learn to deploy or allocate the summarizing resource. And that's actually that is—
Peter Kramer (00:41:40)
What is an example of a summary that you find insightful that you would have had to work harder to get on your own?
Dan Shipper (00:41:49)
Great question. So, here's an example: I don't know, about a year ago, I was writing an article about the FTX scandal and I'm not sure if you're totally familiar with it, but it's just the big crypto fund or exchange that went bust and the people at the helm of that, Sam Bankman-Fried and all the other people were very committed effective altruists, which is a branch of utilitarianism. And I wanted to write an article about does being utilitarian leads you to commit massive financial fraud, necessarily? And in order to write that article I needed to summarize utilitarianism. And I studied philosophy in college, so I 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 reread a bunch of Wikipedia, reread some of my notes, blah, blah, and then and then write the three sentences like utilitarianism is this. And what I did instead is I just put it into ChatGPT. I was like, can you explain utilitarianism in this context? And it gave me the three sentences that I needed and I modified them. But it saves me all that research time. ‘Cause I generally know the answer, I just can't quite put it in the right way.
Peter Kramer (00:43:08)
Well, that seems like an ideal use where you know enough that you would know if a mistake was being made, and you know what you want to say pretty much, and this sort of gets to you fast. Yeah, I mean, I'm so much an on the one hand, on the other hand guy that I do find this a little not useful, I mean, but, you know, I may not be sophisticated enough at framing. So, yeah.
Dan Shipper (00:43:33)
Yeah. I mean, I think it's important to explore sort of the ambivalence. I think my feeling about what you just brought up, which is, you should use it in areas where a lot is very analogous to being a manager—using AI correctly—where managers have to know which resource to use, 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, when to delegate and when to be suspicious and be—I don't 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. 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. So I guess what I'd like to do is we talked about, we always do an explore section of the show. And so what I'd like to do is explore ChatGPT together. ‘Cause I think you haven't used it a ton, and I think it'd be really fun to use it together and see what you think, see how it comports to your expectations, or maybe doesn’t.
Peter Kramer (00:44:52)
So I will say in advance of this, I thought because I'd written this novel about doing psychotherapy with a narcissistic autocrat, I asked ChatGPT different ways to give me a novel plot or a plot outline for that kind of book. And it was so routinized and obvious that no one would want that book and I thought this is really—but, yeah, I think let's have fun with it.
Dan Shipper (00:45:23)
Okay, great. Let's have fun with it. I'm gonna share my screen so you can see it and then I will drive. And you had a specific question in mind. You had a specific thing that you wanted to see how ChatGPT could be useful. So give me the scenario. Give me the thing that you have in mind.
Peter Kramer (00:45:42)
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 are all these Yous with different dilemmas of intimacy. And I thought, you know, psychiatrists are reluctant to give advice. There are situations where they give it, so maybe we would go to ChatGPT for advice and I thought my fantasy was down the road, artificial intelligence would have read so many diaries and so many memoir-like accounts in the popular press that it would know about the consequences of human behavior. So you could go to ChatGPT and say—I'm going to make this a woman, it 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 six months. The relationship is starting to feel a little one-sided, like I'm giving too much. I don't know whether that's the case. Should I stick with him? What will happen? What is the experience of women in my position? Do they do better sticking it out or do they do better leaving? You've seen a lot of people go through these diaries and stuff into their mid-thirties. And I'm someone who wants to be married and have children but only in a warm, loving relationship. 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?
Dan Shipper (00:47:27)
Well, I actually liked—well, okay. So I just want to make sure that this description fits. So, “I'm a 29-year-old woman who is very sensitive to rejection. I don't know what to do. I've been with my boyfriend and we've been together for six 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.
Peter Kramer (00:47:59)
I mean, we would hope in time that some mechanical bot or something would actually be able to give us a much more detailed response to this. Maybe that even be a kind of interview where the—
Dan Shipper (00:48:18)
I think we can do an interview. So, if you want, basically this will do whatever you ask it to do. And so if you want to do an interview to find out more information before it gives an answer, it will probably make the answer better. But you have to ask for it. ‘Cause if you ask it to just answer, it will answer.
Peter Kramer (00:48:35)
Well, let's do it both ways or you take charge. I'm certainly—
Dan Shipper (00:48:37)
Let’s do it 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 You Leave? I'm a 29-year-old woman who is very sensitive to rejection. I'd love some help thinking this through.” Let's just say that. 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 or go. Your sensitivity to rejection is a common issue that many people grapple with, and it can significantly impact your relationship dynamics. In general, women 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: 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 for 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?
Peter Kramer (00:50:34)
What do I think? So, I think I'm not an important enough figure or well-enough represented on the web for ChatGPT to capture anything specifically about me. I would say this is pretty much the advice you would get from a glossy women's magazine.
Dan Shipper (00:50:52)
It’s general.
Peter Kramer (00:50:54)
But it is correct. I mean, it captured a sort of cultural consensus, right? This is what we would expect some average, poorly trained counselors to give right. It wouldn't need much, but you're not going wrong in any way. That makes sense. But I would like something more predictive. So I want to sort of ask, 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?
Dan Shipper (00:51:39)
“I want you to give me fairly specific advice about how long I should let this relationship go on in the manner of Peter Kramer.” Let's just try to push it ‘cause I think it probably has read your stuff. “What would you need to know in order to give me that advice?” It may refuse, depending on its own sort of training. And it looks like, okay, blah, blah, blah. So, it's asking us to give more information about our communication efforts, our changes, personal boundaries, all the things that were said before. So I think we're gonna have to tell it, I've done this, this, and this. So what would you do? How do you want to construct this scenario? And do you think these are the right questions?
Peter Kramer (00:52:50)
“My boyfriend is not abusive in any way. But when we go out to restaurants, I notice him spending a lot of time staring at other attractive women.”
Dan Shipper (00:53:13)
Okay. Anything else?
Peter Kramer (00:53:22)
“We don't seem to have a lot of common interests. He is interested in stereotypically male pursuits like sports and engineering. I'm very interested in art and poetry.”
Dan Shipper (00:53:30)
Let's see what it does with that. I want to just make one modification here. ChatGPT likes to ask multiple questions at a time. And I just wanted to ask one, “How does his behavior of staring at other women when you're out together make you feel and have you discussed this with him?”
Peter Kramer (00:54:36)
“We talk about this problem all the time and he says men are just like that. It has no significance for our relationship.”
Dan Shipper (00:54:40)
And how does it make you feel?
Peter Kramer (00:54:42)
“I feel humiliated.”
Dan Shipper (00:54:48)
“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 re-evaluating the relationship's viability.”
Peter Kramer (00:55:10)
Yeah. I feel ChatGPT is afraid that he or she is going to get sued. This is very namby-pamby. Alright. But I agree it is remarkable to have something that captures standard advice in this way and I don't disagree, by the way, did we hear anything we’d say, oh, my god, that's way off-base? No.
Dan Shipper (00:55:44)
No, I mean, that makes total sense. It is definitely a little bit—
Peter Kramer (00:55:46)
You may be boring this poor woman to death.
Dan Shipper (00:55:54)
Wait, I'm going to ask it not to be namby-pamby. “Please be direct.” How would you characterize the kind of response that you'd like to get?
Peter Kramer (00:56:03)
“How many more weeks should I give this before I leave?”
Dan Shipper (00:56:14)
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 four-to-six weeks, it might be time to seriously consider leaving the relationship.
Peter Kramer (00:56:26)
Four-to-six weeks. Okay.
Dan Shipper (00:56:30)
What would you have said?
Peter Kramer (00:56:32)
How did it come up with four to six weeks? I'm not unhappy with it, but it's kind of a short leash. Okay.
Dan Shipper (00:56:34)
Let’s see. Let's see. “How did you come up with four-to-six weeks?” “The four-to-six weeks time frame 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.”
Peter Kramer (00:57:02)
That's great.
Dan Shipper (00:57:04)
So one of the things that we're seeing here is, 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, I don't want that advice. I want another specific subset of advice. So the more specific terms you have, the less it sounds like Wikipedia, which is, I think, really interesting and—
Peter Kramer (00:57:33)
You know, I do think a therapist might say, let's give it four-to-six weeks and we'll revisit this, but I don't think anybody thinks that people change over the course of four-to-six weeks. You know, they could give a signal that they're thinking about thinking about changing. But that might be just wrong.
Dan Shipper (00:57:52)
That might be tough.
Peter Kramer (00:57:54)
Yeah, so I don't know, but it made me laugh—four-to-six weeks. I don't know what I expected if it said give it a year, would that be better or worse? I don't know.
Dan Shipper (00:58:04)
Yeah. It's an interesting question. I think—
Peter Kramer (00:58:06)
I see, let me say, what I want from it eventually is to say, I've looked at thousands of biographies and letting this go on never works out, or letting this go on works 23 percent of the time. How are you with a 23 percent call rate? Is that enough for you to stay or not? That's up to you. We can give you really specific information based on an enormous range of experience beyond what any therapist or advisor could ever master.
Dan Shipper (00:58:48)
So it's definitely not going to do that because it's not calculating those statistics. It's just predicting the next word in a sequence. So it's very good at given—
Peter Kramer (00:58:58)
No, I would like it to have the information rather than—there's a name for this, which I'm forgetting, but it's like a chip or something. There's a name for a piece of information that gives you a likelihood of what the next word should be—Google is still better at it than ChatGPT.
Dan Shipper (00:59:21)
It's a logit.
Peter Kramer (00:59:22)
A logit? Yeah. There's some other word that I'm not grasping, but, yes. So yeah, it knows what the next word should be, but doesn't know what the information is. I'd like for it to have usable data.
Dan Shipper (00:59:16)
So, one of the things that I've done, which I think you might be interested in is sort of using that it's very good at capturing 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—
Peter Kramer (01:00:16)
Ask what advice Albert Ellis would give. Albert Ellis was very much a tough love, straighten up and fly right.
Dan Shipper (01:00:23)
Okay, so, 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.
Peter Kramer (01:00:32)
It's fine. Believe me.
Dan Shipper (01:00:36)
Great. So, we'll try one more time. So, “Pretend you're the psychologist Albert Ellis. Please give me advice on this situation as he would.” Let's see. So now it's going to give us very Albert Ellis-REBT stuff, instead of doing the personality. So let's see if we can make it his personality. “Please give me, as he would… adopt his personality and style. Don't give me generic advice. Respond directly to my questions as he would see if 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 rational beliefs that lead to emotional distress. In your case, your sensitivity or rejection might be rooted in the belief that you must be universally liked and approved.”
Peter Kramer (01:01:56)
So I think it's capturing something about rational emotive therapy, which is what Albert Ellis did. It's capturing something of that, but it doesn't have his personality, but it's too long ago. It's too long ago. He's not, I mean, I have to think of some contemporary psychologist who's way out on the web a lot. And I'm not, that is, but—
Dan Shipper (01:02:15)
It's also generally, I think it's not good at capturing that thing that like zing—
Peter Kramer (01:02:17)
That irascibility.
Dan Shipper (01:02:20)
Yeah, it's not good at that. But it is good at sort of getting the right information basically, or the general feeling.
Peter Kramer (01:02:28)
Yeah, this is really impressive compared to anything from a few years ago, but I wouldn't use it. I wouldn't employ it.
Dan Shipper (01:02:37)
That makes total sense.
Peter Kramer (01:02:39)
I'm impressed without changing my way of life here.
Dan Shipper (01:02:42)
That's totally fine. That is totally fine. I was not necessarily gunning for that, although it would have been nice if it happened.
Peter Kramer (01:02:48)
Ask it whether antidepressants work in the treatment of depression.
Dan Shipper (01:02:50)
And it's going to say yes. It's going to say like, it's going to give the Wikipedia answer.
Peter Kramer (01:02:52)
Okay. Yeah. Well, I don't know what that answer is, but—
Dan Shipper (01:03:00)
Alright, “Do antidepressants work in the treatment of depression?” Oh, it's answering as Albert Ellis.
Peter Kramer (01:03:07)
That's great.
Dan Shipper (01:03:09)
“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 depression.”
Peter Kramer (01:03:18)
Alright. Okay.
Dan Shipper (01:03:20)
There you go. Okay. So that's our ChatGPT demo. I appreciate you humoring me on that.
Peter Kramer (01:03:28)
I've had fun.
Dan Shipper (01:03:30)
Good, good. So, like I said, the book that you wrote has had a tremendous impact on me and I’m really, really glad to get to share it with other people and get to talk to you. It's honestly a huge honor. Any final reflections before we close out?
Peter Kramer (01:03:52)
No, I’ve really enjoyed this. I think it was a terrific discussion of the really what's at the core of the book, which is technology in itself, so yeah, it's been great.
Dan Shipper (01:04:04)
Cool. Thanks so much for coming on.
Peter Kramer (01:04:05)
Thank you.
Thanks to Scott Nover for editorial support.
Dan Shipper is the cofounder and CEO of Every, where he writes the Chain of Thought column and hosts the podcast How Do You Use ChatGPT? You can follow him on X at @danshipper and on LinkedIn, and Every on X at @every and on LinkedIn.
Subscribe to read the full article
Ideas and Apps to
Thrive in the AI Age
The essential toolkit for those shaping the future
"This might be the best value you
can get from an AI subscription."
- Jay S.
Every Content
AI&I Podcast
Cora
Sparkle
Spiral
Join 100,000+ leaders, builders, and innovators
Email address
Already have an account? Sign in
What is included in a subscription?
Daily insights from AI pioneers + early access to powerful AI tools
Front-row access to the future of AI
In-depth reviews of new models on release day
Playbooks and guides for putting AI to work
Prompts and use cases for builders
In-depth reviews of new models on release day
Playbooks and guides for putting AI to work
Prompts and use cases for builders
Subscribe to read the full article
Ideas and Apps to
Thrive in the AI Age
The essential toolkit for those shaping the future
"This might be the best value you
can get from an AI subscription."
- Jay S.
Every Content
AI&I Podcast
Cora
Sparkle
Spiral
Join 100,000+ leaders, builders, and innovators
Email address
Already have an account? Sign in
What is included in a subscription?
Daily insights from AI pioneers + early access to powerful AI tools
Front-row access to the future of AI
In-depth reviews of new models on release day
Playbooks and guides for putting AI to work
Prompts and use cases for builders
In-depth reviews of new models on release day
Playbooks and guides for putting AI to work
Prompts and use cases for builders
Dan Shipper
Reviews of new AI models on release day