Chantel Prat, cognitive neuroscientist and author of The Neuroscience of You, joins Beyond the Prompt to explore why curiosity may be one of the most important skills in the age of AI. The conversation covers how brains learn, why people respond so differently to new technology, and why feeling threatened by AI can prevent the very learning needed to benefit from it.
Chantel Prat studies how different brains make sense of the world. Her work starts from a simple idea: every experience leaves a mark. The inputs we consume shape how we think, what we notice, and ultimately who we become.
The conversation explores why people often choose familiar rewards over uncertain opportunities to learn. Chantel explains the tension between exploration and exploitation, why curiosity is essential for growth, and how fear can prevent us from engaging with new technologies like AI.
They also discuss theory of mind, cognitive offloading, and what happens when we increasingly rely on AI for thinking. The goal is not simply to do better work, but to use AI in ways that help us become better versions of ourselves.
Key Takeaways:
Chantel Prat: linktr.ee/chantelprat
The Neuroscience of You: The-Neuroscience-of-you/book
00:00 Curiosity Versus Threat
00:31 Meet Chantel Prat
01:02 Why Input Shapes Brains
04:08 The Output Pressure Trap
05:52 Exploration Versus Exploitation
10:05 Average Brains And Teams
15:35 Theory Of Mind Defined
22:12 Practicing With AI Feedback
24:31 Offloading Thinking To AI
29:50 Humans In The Loop
35:16 Age And Tech Reactions
42:15 Why Curiosity Requires Safety
48:15 Personal Codex And AI
50:54 Becoming More Yourself
54:34 The Debrief
📜 Read the transcript for this episode: why-fear-kills-curiosity-and-what-that-means-for-ai-with-chantel-prat-cognitive-neuroscientist/transcript
[00:00:00] Chantel Prat: The truth is that if you think you already know the answer, you will feel zero curiosity. And if you feel zero curiosity, your brain is not set up to learn. And that's just if you think you already know the answer. If you feel insecure, you think it could be better than you, you think it could take your job, then it's like not just not knowing the answer, but feeling some kind of threat.
That would cause you to actively sort of create these defense mechanisms and move away from this new experience, whether it's AI or a coach.
Hi, I'm Chantel Prat. I'm a cognitive neuroscientist who studies individual differences and the way brains make sense of the world. I'm really excited about this conversation and learning how AI can help us be more of ourselves and more of who we want to be in the future.
[00:00:47] Jeremy Utley: How do you think about, as a neuroscientist with significant output requirements, why is input important? And then how does one, knowing that it's important, how does one protect it?
[00:01:00] Chantel Prat: Oh, that's such a good question. Okay. First of all, why is input important? Like, this is my absolute favorite thing about... Well, gosh, I have a lot of favorite things about the human brain, but I think one of my favorite things is how adaptable it is, right?
Like, when we talk about... And I should s- I should, like, sort of center this by saying my specialty is individual differences, so I'm always focused on the unique ways our different brains make sense of the world. And every brain really is different, shaped by our genes and our lived and imagined experiences.
And I think there are lots of kind of pop psychological examples of how we miss this. We think that other people think like us, whether we think in words or pictures or, you know, how we make sense of things. But in fact, our brains are shaped by our inputs from the beginning, you know? So language is one of my areas of expertise, and a newborn infant can hear all of the sounds in every language that any human makes.
Not perfectly, but they can make all of the distinctions. Mm. And they become tuned to their environment, right? So in the place of instincts that let other animals kind of hit the ground running, we have a brain that's born to learn about our environment, how to succeed in the place that we were born in.
And then from every input, our brain restructures in ways that it believes is setting us up for success in similar environments, right? So we're always learning. Every thought is a learning event. And, and that learning is preparing you for some imagined future.
My husband, I'll give him credit, Andrea Stocco, he teaches cognitive psychology at, at our university, and, um, he uses the, the metaphor of a footprint in the sand as the way that a memory or an experience changes your brain. So it's not the case that any idea or any input has a, a local effect. I mean, we've got 86 billion neurons that are, you know, tasked with this human exper- experience, and much like a footprint in the sand, a- an experience makes its mark on your brain by moving probably hundreds of millions of neurons into different patterns of connectivity with one another.
So newness is really, to me, it's very, very important because you, you might have this experience, you might remember it, you might say, "Oh, when I did this, this happened." But it also changes the shape of you. It changes the statistics that your brain is taking in a way that you may not be able to appreciate consciously. So for me, input is just everything.
[00:03:55] Jeremy Utley: Well, uh, I agree. I mean, I'm obsessed with input. Uh, anybody who listens to this podcast and anybody who listens to Mind and Head versus Private Conversations, I would say they're primarily input oriented. I mean, we are, uh, input obsessives. So here's my question. As someone who has been to South by Southwest many times, you're just coming off of your South by Southwest kind of afterglow, um, why is it even for those of us who love input, there is this really challenging pressure towards output?
So the way that manifests for me is when I'm at South By, I'm looking at the schedule of unknown inputs that can have indiscernible impact on my life in lots of ways, and then I've got phone calls that I can join. Mm-hmm. And I've observed in myself, sometimes I sit in my hotel room and I, I schedule a stupid phone call that I could do any time.
It's almost like this out of body experience where I'm seeing this being who supposedly, and at least theoretically, values input, who's in an input rich environment and goes, "Eh, you know what? I'm just gonna, I'm gonna, I'm gonna go to what's comfortable." Why is that? Mm-hmm. Why is it so hard to make space?
[00:05:02] Chantel Prat: We can go all the way from society to habit to answer that question, right? And I think it's quite interesting 'cause on the one hand, you know, I think I didn't answer the second part of your question, how do you protect time for that? Yeah. Right? Because we're rewarded for output. Mm-hmm. Right? And in fact, even your brain is taking your inputs and it's using them in ways that it thinks will maximize reward.
So it doesn't just take this as it is. Your brain has no interest in representing your experiences as they are. Zero. There's no, like, in fact, doing so would be disadvantageous. Like, if you tried to process everything that was going on, most of it is not important to you. Mm-hmm. Mm-hmm. And that's how that adaptability, that experience thing starts to feed itself and make you...
It can make you pointier, right? Like, if you keep doing the things that have been rewarding in the past, you can get really, really good at fewer and fewer things because your brain is saying, "Hey, last time you had a call, you got this done," or, "Last ti- you know, you made this connection, you opened this door."
Um, and so there's always this trade-off between let me try something new, um, let me get some new inputs that might sort of shift the way I understand the world, and this thing gets me-
[00:06:20] Jeremy Utley: With a known output- ... success ... known deliverable. Yeah.
[00:06:22] Chantel Prat: And that's exploration, exploitation.
[00:06:25] Jeremy Utley: How do you personally protect it?
[00:06:26] Chantel Prat: Practice and, you know, and not al- it's not always easy. I remember in my podcast with Adam Grant, who I met at South By, which was amazing, um, we actually ran into each other on the escalator, which was funny. But, um, he said, "Wait, neuroscientists have a knowing-doing gap, too?" Yes, of course, we do, right? Mm.
Like, the thing that we're trying to study is, is ruling us as well.
[00:06:47] Jeremy Utley: Sure. I
[00:06:47] Chantel Prat: think that people differ. So you're asking how I protect it. For me, learning is really important to me personally. I mean, it's what I study. Um, but it's a value I guess I would put it as a value, so because learning feels like a reward to my brain, it makes it easier.
Like I would say I need to eat, I need money to support, you know, my lifestyle. But in fact, the brain treats information re- rewards exactly like food rewards. The dopamine pleasure centers respond to a question that makes you feel curious just like it would to a piece of food if you're hungry. And that varies depending on the, the information to be gained, and it also varies, uh, depending on the person.
But for me, learning is a primary reward. So I think that I could s- I could sit here like I'm preaching- You can do it for its
[00:07:43] Jeremy Utley: own sake, yeah ... this
[00:07:43] Chantel Prat: value, but it's just in my brain that's already intrinsic. Intrinsic.
[00:07:46] Jeremy Utley: One thing you made me think of is you meet Adam Grant. If you had known in advance you'd meet Adam Grant on the escalator, it's probably worthwhile, right?
Like last time I was at South By, I met Bob Metcalfe. I was in a session and somebody referenced Bob Metcalfe and said, "He's actually up here," and he waved his hand. And I'm like, "Wait, that's Bob Metcalfe? Like Metcalfe's Law Metcalfe?" Yeah. So after the session, I just walked up to him like, "Hey, you don't know who I am, but I'm a huge admirer of your law."
I... Oh, and then we had this amazing conversation, right? If you had told me a priori, "Hey, dude, you're gonna meet, like, a legend of computer science," I would have put more valence on the open-ended task of inspiration. Part of the challenge is it's an undefined reward, and your bra- when you're making decisions between...
It's like m- it's like a roll of the roulette wheel. It's a one in 100 chance I'm gonna meet someone world famous here, or it's a guarantee that I'm gonna move some initiative forward 1%. Mm-hmm. That's exactly, exactly. We often will, we often will take that, that 100% guarantee of a 1% improvement
[00:08:52] Chantel Prat: And that's exactly the space of exploring versus exploiting.
And the other thing is, like, the older you get and the more experienced you get, the more data you have, right? So, um, I wrote about this in my, in my second book. I was like, do you go to the same restaurants over and over and order the thing that you know you love? Or do you like to try something new? And, you know, thinking about how I love eggs benedict, but it, you know, and I'm 51 years old, and I've eaten a lot of eggs benedict.
So I have a lot of data now about, like, how good an eggs benedict could be. You know, when you're a kid, it's like, I don't know, you got smushy this or smushy that. You don't have a lot of data, so you don't know, like, how good the, the thing that's for sure is. Um, but, you know, the better you get at the game of life and the more you know, like, okay, this eggs benedict is really this good compared to all of these other potential eggs benedict, it kind of becomes a, a bigger gamble to try a new eggs benedict at a new place.
Huh. But there could be a better one out there, right? So that's, that's the explore, exploit, um-
[00:09:52] Jeremy Utley: Dilemma ... game. It's a personal breakfast dilemma, and it's a- ... it's a corporate existential dilemma, and everywhere in between.
[00:09:58] Chantel Prat: Heinrich, I think, Heinrich, am I s- I'm murdering your name. Say it for me, please. Please.
It's cool.
[00:10:03] Henrik Werdelin: I love the energy. I'll change the topic then, 'cause I was curious on some of your, uh, research about the brain of being average. Mm-hmm. And I was curious about your, uh, thoughts on us all using the same model when our brains are this different. Is that a mistake? And is there anything we should do to kind of make sure that these models kind of adapt to our unique brains?
And maybe just talk a little bit about the research if that's relevant, uh, to- Yeah. Yeah,
[00:10:37] Chantel Prat: yeah, please. So I'm happy to talk about the research. So I would say the idea that we use- An average brain, but I think more so we tend to assume that other people's brains work like our brains and that our brains are average.
Like, both of those things are, I would say, mistakes. Social, the field of social neuroscience has shown us in the past, like, three years that you can predict how, uh, closely people will affiliate, how good of friends they'll be, um, based on how similarly their brains work. So the first study was actually done in a cohort of 350 or so graduate students and, um, they put them in a scanner and showed them lots of little vignettes and watched their brains change in response to these different inputs, to go take it back to inputs.
And you could decide if somebody was going to be friends or if maybe they had one friend in common. How, how, how far away on the social network they were was highly correlated with how similarly their brains respond. And I think, you know, whether it's a mistake or not, I think what happens is when we assume that other brains work like our brains, it helps us, we get the inferences of understanding someone else right more frequently when they do work like we do.
So new research is showing that teams with brains that work differently are more effective than teams with people who have, you know, five brains that all work the same as long as we can understand one another. So I think is it a mistake if you're trying to work with somebody whose brain doesn't work like yours?
That limits you
[00:12:15] Henrik Werdelin: a lot. And do you think that some people's brain works, uh, less well with the AI models? Yes. If we assume that they act, uh, like a, like a brain.
[00:12:26] Chantel Prat: So I would say some people can understand the AI better. The AI is this aver- I mean, in, in some ways we would say it's Unlike each of our brains, which are kind of this kaleidoscope of differences, the AI is closer to an average.
So I think what we learned about a face is if, if we take all the human faces and morph them together, we'll think that face is pretty attractive. So in some ways I think the average AI is kind of l- the optimal attractiveness to all of these different brains. But I do think some people will be able to understand it better, have a better model of the AI.
And I think the thing that makes us... I'm thinking about this a lot, actually. I think the thing that makes us a successful teammate with other humans probably also drives our ability to be a successful teammate with AI. And I think that that's about knowing what you're uniquely good at, and knowing which areas you wanna grow in, and knowing where other, other intelligences might be able to help you.
[00:13:27] Jeremy Utley: Could you explain theory of mind? Because I think it's related to this point. For folks who aren't familiar with, we can talk about some of the research related to theory of mind and AI performance, but just start, as a starting point, what is theory of mind in the world of neuroscience?
[00:13:41] Chantel Prat: It's amazing because that's a, a natural next step to the question I just asked about are we getting it wrong when we try and understand others.
So the most instinctive way of understanding another is through mirroring. So we have these mirror neurons. Um, you know, we know that newborn babies will kind of mimic faces of the, you know, their caregiver, and this is probably the source of empathy or feeling with, is that we have a set of neurons in the brain that when you see an action of someone else, the same neurons that are involved in causing you to do that action are involved in perceiving it in someone else.
So they accidentally discovered these neurons in a monkey lab. They were looking at motor control and hand grasp, and the experimenter, it wasn't even a monkey, it was a human, made the same motion to put a treat on the, on the board for the monkey, and the neurons that are involved in the action went, started firing when they saw someone else do it.
So, so we learned that, um, we have these set of, special set of neurons that are both involved in recognizing a behavior and in executing that behavior. So to me, that's kind of like the putting, the idea of putting yourself in someone else's shoes. And what I think is that This is our effortless way of understanding one another.
This is not theory of mind. This is like a, a simulation, and I think that that works well if your brain works like that brain, right? Because- Mm-hmm. Mm-hmm ... if you see someone doing that and you're thinking, "That is so idiotic," or, "That makes no sense whatsoever," that's what putting yourself in someone's shoes is.
It doesn't work if they're trained on a different set of inputs. It doesn't work if their brain has different set of weights. Um, but it works if they have, you know, similar brains, which is why we find each other in the wild. Theory of mind, the way I like to describe this, is that it's a way of modeling.
It's an energetically expensive way of understanding another person, like you would try and understand quantum physics. It's a way of learning the facts and building a model of something you can't directly see. In this case, it's someone's mind. So when we're trying to understand someone else, we have a set of observables.
We have the things they say to us, the ways they behave, and rather than just saying, "Oh, if I were doing this, I would feel this way," or, "I would," you know, "This is what would cause my brain to do that," you can say, "Okay, given what I know about psychology or neuroscience or teams or blah, blah, blah, I can imagine, if I take the simple premise that this person is behaving in a way that is rational to their brain based on their experiences, what might I reverse engineer?"
And that is what I think a theory of mind is. I think it's a frontal lobe function. I think it's energetically expensive. It's a way of trying to say, in what circumstances could this person's behavior make perfect sense based on their experiences and their values and the way their brain works? Put it together.
[00:16:57] Henrik Werdelin: And how good do you think AI is to do that? So when you ask it to mimic theory of your mind on yourself- Mm-hmm ... do you think the output is, uh,
[00:17:07] Jeremy Utley: real? Well, well hang on, Henrik. So just so you're, just so you're aware, the reason that I mention this is because there's some studies that are showing now that folks who have good theory of mind for AI, not the AI has good theory of mind of them.
Hmm. But individuals who possess this sense of, "I think I know what it's thinking right now," actually outperform people who don't have good theory of mind.
[00:17:31] Henrik Werdelin: Oh, interesting.
[00:17:32] Jeremy Utley: And so I, I don't, I don't know how much it's related to empathy. I would actually be curious to hear your thoughts on that, Chantal.
Like, is, is theory of mind- Close to empathy? Is it synonymous?
[00:17:44] Chantel Prat: No. I think it's actually different. And I keep citing this book, Paul Bloom wrote a book called Against Empathy: A Case for Rational Compassion. The title, I think, is very evocative because again, empathy works when the person works like you. It helps you feel with somebody who works like you.
But all of the things that enhance empathy also drive in-group, out-group biases.
[00:18:09] Jeremy Utley: Interesting. Um, whereas- Tribalism ... rational
[00:18:11] Chantel Prat: compassion... Exactly. And so it's, I don't think empathy is the answer that everyone wants it to be. I don't think it pulls us closer, whereas rational compassion is like thinking yourself into a place where you can- Yeah
feel with. Does that make sense? Yeah, yeah. Like, I think empathy is fast and automatic, and will bring you closer to people who you might have already been predisposed to be closer to, whereas theory of mind is a way of trying to do the work To understand somebody who might be really showing up in a way that's really different from
[00:18:43] Jeremy Utley: you.
And so it's c- maybe take a stab at hypothesizing why it is that individuals who possess a strong theory of mind ability end up doing better with AI than those who don't.
[00:18:58] Chantel Prat: I would say that you're forming a, a more accurate model of why this algorithm is behaving the way it's behaving, right? And I think that especially with LLMs or chatbots or, or things that are designed to behave, they've been trained on our linguistic behavior, so they're designed to emulate humans.
So I think that's really gonna bring out our bias to, you know, this is like a f- kind of like a thinking fast and slow thing. Our bias is gonna be, well, if this thing is speaking like me, then it knows things like I know, and it's behaving because X, Y, and Z. In a way, I think there's a good reason to believe that theory of mind sometimes requires overriding that faster assumption that this thing is behaving- Yeah
like you. And so, so a little bit more research is that it develops. It's not something we're born with, and it's seems to be, like the most compelling study I saw had 1,000 pair of five-year-old twins, so that's like 2,000 individuals, and theory of mind seems to be entirely learned. There was like zero genetic factor, which I've never seen like a, a cognitive thing that had a 0%, uh, genetic contribution, and this was a big study.
[00:20:20] Jeremy Utley: Okay, so, uh, so then, I mean, it begs, begs the question for me of how do we learn it? Right. If you take, if you take as a given, one, that folks that possess strong theory of mind outperform with AI, and two- Mm-hmm ... that's part of our kind of purview on this podcast. Right. You are a theory of mind expert. You just said it's 100% learned.
Right. So to me, this is actually, this is really interesting. How do you learn theory of mind?
[00:20:43] Chantel Prat: Yeah. So I think you need to learn, just like quantum physics, you need to learn the operating principles. For a theory of mind, it's like you need to understand not only how minds work, but how different minds work, right?
Because you're al- you've always got the observables. I mean, in a lot of ways, I think that the human mind is harder to understand than quantum physics because there are so many parameters, and again, because our default is to think, uh, that we're the same. We're all the same. We all think the same. So for me, you know, the idea is get the data.
In my book, The Neuroscience of You, which is about brain differences and how different brains make sense of the world, goes through sort of there are seven chapters. It's called The Neuroscience of You 'cause it's supposed to help you understand your own brain, but in every facet from, like, neurochemistry to, um, laterality, how the two halves or hemispheres of your brain work differently.
It talks about variability. There are little tests you can do, and so you start to learn how people differ. You know, I, I alluded to this idea that some people think totally in images. Some people can't create an image in the mind, and just knowing that, you know, the more you learn about these differences, the more tools you have for sort of reverse engineering.
So for AI, I think it's the same thing. You need to understand how they differ. They also are, are not all the same. They don't have the same inputs. They're not designed for the same job. They're not optimized for the same rewards. So understanding, you know, what that problem space looks like, and also I, I think this is something, Jeremy, that you sort of pointed out to me, like asking your AI questions about itself.
Maybe that was in your talk, right? Mm-hmm. Like they're an expert in them, like ask them how they work. And I think that we can do that with people, too. In some ways, I might make the argument right now, and I have not had this thought for longer than five seconds, so I hope it's a good one, that AI might be a good way to practice this theory of mind building because with people, something that I feel very fortunate about is having a trusting relationship where you can ask and get it wrong and get feedback.
[00:22:45] Jeremy Utley: Mm-hmm. Mm-hmm. Like
[00:22:46] Chantel Prat: the I, "I think you're behaving this way because blah, blah, blah." Like, "I'm glad you said that 'cause that's totally wrong," you know? Usually you just sound like a jerk, and you get it wrong, and then that person thinks- You know, you're, you're a bad person or whatever and they never wanna talk to you again.
But your AI, it's like you can start with a, it's probably not a trusting relationship, but it's, it's a place where you can get feedback and learn how to think about thinking.
[00:23:10] Jeremy Utley: Yeah. Non-judgmental- Yeah ... for sure. That's one thing I hear from a lot of people is I, I f- find myself saying to AI things that I wouldn't say to a human because I'm self-conscious to say it to another person, but I'm not concerned about AI judging me.
[00:23:23] Chantel Prat: Yeah. So getting that feedback about how do you work, how do you learn, i- as a practice of working it out with a per- in front of a person where you're like, "Here's, you know, what I think is going on." Because in trusting... I think just understanding what's an assumption and what isn't, um, and then getting f- we learn from feedback, right?
To be good at theory of mind at AI, I would want to learn how this AI was trained, what it's optimized for. Like, for me, um, and of course my brain is tuned to hu- the human experience, so I probably err too far on the other side of like this is a r- this is a, a non-living thing. Like, my husband is constantly angry at me for being rude to Alexa.
I'm like, "I'm not rude to her or them. It's just, like, a machine," so I don't know. I'm just, like, being efficient. Um- It'll
[00:24:17] Henrik Werdelin: come back to haunt you on one point. Suddenly, like, Alexa go like, "Remember this five years ago when you were being mean to me?"
[00:24:24] Chantel Prat: Call... When you called me Siri? Exactly. I never forgot.
[00:24:29] Jeremy Utley: That's hilarious.
[00:24:30] Henrik Werdelin: Can I ask you a question on the offloading stuff? Um, when you talk about- Please ... y- you mention the systems one and system twos, uh, from Thinking Fast and Slow. I've heard some people talk about AI now as being system three- Mm-hmm ... which is basically this thing you now can offload part of your brain to.
What's gonna happen to the brain when we start to offload a lot of the thinking or I guess already are offloading a lot of our thinking to it?
[00:24:56] Chantel Prat: I think exactly what, what you suspect. So this is kinda like what I'm thinking about in centering the human in this AI teamwork. Remember that, you know, I started this conversation by saying we adapt to our environment.
I think the human brain and the AI are co-adapting. So if, for instance, you occupied a world where you never made a decision or you never made a decision without heavy input from AI, you would get worse at it. That's, that's just- True. So I think that, um, when I'm thinking about my future, my partnership with AI, I'm thinking not only what am I uniquely good at, like what is, what do I think that it can and can't help with?
I'm also thinking really intimately about what I wanna be better at and what I don't care about being better at for these reasons, because we are going to co-adapt and whatever we off-put, it should be, I think, as a discerning human, I think we should think or learn and acknowledge our biases in these areas 'cause I think it interacts with expertise.
Do I think this is something that AI could be better at than I am? And certainly there are plenty of spaces for that. And is this something I don't care about being good at? So, like, that to me is the, is the magic place for if we're gonna offload something. And I think otherwise it's like, what do I wanna be better at?
How can AI help me be better at something, or even point out thing, areas for me to be better. But I think th- we have to be really intentional about the difference between being better and doing better.
[00:26:37] Jeremy Utley: Mm. Mm-hmm. Mm-hmm. Yeah. '
[00:26:39] Chantel Prat: Cause if you're, you know, like if you're, let's say for instance, something that doesn't happen anymore but used to on social media, you'd like make a post and then they would say, "Do you want this to be funnier?"
And I, I would be like, "How rude." Like, absolut- you think you can be funnier than me, number one. Number two, do I want my post to be funnier or do I wanna be funnier? That's different, right? 'Cause if you start just writing boring stuff and then letting it make your post funnier, I think that takes... Maybe not.
Maybe not. Maybe you could be like, "Oh, that really is funny," and I learned from that and now I wanna, you know, write a post more like that. No, I think that's a very good point. But I think you have to think about whether you want your thing to be-
[00:27:19] Jeremy Utley: Well, you, I mean, o- one thing I, I really resonate, Chantelle, with this idea of you're not gonna get better at whatever you give the AI, but it's, it's worth asking what do I wanna be good at?
And for example- Mm-hmm ... I mean, I remember when we were in high school or college, you know, it, it was a bragging right like how many phone numbers you knew, right? Yeah. Yeah. I don't know if you remember this. Like- Yeah ... and it's kind of, it's a proxy for how close is our relationship, like whether I know your phone number, right?
Yeah. Right. I no longer pride myself on whether I know someone's phone number because- Right ... it's irrelevant. And I've, I have offloaded that task to a smartphone and it's no longer th- there was a point in my life when I did want to be good at it. Mm-hmm. It's no longer needful. But to your point, the question isn't am I better at knowing phone numbers, it's am I better at being in touch, right?
Mm-hmm. Am I... If I'm more in touch with people whose numbers I know less, then actually that's n- a net benefit, right? Correct. If I'm less in touch, then maybe I should devote more attention to memorizing phone numbers. I don't know what the or the- Right. Right ... you know, corollary is.
[00:28:20] Chantel Prat: Yeah. And I think it also goes to like navigating, right?
It was so funny 'cause I just had this whole c- conversation with my daughter who's visiting about, I don't know why somehow like a picture of a dance club in the '70s came up. I have no idea how this happened. The idea that people danced in the '70s, that was before my time even. But then in my time, how did we arrive at a place?
How did we navigate to a place without a phone?
[00:28:45] Jeremy Utley: Right.
[00:28:46] Chantel Prat: And how did we, you know, how did we do all of these things? And, and I realized that navigating is a skill that to me is like if you do wind up in a, a new place and you don't have cell signal or your phone dies and people like forgot how to ask for directions, or you have n- no sense of where your hotel is compared to where you are, and that's interesting.
And not everybody cares about that, and some people do, but almost everybody is worse at it. than they were when I was a kid, you know? Sure. And I told her about, like, me one time trying to... I was skipping school and trying to wind up, I'm from California, I was trying to wind up in Santa Cruz at the beach, and I went the totally wrong way and wound up at my dad's house in the middle of a day where I was supposed to be in school, and it was, like, 200 miles the wrong direction.
But I thought these things were familiar, and I could navigate in this way, and I could learn, but it was just a, it's different, right? So it's not that that's better or worse, I just think that we wanna be mindful about what we wanna be good at and what we wanna do well, because those are not, like, produce well, because I don't think those are the same.
On
[00:29:51] Henrik Werdelin: the thing on skills, um, maybe this is just a- an ask for- Yeah ... for help. So I've been going through a bunch of organizations that are trying to kind of re-found themself, you know, and to become AI native. Mm-hmm. And one of the things we've done is we've identify where do you need humans in the loop, and where do you need specifically humans in the loop for that organization, because it could be humans in the loop somewhere else, right?
Like, so for Bark Box, for example, we make dog toys. Mm-hmm. And one of the humans we need in the loop is a guy called Derek. He make, you know, he's the head of design, and he makes what we think as Bark magic. He makes these crazy, quirky, uh, wonderful toys that no AI algorithm can make right now, right? He and his team.
But-
[00:30:32] Chantel Prat: Like this? Like my radish? Oh. My Bark Box radish. Ulrich. Aw, that
[00:30:37] Henrik Werdelin: makes me so happy. That's so cute. Thank you, Derek. Aw, thank you, Derek. But here's the question for you, and here's, like, the- the- the- the thing that's a mind knuffle for me. The more that I'm trying to define what is it that Derek can do- Almost the less I need Derek to do it because the sheer fact that it's difficult to explain what his cryptic mind kind of creates, it makes it more difficult for me to get an AI algorithm to do it.
And so how should I think about this thing of the uniqueness of the brain when there's like this urgent need for corporate America to define everything, put OKRs on it, put K- I've literally had this, uh, conversation where somebody asked me, "What's the KPI of Bark Magic?" And you're like, "You can't freaking define it.
That's the whole point." Do you understand what I, what I'm trying to get to? It's like we have this interesting thing where you say, "What do you wanna be good at?" If you could define it, it wouldn't be magic. Exactly right. Um, so we have this interesting conundrum now where I might have things that I feel that I'm good at, but I might not have the vocabulary to really explain it, and the better I become of explaining it, the less it becomes a unique quality of mine.
[00:31:52] Chantel Prat: Yeah. I would say I, I think this is a problem I can't solve for you because I'm really in it, so I can give you my version of this is like, there was an experiment that came out in Nature in 2025 where a large language model was trained on like 60,000 pieces. Maybe this is an extreme example, but I believe it's, uh, uh, my, my lane of the same question.
So they fed this large language model experimental data from 60,000 different tasks. There were 625 tasks and tons of, tons of undergraduates doing these tasks, and the model could learn to predict behavior Better than any existing cognitive model designed for one type of task. And so I found myself asking the question, like this model does not think.
It really does not have... First of all, it's doing dramatically different tasks in exactly the same way using exactly the same statistics. So what- whatever it's doing, it's doing, it's behaving like a human using a very different mechanism that humans, um, use. But I found myself asking like, okay, I'm a cognitive psychologist, so my whole jam is understanding the black box, understanding how attention and memory and procedural, um, knowledge, all the things you can and can't describe, limit, constrain, and support human behavior.
But I'm like, if we don't need it, you know, if this model can do it without those things, am I trapped? Is cognition real? Is that like a, we don't know what consciousness is, ri- right? So like I think that what you're describing is like how do we think about this thing if we can't quantify it, if we can't put it into words, which is now the coin of the realm because the, the l- large language models know what they know by reverse engineering our language.
Is it important or what is it? And so then I'm like, wait. You know, so I'm in the, I'm in this sort of identity crisis where it's like, am I the human mind trapped in the hu- is the human mind even like useful? Is it limiting? You know, so, so I, I'm sorry if I can't solve that problem, but I think I understand it deeply.
It, it sounds
[00:34:06] Jeremy Utley: like at the very least Derek has job security, and I think that's really the important thing here. He definitely have job security. I had one question that I wanted to go to, like a totally different direction, but Henrik, do you wanna wrap this one? You started talking after I...
[00:34:19] Henrik Werdelin: I think, uh- No.
But if somebody, I mean, I tell you like after all these years, when somebody shows your product in the, in the wild, it's still like makes your day. Like it makes me so profound happy. So like I'm still a little bit high over that. So thank you so much. No, take it away, Carrie. Oh, and I
[00:34:35] Chantel Prat: have, so you, so I didn't even know that about you, but I've been a BarkBox subscriber for maybe eight years or seven years, so, so I mean, and I'm tailoring that to my dogs, and I know if they're a thrasher or a thiser or a thater, so you got the individual differences in there.
We got a three-dog subscription. Come on.
[00:34:54] Jeremy Utley: Oh my goodness. And, and they
[00:34:55] Chantel Prat: know like now it's a bark bag. It's not a BarkBox. But they somehow know when that thing comes in the door, they're just like
[00:35:04] Jeremy Utley: And it's because don't you spray it with, like, bacon perfume, Henrik? Isn't that right? That's how they know. We did, we did used to do that.
Okay, so, so I have, I have kind of a different direction that I wanna take this. I, I'm very curious fro- from your kind of psychology expertise and what we were talking about earlier, kind of going back to the beginning of the conversation with novelty, with learning, and I wonder if there are differences in regards to age.
You know, and you mentioned that, like, a Hollandaise at 51 or is- Yeah, yeah ... very different than a Hollandaise at 15. So I, I'm, I keep thinking of this Hitchhiker's quote, or I'm, I'm actually gonna read it. It's from Douglas Adams, who's the author of The Hitchhiker's Guide to the Galaxy. He wrote this in a nonfiction piece.
He said, and I quote, "I've come up with a set of rules that describe our reactions to technologies as people. One is anything that's in the world when you're born is normal and ordinary and just a natural part of the way the world works. Two is anything that's invented between when you're 15 and 35 is new and exciting and revolutionary, and you can probably get a career in it.
Three is anything invented after you're 35 is against the natural order of things." Which is just perfect, right? And I wonder, could you talk about why that is from a neurological perspective, and then what are, what are the... And then maybe we could just have a conversation on what are the implications of that on humans today in the age of AI?
But, like, why does that progression happen from it's normal to it's exciting to it's against normality over the course of, say, 40 years of life?
[00:36:32] Chantel Prat: I think your intuition, um, and what a brilliant, what a brilliant quote. I think your intuition that it's about eggs benedict is correct, or that it's related is like, you know, of course, like, when you're born, you have no statistic- you have no expectations.
And so if it was there when you were born, then it's in the water. And I think this is the same thing also with, like, it might have been there is, like, when we have cross-cultural experiences, right? Or we meet somebody who works differently than us. Like, if it's true of us, if it's true of the environment we inhabit when we're born, it's in the water.
It's taken for granted. It's normal. Um, but you know, up through sort of our, our teenage and becoming years, again, it's y- you could actually probably plot this. So I, I would go, like, one step nerdier, if you don't mind- Please ... and say that it depends on the type of input. So things like wavelength of light, for instance, you know, you, you might be like three or four years old, and that you've, like, experienced var- all the variety there is.
You've experienced all the wavelengths of light that your retina can detect, and so your brain kind of closes and it's like, there are this many colors. That's it. And by the way, people actually, some people have four cones, some people have two cones, so, like, even that is a thing that we may experience differently, and we don't even have names for the colors that people who have four cones.
But anyway, I digress. But whatever- Wow ... you have, when it stops changing, your brain goes, boop, that's the way it is, and I'm not taking in new things. So there are different types of input that vary over your lifespan, and once you've gotten, like, this many statistics and you're like, "This, I've, it's not changing anymore," like, then that kinda concept crystallizes.
And so In teenage years, these kind of open things, you know, these relational things, these what are my future potential things, I think are still really, you're trying to figure out who you are, how you're different from others, and who you wanna be when you grow up. So I think that the tools around you are just an exciting kind of part of- Way to pull
[00:38:32] Jeremy Utley: it in, incorporate it, yeah.
[00:38:34] Chantel Prat: Right. And not even only the tools, but the culture around the tools.
[00:38:37] Jeremy Utley: Mm-hmm.
[00:38:38] Chantel Prat: Right? So I think that, like-
[00:38:40] Jeremy Utley: Interesting ...
[00:38:41] Chantel Prat: if there was a new tool and somehow it became unpopular with a, like a, a up-and-coming generation, then I think people... Especially imagine that a, a new technology was created specifically for older people to like, I don't know, monitor their children or so.
You know, like, I think that it's probably, like, a cultural thing too. But I think that this statistic taking and, like, what do I think the likelihood of this technology or this kind of input leading me to good things, as you become older, you think you've figured the game out. You really feel pretty confident about your eggs Benedict and your career and your skills.
Not everybody, but on average, you know, with more data, things start to level out, and especially if you have a kind of narrow pocket of expertise, I think. Um, then this new disruptive thing, right? I think truthfully, um, the likelihood that any new thing is gonna be better than what you already know goes down as you have accumulated a lifetime of experiences.
So I think that new technology, including AI, is, uh, is a way of highlighting our own, the biases that we already have toward newness or oldness as we move through our experiences. And I, I'm seeing this in my lab, and after our first conversation, Jeremy, I'm thinking about this a lot because what I'm observing in the people around me is that some younger people who don't know the game and don't know what works are very eager to have other people tell them what works They're very eager for advice, and those I think are more eager to sort of have AI do a thing for them or show them how to do a thing or make decisions for them because they don't have a history of success.
And I think that, you know, as I checked in with myself and on this panel I was, uh, on in AI, um, at, at South by Southwest, I said, "You know, these same students are, are probably too willing to trust me when I tell them a, like a thing is the right thing to do." Mm-hmm. And so I'm really thinking about teaching them how to discern advice, whether it's coming from me or AI or their boss or whatever, like how to be a critical thinker in the face of advice.
Whereas someone like me or somebody, you know, 10 years older than me might be overly convinced that they already know the right thing. So I, I talk about intellectual humility and how, you know, I said learning is a key value for me, but the truth is that if you think you already know the answer, you will feel zero curiosity.
And if you feel zero curiosity, your brain is not set up to learn. And that's just if you think you already know the answer. That doesn't have anything to do with if you have a skin in the game. You think this thing could... You know, if you feel insecure, you think it could be better than you, you think it could take your job, then it's like not just not knowing the answer, but feeling some kind of threat.
That would cause you to actively sort of create these defense mechanisms and move away from this new experience, whether it's AI or a coach.
[00:41:54] Henrik Werdelin: But it doesn't seem to be just an age thing if people are good or bad at AI. There seem to be- Mm-hmm ... almost a personality that some people seem to really be curious and like the newness and lean into it and so on.
What do you think makes the difference between people who are kind of good at it or not?
[00:42:14] Chantel Prat: Well, I would say, I don't know. Um, that's the first thing I would say in the space of intellectual humility. Our brains do give us a recipe for curiosity, which I think is important. What I don't know is, are people better or worse at it at time one, or are people more or less willing to learn how to work with it?
So from my, you know, perspective, I would guess that the biggest difference is how willing someone is to learn and to partner with it. So one of my favorite models, uh, about, of curiosity by, uh, Mathias Gruber and Charan Ranganath is called the PACE model. And like I said in the beginning of our conversation, when your brain sees that there's a possible information reward in the future, it responds in the same way it would if it sees calories out there to be gleaned.
Hmm. It gives you dopamine. It gives you something that... Dopamine is a feel-good chemical, but it also energizes learning, and it's also a neuroplasticity catalyst. If I asked you two questions, um, give you two trivia questions and asked you do y- and you didn't know the answer to either, but you were more curious about one than the other, and then I give you the answer to both, you're gonna remember the one that you were more curious about, 'cause your brain is saying, "I think that reward is gonna be more important," so it gives you dopamine upfront, which actually lets you lay the memory.
So I think curiosity is really key, but curiosity requires some preconditions. And the first precondition, so the P in the PACE model is prediction. So you need to know that you don't know. You need to be aware that this is something new, so unfamiliar, so your previous predictions fail And that's where if you think you know the answer or you think you're already perfect at this thing, you think there's no growth edge, you won't feel curious.
But the second is really critical, and I think this is probably the biggest barrier for AI, and that's appraisal. A is appraisal, and that's appraisal of safety. So imagine that you find yourself in a new neighborhood. You make a wrong turn. You thought this was gonna be the place where your favorite... I was gonna say record shop.
Like, I have not been to- Go
[00:44:25] Jeremy Utley: there ... a record shop in so many years. It's a Tower Records. Yep.
[00:44:28] Chantel Prat: You know
[00:44:29] Jeremy Utley: that you're safe. There's a Virgin Records store, totally. We're tracking with you.
[00:44:32] Chantel Prat: That was strange, my brain. But, um, you know, are you excited to explore this neighborhood, or are you trying to get the hell out?
There's a safety component there because, you know, we think curiosity is good, and it is, but also we know curiosity got the cat killed. When your brain is curious, it actually turns down the signal on things associated with risk of the unknown.
[00:44:54] Jeremy Utley: Mm.
[00:44:54] Chantel Prat: So before it makes you feel curious, it says, "Am I safe here?"
And I think that in the space of AI, we're talking about psychological safety- Mm ... job safety, evaluative safety. Like, is someone else, you know, monitoring my performance, all of that stuff. So if you understand that there's a growth opportunity, and if you feel safe, then that's predict, appraise. The curiosity is the third step And the fourth is exploration.
So I suspect that it's mostly, um-
[00:45:28] Jeremy Utley: Safety ...
[00:45:29] Chantel Prat: safety, yeah. Mm-hmm. Because also, you know, some people might be really motivated to learn AI because they think maybe they're not curious, but they think this is gonna be the way I get employed, right? Or like their, their employer has thrown a carrot, like the first person to use this is gonna blah, blah, blah.
So you could be motivated for different ways, but that the curiosity is intrinsically a driver for E, which is exploration. And it also just helps you learn better. So given the same number of experiences, the same skills, the same background, if you feel curious instead of threatened, the benefit you'll get from each interaction with AI will be higher.
[00:46:07] Henrik Werdelin: My last question for you is you worked on, uh, kinda neural linking stuff, right? Brain-to-brain stuff. Mm-hmm. So how far are we from us being connected to these AI brains? What will happen? Give us kinda like the, the need-to-know of, of what we're about to endeavor.
[00:46:27] Chantel Prat: Far. We're really far, honestly. And even then, I mean, the resolution...
I say we're closer 'cause we have different technologies for stimulating the brain. So we can read out what's happening in a brain with very high fidelity. Like, we can tell which object a person is thinking about with pretty high accuracy in a, if you put them in a tube and they lay still, not when they're moving around in the world.
But to encode an idea into someone's brain, we have like, we're just very, very far away from the technology. I mean, the things that we did once upon a time were sledgehammer level. Like, you know, make a, a wrist twitch and make someone hit a button, like nowhere near sort of tip inception level thoughts.
What I said then is still true, like the best kind of mind control is words. And even that is very, very imperfect, but like what we could do by forcing someone to sit still and tipping neurons out of whack is so much more expensive and less effective than saying a thing to them, that I can't imagine that that's gonna change our world dramatically.
I think the reading out and translating, maybe having an AI in the middle, somebody who can read from my brain, read from your brain, and get some kind of a codex that helps you under- communicate and find alignment if your theory of mind sucks, I think that would be
[00:47:48] Jeremy Utley: We've actually, we've actually talked on this show about, with, a couple of times about this, the opportunity space around call it English to English translation- Yeah
meaning Jeremy English to Shantel English, right? Or the, the transmission and the reception. You know, you could encode it by Enneagram, right? If you're an Enneagram- Mm-hmm ... five, the way you hear something is r- or whatever. But, uh, I think that codex idea is really cool, and I don't think many people are kinda scratching the surface right now.
But I don't know why you wouldn't, in a team especially, why you wouldn't have your whole team's Enneagram numbers and then say, "Hey, how am I gonna tell Henrik this?" Well, the way to communicate with Henrik is XYZ. Just to kind of, like, add a little bit of,
[00:48:26] Henrik Werdelin: of layer to that is I've seen, uh, recently on a bunch of these founder email chains that people are, uh, creating these markdown files that we are creating anyways for all these agents we are creating.
So the soul.md and all these, like, person.md files they're creating out of OpenClaw. People are now starting to share that with each other- Hmm ... so that as a way to basically say, "This is who I am, and so if you're gonna try to communicate or have your agent to communicate with me, here's a way for you to do that exact thing."
And so a-
[00:49:00] Jeremy Utley: It strikes me, it strikes me that that's, there's some hubris there because in a way you set yourself up to being uniquely exploited. It's like, "Here's my DNA. Feel free to do something." You know, you know what I mean? Like, if someone knows exactly how to communicate with you, it, it stands to reason they, they probably know exactly how to exploit you.
I mean, going back to Shantel, your statement that words are the best way to incept an idea. If someone knows perfectly how... I mean, I don't, I don't have the scientific language for this, right? But it's like I know that a acai berry absorbs a free radical, right? It's like- Yeah ... if you know the acai berry that's gonna make Henrik say yes-
it, it's somewhere in that soul.md- It's the one covered in chocolate ... yeah, it's like the soul.md file there, right? Mm-hmm
[00:49:49] Chantel Prat: I wanna add one little thing to that. I think it's fascin- this idea is fascinating even though I can only barely understand what you guys are... I do know what a markdown file is, but I can only barely understand what format that would take.
It strikes me that, like, my Instagram algorithm understands things about me that I don't understand about myself. Mm-hmm. And I try and reverse engineer it. I'm like, wow, I'm getting a lot of, like, porcupines eating crunchy foods. What does that tell me about me? A- and so in some way, like, having... It is a huge amount of feedback.
These things that have your behavior and that learn to, to give you the inputs that work for you have a different level of awareness. It's kind of like your, you know, your partner. I- m- not all of it is... When you were trying to describe the job of Derek, is that who? The, the, the magician. They have a lot of information about you that they can make explicit that you might not have about yourself, and I think that could be a really important, even a self tool.
And then you decide what you wanna share or how you wanna share. You know, it'd be great if there was, like, sliders. I'll take 75%, 25% transparency or only sharing this, not this. Um- Yeah, and I
[00:50:54] Henrik Werdelin: think, I think, you know, because I do think that obviously with the best way to predict the future is to make it, right?
And I think in many ways when we created all these social media tools, we thought we were just making it exciting to share pictures of food and little did we know we were, like, uh, destroying a whole generation's, uh, mental health. But I think if you take AI, I think obviously a lot of us who are building in this space now are trying to be very self-aware, they're not making the same mistake.
And I do think that this can go two ways. To Jeremy's point, it can be used as something to be super manipulative, but it could also be something to really communicate much more about our true self, right? Whereas social media, if you're showing me in a dating profile through my social media profile, it would be all the vain thing that I want the world to see about me.
If you ask an AI algorithm to completely mine everything I've ever written, everything that's on my desktop, all my browser history, there'll be a much more truth kind of self-exploration. This is a Dan Shipper kind of, uh, point, and so I'll make sure to give him credit. But, and so I do think that there's this interesting argument that AI, while much more kind of intimate and intrusive, could also be something that could make us much more earnest and show much more of a broad spectrum to the world of ourselves, and that could be a good thing.
[00:52:16] Chantel Prat: Right. And to Jeremy's point, objective and non-judgmental 'Cause we're always gonna turn up the volume on the things that society has told us is good. Right.
[00:52:25] Jeremy Utley: Right.
[00:52:26] Chantel Prat: I really like that in the space of what do you wanna be better at, I like the idea of partnering with AI to learn to be better at yourself and maybe to learn how to teach others how to communicate with you.
Also sharing the, the sort of vulnerability and skepticism about giving people that coda.
[00:52:42] Jeremy Utley: Well, you know, I'll build on that and maybe we can wrap here and then do our closing remarks. But I was at a company on the East Coast a couple weeks ago on a... I was being interviewed in a fireside chat, and one of the questions, it was an entertainment company, and so one of the questions that had come up is, is AI more creative than people?
You know, it was kind of like a typical question. And I found myself saying in that moment, I'm actually, the comparison that I want to make is am I Jeremy-er? Not is it more blank than me, but am I more me than me with this, right? Mm-hmm. And to me, that's actually the, the exciting kinda possibility space- Mm-hmm, mm-hmm
is how do we become more ourselves? Mm-hmm. Um, I, I'm not so interested in the comparison between me and AI, but wow, Jeremy augmented by AI- Mm-hmm ... to be more of who he wants to be, that's quite exciting to me.
[00:53:40] Henrik Werdelin: Can I just add one point, and then I think- I totally agree ... to wrap up to get you out on time. I just wrote, uh, a piece on this, and I do think it's very important.
One of the things that I've learned as I use AI a lot is that I need to understand myself much better for AI to really be an amplifier of what I want to be more of, right? But that requires all this self-work to figure out what are my core principles, you know, what are my first principles- Mm-hmm ... and all those different things.
And so I've been trying to take basically all the work I do with my coach and then try to morph that into instructions to the agents that I have working for me so that they can now come back and present stuff that is more me. Which then is a little bit of a m- mind fuck- Mm ... when you then say, "Well, I also need like a much more diverse set of thinking," so now, like, I need to make versions of me that are different, but we'll get back to that.
[00:54:27] Chantel Prat: More me and less me.
[00:54:29] Henrik Werdelin: Yeah, exactly. Much less of
[00:54:30] Jeremy Utley: more me. Thank you. You're amazing. This was so fun. Jeremy Utley.
[00:54:36] Henrik Werdelin: Na, na, na, na, na, na, na, na, na, na. People thought they'll get like a conversation about the conversation, but no, they're getting a, a singalong.
[00:54:47] Jeremy Utley: Well, it's, you know, hopefully we are pitch perfect, as they say.
[00:54:50] Henrik Werdelin: That's because Chantel has such a good energy, huh? Like you kind of feel- Oh, yeah ... a bit chipper after the conversation.
[00:54:55] Jeremy Utley: Yeah, absolutely. Anybody coming from South By, I feel like they get the South By vibes. She clearly was inspired. You know, for you and me both, right, it's this idea of seeking input. I mean, she...
Kind of before we were recording, and then it was kind of we started recording, and then we weren't sure w- if we were recording yet, so I don't know exactly what folks will hear. But the whole premise of not knowing what you're gonna learn, going somewhere for the purpose of provocation, but not exactly knowing what the provocation will be, and how do you grapple with that as a successful person who's, who has opportunity cost to their time and who has kind of known deliverables with known value?
How do you carve out space for inspiration? It's something that I feel like in the totality of our experience we had with Chantel was a- was foregrounded a lot. What do you think about that, Henrik?
[00:55:46] Henrik Werdelin: No, I think you're 100% right. I think without good input, there will be no good output, and I think sometimes when I feel a little bit uninspired, it is basically because I've said the same bullshit over and over again, and I haven't really been out- Right
sucking in some information. And so I think for me it's about, you know, making is thinking, so for me it's about making things, and that's where I learn stuff. And I think you have to be disciplined about it, to your point. It's so easy to say, "No, I should not listen to a podcast," or, "No, I shouldn't read an extra book," or, "No, I shouldn't go to this conference 'cause I've other things to do."
But yeah, it's almost like working out. You need to mentally make sure you don't get obese.
[00:56:24] Jeremy Utley: Yeah, there's always... I mean, at the same time, what's interesting is consumption can be really gratifying in and of itself. You know, it's, it's almost consumption for consumption's sake rather than consumption for the purpose of advancing.
You know, I, I heard Danny Meyer say once, "ABCD to ABCD, always be collecting dots so you can always be connecting dots." Mm. And the col- the consumption is really collection, but the purpose of collection is connection. And so I think there's that- You know, tension and almost commitment to yourself, both to get inspired, but also to do something with the inspiration
[00:57:03] Henrik Werdelin: Maybe that's the yin-yang, I think, of work that you do, and I guess I do in part, is that we get asked to go out and speak places, and that forces you then to create an output.
A- and so that is often the way that you connect the dots, right? Because- Right ... you're being paid to say something interesting. Speaking of interesting- Right, right ... what do you think we should highlight from, uh, Chantal?
[00:57:25] Jeremy Utley: You know, I thought that the, the whole conversation around curiosity was quite interesting to me, that your question I thought was quite insightful about if it's not age, what accounts for the difference in, you know, someone's willingness to engage with AI?
And the, the idea of curiosity, and she had mentioned the PACE model: prediction, appraisal, which precede curiosity and exploration. I thought the whole question about safety was really fascinating because AI is so often framed as a threat and is so often framed as something to be defended against. It's almost, it's
The way it's framed prevents curiosity, and I thought, wow, if organizations really knew that speaking about AI in a way that is threatening keeps their people from being curious, I think they would speak about it differently
[00:58:15] Henrik Werdelin: I 100% agree. I mean, that was my biggest takeaway also, but I think it was quite a big unlock.
I have this thought, and I think I... maybe we took it from something you said, but that there is these four steps of people going through AI learnings. Maybe I think it's Bryce whom- who brought this up in a conversation we had, where one is- Mm-hmm, mm-hmm ... you kind of use AI as a Google; two, you start to have a conversation with it; three, you start to add context to it, maybe connect it to databases or MCP servers; and four, you start to kind of program or you create autonomous agents that go out and do stuff for you.
And the, the thing that becomes increasingly complicated is that the more you're kind of in phase three, four, the harder it is for people who are in phase one, two to understand what the hell you're talking about. There's no emotional really connection. Mm-hmm. It's like talking about having children to people who don't have children.
But what I don't think I ever realized was it's not necessarily just a, a understanding. It's also a, a feeling of safety, that when you start to talk about all these amazing agents you have running around doing stuff, then people are not getting curious about it. They're just getting increasingly scared.
And so you need to find- Yeah, yeah ... a way, even as an AI kind of educator, to, to- Enthusiast, yeah ... maybe even just... Y- yeah, or maybe even just kind of like label that and saying, "Hey," uh, and I think that's w- some of the stuff you s- do very well at your talks is that you do talk about these things in a very non-scary way, which I think is the reason why you invite people on that learning journey.
[00:59:54] Jeremy Utley: It's, it's amazing too how, um, maybe because of the technical language or because of the number of things you can do, I think fear is an, is a very natural response, just 'cause it can be overwhelming. And it's, it's a worthy kind of challenge. I don't have any ready answers, but it's a worthy challenge to say, how do we spark people's curiosity rather than trigger their fear?
I think, in fact, maybe just like sitting with that question and asking our audience even to sit with it, especially, you know, folks who listen to this show who are more AI-forward, more AI-native. It's almost like a good soul-searching question. Is my enthusiasm scaring people or getting them curious? And what can I do with my enthusiasm to shift from kind of a threat to excitement?
[01:00:42] Henrik Werdelin: I did a workshop the other day with some folks, and I in a q- little bit of a different way, but kind of started the whole conversation about that saying, you know, labeling this thing that I felt often that people got more scared than they got excited. Um, and I think even by saying that out loud, it kind of had a big result.
And so I think what you're saying resonates a lot. And so I think maybe, yeah, that would be my biggest takeaway of the, the conversation
[01:01:08] Jeremy Utley: today. We'd love to hear, maybe this is a cool spot for if folks listen and they have thoughts, drop comments in the, uh, YouTube comments and the podcast comments, or feel free to write us as well.
Let us know what are ways that you have found helpful to help people overcome fear and, uh, and, and spark curiosity instead. I think we'd love to treat this as a community learning opportunity.
[01:01:33] Henrik Werdelin: And with that, I think we're gonna say goodbye.
[01:01:37] Jeremy Utley: Bye-bye.
[01:01:38] Henrik Werdelin: Bye-bye
Â