The Seminar Continues

on personality psychology, arguments ensue

To get a sense of how the user can jump into the conversation, you can try a mockup. In the mockup, after each chunk of dialogue you can click “continue” or “call on me!” You may have to scroll down to see the buttons. If you click the latter button, you get to type in a comment or question. But don’t put thought into it: in the mockup, it will ignore what you type and show some pre-canned responses instead. It’s a mockup! Try it here

The idea here is for a seminar in which you immerse yourself along with an AI professor and four AI-generated students. Claude and I collaborated on a script below.

If this were fully developed (a) there would be a working “call on me!” button to allow you to interrupt at any time with a comment or question, and have the professor and the other students react¹; and (b) it would be multimedia, not just text. I think (a) is necessary to have a Minimum Viable Product.

I would appreciate it if below you would submit the sorts of comments and questions you might ask at this seminar, so that I can see how Claude handles them. Claude is amazing at starting from a set of essays on the topic that I have written and coming up with a script/screenplay.

Also, as I reviewed some of the scripts, I thought that everyone in the seminar was too nicey-nicey. So I instructed Claude to introduce some conflict. It may be a bit overdone now, but I hope it makes the AI characters seem more vivid.

Eventually, I imagine that one of the big corporations is going to put together a sort of studio in which an author/director like me can create not just a seminar but other shows that allow user participation with AI characters responding and the plot taking different directions. In the meantime, it still might be fun to see what I can do with Claude to try to develop this particular instance. Claude wrote the code for the mockup—it would have taken me weeks to do it the old-fashioned way.

Chapter: Personality Psychology

[Continuing from the correlation discussion where Blake dismissed personality psychology as "baloney"...]

Professor Hartwell: You asked about personality psychology during our correlation discussion, Blake. Is it at all useful?

Blake: I mean, it just seems like horoscopes with better marketing. People want to put themselves in boxes.

Drew: immediately bristling That's exactly the kind of dismissive attitude that—

Professor Hartwell: Hold on, Drew. Blake, let me ask you this: in order to thrive, humans have to be able to predict, explain, and change what we encounter in our environment. And what is more important in our environment than the behavior of other humans?

Drew: speaking up That's exactly why I think personality psychology matters. Understanding people isn't just academic—it affects how we treat each other, especially people who might be—

Blake: interrupting Ok, ok. And I noticed that you used the same three verbs—predict, explain, and change—to describe correlation. So do you think of humans as constantly working with correlations?

Professor Hartwell: Yes I do. When we do this formally, using certain methods, we call it science. But we do it informally all of the time. Informally, everyone is a personality psychologist. We try to come up with ways to predict, explain, and change the way other people around us behave.

Blake: But other people are trying to do the same thing to us!

Casey: dryly Welcome to the human condition, Blake.

Professor Hartwell: Indeed. And how you behave is likely to depend on your beliefs about personality psychology. So if I am going to predict, explain, and change your behavior, I may need to know how you think about personality psychology.

Blake: You're saying that my personality includes my beliefs about your personality. That gets complicated. Does formal personality psychology get into that?

Avery: jumping in It's obviously recursive—personality theories become part of personality formation, which affects how theories develop, which...

Blake: cutting off Right, but that just makes it seem even more made-up.

Professor Hartwell: Hmm. Maybe indirectly. But first let's talk about informal personality psychology. For example, what would you say it means if I describe someone as "manipulative"?

Drew: immediately I guess it means that the person gets me to do things that are in their interest but not necessarily in mine.

Casey: But that's a loaded term, isn't it? What's the difference between "persuasive" and "manipulative"? Seems like it depends on whether you agree with their goals.

Blake: sarcastically Great, so now we can't even agree on what words mean.

Professor Hartwell: So that would be a useful thing to know, right? Now, some people might be manipulative in certain situations—like a car salesman trying to make a commission. But others might have manipulative behavior as a consistent personality pattern, regardless of the situation.

Drew: So there's a difference between someone acting manipulative and someone being manipulative?

Avery: Obviously there's situational versus dispositional behavior. The question is whether dispositions actually exist or if everything's just situational responding to—

Blake: exasperated Can we stick to one thing at a time?

Professor Hartwell: Exactly, Drew. And that's where formal personality psychology comes in. It tries to distinguish between situational behavior and persistent personality patterns. Formal personality psychology doesn't define a "manipulative personality" per se. Instead, it looks at manipulative behavior as a characteristic of certain personality disorders, including antisocial personality disorder and narcissistic personality disorder.

Blake: What makes these disorders more scientific than just calling someone manipulative?

Casey: Well, for starters, they have operational definitions instead of just folk psychology labels—

Blake: interrupting "Folk psychology"? You mean how normal people actually think about other people?

Casey: getting slightly irritated I mean unscientific categorization based on cultural assumptions rather than systematic observation.

Professor Hartwell: Let's take antisocial personality disorder, or ASP. It's defined as someone who persistently disregards and violates the rights of others. Although it's diagnosed in someone at least 18 years old, the pattern usually traces back to childhood. So unlike situational manipulation, someone with ASP shows this pattern across different contexts and relationships.

But here's the catch—not everyone with ASP is manipulative.

Blake: Uh-oh: sounds like the First Iron Law. Sometimes it's this way, and sometimes it's that way.

Drew: frustrated Blake, you can't just dismiss everything because it's not perfectly predictive. Most useful knowledge involves probabilities, not certainties.

Blake: I'm not dismissing it, I'm just pointing out that if your scientific categories don't actually predict behavior...

Professor Hartwell: I'm afraid Blake has a point about the First Iron Law. There are several behaviors that can indicate ASP—criminality, impulsivity, aggressiveness, manipulativeness, and others. You can be diagnosed with ASP if you exhibit any three of these behaviors, with or without including manipulative behavior.

Drew: concerned So what can you actually predict about someone with ASP? Can you change their behavior?

Avery: The whole system is set up wrong. You're trying to categorize continuous variables into discrete boxes, when personality obviously exists on spectrums with complex interactions—

Casey: mildly exasperated Avery, not everything is a systems design problem.

Professor Hartwell: Well, one thing you can predict is that it will be really hard to change such a person's behavior. One important motivation for changing is wanting to be better liked by others, and ASP by definition means you lack that motivation. You can predict various adverse outcomes—higher probability of incarceration, lower probability of stable employment or marriage.

Drew: That sounds pretty extreme. How do personality psychologists deal with normal people who aren't disordered?

Blake: Yeah, most people aren't psychopaths. What about the rest of us?

Professor Hartwell: That's where we move to the other end of the spectrum. Have you heard of Myers-Briggs? It has some problems, which I'll get to, but it treats everyone as normal. It says personality differences explain why people clash, but neither type is inherently better or worse.

Casey: joining the conversation I think I've heard of it. Something about introverts and extraverts?

Avery: dismissively Myers-Briggs is completely unscientific. The categories are arbitrary, the test-retest reliability is terrible, and—

Drew: cutting in But millions of people find it useful for understanding themselves and others. Doesn't that count for something?

Avery: Horoscopes are popular too. That doesn't make them valid.

Professor Hartwell: That's one dimension, yes, Casey. Let me give you a few examples. If you're a J-type, for "judging," you tend to want cut-and-dried conclusions. If you're a P-type, for "perceiving," you prefer to explore nuances before deciding. In a business meeting, the J-type wants to stick to the agenda and check off decisions. The P-type wants to keep revisiting issues the J-type thinks were already settled.

Casey: That sounds like torture for both sides.

Blake: sarcastically Great, so now we have an excuse for why people are annoying.

Professor Hartwell: Exactly, Casey! But Myers-Briggs says both approaches have value. Here's another one: S-types need details to understand something, while N-types need the big picture. To an S-type, an N-type seems stupidly careless. To an N-type, an S-type seems stupidly slow. But typically you need both perspectives.

Drew: So you're saying people aren't trying to be difficult—they just process information differently?

Avery: impatiently But these are just made-up categories that don't correspond to actual cognitive differences. There's no neurological basis for—

Blake: interrupting Avery, do you always need everything to have a neurological basis before it can be useful?

Professor Hartwell: Right, Drew. There's also the thinking versus feeling distinction. A T-type wants the logical argument for, say, universal basic income. An F-type wants to hear how it will make people's lives better.

Casey: But surely good policy needs both logical arguments and consideration of human impact?

Drew: Exactly! That's why these differences could be valuable rather than just divisive.

Professor Hartwell: And then there's extraversion versus introversion. Myers-Briggs uses these terms differently than the original psychologist Carl Jung. For Myers-Briggs, it's about energy: extraverts recharge by being around people, introverts recharge by being alone.

Casey: I've read about Jung. His idea of extraversion was more about the person being outgoing, with the introvert being more reserved.

Avery: jumping in So what's wrong with Myers-Briggs if it seems so useful?

Blake: under his breath Finally, a question worth asking.

Professor Hartwell: The problem is reliability and validity. If you're going to assign someone a personality type, your process needs to meet certain standards. First, there isn't one standard Myers-Briggs survey—there are many versions, and they don't all give the same person the same classification.

Blake: perking up So I might get different results from different tests?

Drew: defensively But that could just mean the tests need improvement, not that the underlying concepts are wrong.

Avery: No, it's deeper than that. The whole typology approach is fundamentally flawed compared to dimensional approaches that—

Casey: sighing Can we hear the professor's explanation before you redesign the entire field?

Professor Hartwell: Exactly, Blake. And even with the same test, if you take it today versus next week, you might get different results. You might be in a mood today that leads to T-type responses, but next week you give F-type responses.

Also, your behavior might be situational. Your patience for details might be high when you're decorating your living room but low when you're planning a dinner party.

Blake: So Myers-Briggs might be measuring my temporary state, not my personality?

Drew: But couldn't that just mean we need better tests, not that personality doesn't exist?

Blake: getting annoyed Drew, you keep defending things that don't work by saying they could work if they were different.

Professor Hartwell: That's one problem, Blake. Another is that Myers-Briggs is either/or rather than quantitative. You find out you're an N-type, not an S-type, but are you strongly N or barely N? And how does being strongly N differ from being weakly N?

Casey: So what does formal personality psychology do differently to avoid these problems?

Avery: Finally! Use continuous variables instead of arbitrary categories, measure multiple dimensions simultaneously, establish proper psychometric properties—

Blake: interrupting Avery, we get it. You think everything should be more scientific. But does it actually work better?

Professor Hartwell: Good question, Casey. Formal personality psychology tries to use more reliable measurement tools and looks at personality as existing on continua rather than discrete categories.

River, do you have a question or comment, or would you like me to continue?

[Reader interaction point]

Professor Hartwell: The most widely accepted approach today is called the Big Five personality model, also known as the Five-Factor Model.

Blake: Let me guess - it uses continua instead of categories?

Casey: And I suppose it has better statistical properties than Myers-Briggs?

Professor Hartwell: Exactly right, Blake. Instead of saying you're either introverted or extraverted, it measures how extraverted you are on a scale. And instead of four dimensions like Myers-Briggs, it uses five: Openness, Conscientiousness, Extraversion, Agreeableness, and Emotional Instability.

Avery: eagerly Does that spell something?

Drew: rolling eyes Of course you'd care about the acronym.

Professor Hartwell: Well, researchers used to call the fifth dimension Neuroticism, which made the acronym OCEAN. But that term had misleading connotations - it sounded pathological when it's really just a normal dimension of personality. So now it's more commonly called Emotional Instability, though you'll still see the old term in some research.

Blake: So someone could be highly conscientious but also highly neurotic?

Casey: Blake, weren't you just complaining about using the pathological term?

Blake: defensively I'm using the professor's old terminology to make a point about independence.

Professor Hartwell: Exactly, Blake. Or highly agreeable but low in openness. The dimensions are independent. Now, let me walk through each dimension and its facets...

[Professor continues with detailed explanation of Big Five dimensions]

Casey: So this includes Jung's idea about being outgoing, but it's more detailed?

Avery: Right, and it's actually measurable instead of just impressionistic. Someone might be high in warmth and sociability but low in assertiveness. Or high in excitement-seeking but low in positive emotions.

Blake: sarcastically Great, so instead of four boxes we have five boxes with subscores.

Drew: Blake, you're being deliberately obtuse. The professor just explained these are continuous dimensions, not boxes.

Professor Hartwell: Conscientiousness includes: competence - how capable you feel at handling tasks; order - how organized you keep your environment; dutifulness - how much you stick to your moral obligations; achievement striving - how hard you work toward goals; self-discipline - how well you resist temptations; and deliberation - how carefully you think before acting.

Drew: That sounds like it would predict a lot of life outcomes. Organized people probably do better in school and work, right?

Blake: skeptically But correlation isn't causation, right? Maybe successful people just learn to act conscientiously, rather than conscientiousness causing success.

Casey: That's actually a fair point. Which direction does the causation run?

Avery: impatiently Does it matter? Either way, conscientiousness correlates with academic performance, job success, health behaviors, longevity—it's basically the best personality predictor we have.

[Continues through remaining dimensions]

Blake: smirking after agreeableness description I'm guessing I score low on that one.

Drew: pointedly Your skepticism might suggest low trust, Blake. But maybe some skepticism is healthy.

Blake: Thanks for the personality diagnosis, Drew.

Professor Hartwell: chuckling Your skepticism might suggest low trust, Blake. But remember, both ends have value. Low agreeableness can mean you're good at negotiating, catching deception, making tough decisions.

Casey: Though presumably there are costs to being too disagreeable?

Avery: Of course. Low agreeableness predicts relationship problems, workplace conflict, reduced cooperation. It's not inherently good or bad, but there are trade-offs.

Professor Hartwell: after explaining Emotional Instability Now, Blake, you asked earlier about prediction. How well do these scores actually predict behavior?

Blake: Right. I mean, does knowing someone's Big Five profile really tell you how they'll act?

Drew: That seems like the crucial test. If it doesn't predict actual behavior, what's the point?

Avery: impatiently You're thinking about this wrong. Personality predicts behavioral tendencies across situations and time, not specific behaviors in specific moments.

Casey: Blake, probabilistic prediction is still useful. We don't need perfect prediction to gain insight.

Professor Hartwell: The Big Five is better at predicting general patterns than specific behaviors. And the reliability varies by outcome - extraversion predicts leadership emergence, agreeableness predicts relationship satisfaction.

Blake: But who's to say those five factors are the "right" ones? Maybe they just reflect how English speakers happen to think about personality.

Casey: That's actually a sophisticated critique. How do we know these categories aren't just cultural artifacts?

Avery: getting animated That's why researchers tested this across many languages and cultures! The same five factors keep emerging.

Drew: But couldn't the researchers have unconsciously biased their studies toward finding what they expected?

Blake: dryly Everything's a system to you, isn't it Avery?

Casey: Though thinking about personality as multidimensional coordinates does help explain why it's so complex.

Drew: But does this actually help people? I mean, knowing someone's Big Five scores - can you really use that information effectively?

Blake: Yeah, what's the practical payoff beyond academic satisfaction?

Avery: getting excited Of course it helps! Understanding personality differences explains why people clash, suggests better ways to motivate different types of people, helps with team composition—

Casey: mildly Avery, slow down. Let the professor answer.

Professor Hartwell: That's the practical question, isn't it, Drew? But remember what we learned about correlation - knowing general patterns doesn't guarantee specific predictions.

Blake: So if personality psychology can't make reliable predictions about specific situations...

Drew: frustrated Blake, you keep acting like probabilistic knowledge is worthless. But knowing tendencies is still useful.

Casey: Even physics can't predict exactly where a particular molecule will be. That doesn't invalidate physics.

Professor Hartwell: The Big Five gives us much better reliability than Myers-Briggs, and it predicts meaningful life outcomes. But Blake's skepticism about specific predictions is warranted.

Drew: How stable are these traits over time? Am I stuck with my personality, or can I change?

Avery: jumping in Personality shows remarkable stability in adulthood - your ranking compared to others tends to stay consistent. But there are predictable changes. People generally become more conscientious and agreeable as they age, and less neurotic.

Casey: That makes sense. Life experience and social roles would gradually shape behavior patterns.

Blake: Or maybe people just learn to fake being more conscientious and agreeable as they get older.

Drew: sharply Blake, why do you always assume the cynical explanation?

Blake: Because I've noticed that people often say they've changed when they've just gotten better at impression management.

Professor Hartwell: Fascinating question, Casey. The Big Five traits are relatively stable in adulthood - your ranking compared to others tends to stay consistent. But there are some predictable changes. People generally become more conscientious and agreeable as they age, and less neurotic.

Drew: Can knowing your Big Five profile actually help you? Like, if I know I'm low in conscientiousness, can I work on that?

Casey: That assumes personality change is desirable. Maybe the goal should be understanding your traits and working with them rather than against them.

Avery: enthusiastically Both approaches make sense! Some research suggests you can develop skills and habits that help you function better, even if your underlying tendencies don't change dramatically. A naturally disorganized person can learn organizational systems.

Blake: But is that really changing your personality, or just developing better coping mechanisms?

Drew: Does the distinction matter if the outcome is better functioning?

Professor Hartwell: Good question, Drew. Some research suggests you can develop skills and habits that help you function better, even if your underlying tendencies don't change dramatically. A naturally disorganized person can learn organizational systems, for instance.

River, we've covered a lot about personality measurement. Do you have questions about the Big Five, or would you like to explore how personality affects group dynamics?

---

Navigation Options:

• Continue to: HEXACO and the Dark Triad

• Jump to: How Personality Shapes Political and Social Behavior

• Explore: The Limits and Applications of Personality Psychology

• Return to: Correlation and Prediction Overview

1. Actually doing this would require having the script be on a web page with calls to an API so that you can get the AI characters to react in real time to the user’s comment or question. That kind of coding is above my pay grade, but Claude assures me that it can do it.