Turi Munthe

How do we escape the confines, the bounds of our own consciousness, of our own strange hallucinations of reality, of our own deep, deeply predisposed opinions and beliefs? And the only way that we

me well.

have any freedom is with the interaction with the other.

The only we and way in which we can break out of our consciousness bubbles is by butting up, bursting them against

you

the consciousness bubbles of other people. And that I find

That I find.

very beautiful as an idea, full stop. But I also find

Mm

it it's more than beautiful as an idea. I feel it's quite an important political statement at the moment when we are fracturing into really highly polarized political tribes and and increasingly.

don't really believe that the other side has anything to advance that's has brings nothing to the table. Actually the idea that we can't think without engaging with the other side is is a salutary lesson for all of us and it turns out to be true.

Welcome to Tastebench. I'm Bruno Marnette. My co-host is Philip Zahn and this is a podcast about AI and taste. Our guest today is Turi Munthe a British writer, journalist, and media entrepreneur. He has written for The Economist, The Guardian, The Times. He was the co-founder of Demotics, a very successful photojournalism platform, and was acquired by Bill Gates' company Corbis back in 2012. Since then, he invested in many media ventures.

And he is also the author of a recent book called Why We Think What We Think, which explores the hidden forces that shape our beliefs and opinions.

hello, Turi Such a pleasure to talk to you today. There many reasons why we're excited to talk to you. First, you had a podcast called On Opinion, which has some resemblance to the kind of topics we're passionate about. But maybe even more timely is you recently released a book called Why We Think What We Think.

And

I'd love to maybe start a little bit with this. In particular, was there's something that I found quite impressive with this book, which is you you're able to cover two topics which were quite big. And I think often often they would be covered in different books. So you have an entire part of the book about what one could call maybe individual intelligence or maybe individual level kind of rationality. Like, why would a single person end up believing something? And then you have another part more about

maybe what you'd call collective intelligence, you how are the importance of debate and so on. And I guess I'm curious, why did you decide to do two books in one instead of, you know, doubling your sales and having two different books?

No, Philip, great to be on the podcast with you. I'm thrilled to come and chat and and looking forward to this conversation. And and very, very good backhanded insult. I like that. You were like, yeah,

Yeah

such a clever thing to to to screw up a book by putting two topics in the same one. Like it, I

haha

saw that. I see what you did there.

Ha ha.

You

yeah, so so sort of high level. The the first part of the book tries to explain the non-rational influences on our opinions. And it sort of takes as a standard model the fact that most of us wander around assuming that our opinions come from some combination of reasoning,

Mm-hmm.

which humans are good at, and socialization, right? You know, what our parents think and what our friends tell us and kind of

Hmm.

the contagion we get from the media and everything else.

And that model is true, but it's just a tiny bit of it. And so there are many other things which influence the ways in which we think what we think. I could have stopped the book there, but it would have had

Mm-hmm.

a massive and then what question. If I'm proving

Yeah.

to you that, you know, the agriculture of your most distant ancestors has an impact on your very deep values, if I'm proving to you that

The number of taste buds on your tongue correlates with your political leanings. If I'm suggesting

Mm.

to you that an enormous number of your most cherished beliefs, whether they're ethical or spiritual or political or anything, may be profoundly connected to your gene pool,

Mm.

I'm kind of fracturing the world. I'm and I'm also and I'm also kind of destroying this any suggestion that we might be able to talk to each other if our because there's a

Mmm.

sort of hint of predisposition.

Predetermination in what I'm suggesting. It's not predetermination, it's predisposition that I'm interested in. But so you need to have then a kind of a a what now? If we are profoundly only a little bit rational creatures, if we are deeply predisposed to our opinions, you what does it make sense to argue? Does it make sense to have politics? Does it make sense to have democracy? Does it make sense? I mean, any of these things. And so that's why I needed part two. And the part two is okay, what do we do with this crazy stuff? If we're not that responsible.

In an active agentic way for our beliefs. How do we get over them?

Mm.

And how do we set ourselves free from ourselves?

How do we escape the confines, the bounds of our own consciousness, of our own strange hallucinations of reality, of our own deep, deeply predisposed opinions and beliefs? And the only way that we

me well.

have any freedom is with the interaction with the other.

The only we and way in which we can break out of our consciousness bubbles is by butting up, bursting them against

you

the consciousness bubbles of other people. And that I find

That I find.

very beautiful as an idea, full stop. But I also find

Mm

it it's more than beautiful as an idea. I feel it's quite an important political statement at the moment when we are fracturing into really highly polarized political tribes and and increasingly.

don't really believe that the other side has anything to advance that's has brings nothing to the table. Actually the idea that we can't think without engaging with the other side is is a salutary lesson for all of us and it turns out to be true.

So that's I think why I disappointed you and put two books in one.

Yes. And I'm curious, what led you to decide what to write about? Did you start in your mind with the second part, where you initially motivated more with the kind of democracy kind of focus and so on? Like, how did you How did you start being, I guess, interested in both aspects?

Bruno, great, great question and thank you. Yes. So I didn't I didn't initially start trying to write a book. I initially started to try and build a billion dollar media company. but as you know, maybe if you s if you aim very high when you when you eventually hit, it still has some value. Anyway, my poor

Yeah.

investors, yeah, they they invested in a media titan, collaborative

Mm-hmm.

kind of wiki media titan. I was gonna map all the world's arguments and I've sent them both hard copies.

of of the book. I'm not sure it's a great ROI for their LPs. Anyway,

Mm-hmm.

so yeah it started off as a media project. I was trying to map the world's opinions and the people who held those opinions. And and quite quickly I realized what a catastrophe that was because I was wandering around with an 18th century model in my head. I was trying to build a kind of a Circle de Lumière enlightenment

Mm.

encyclopedia and

Turns out people don't think that way. And so I s the I I came to this book and via a podcast as I was realizing that my giant media venture was floundering. I st I started a podcast trying to understand actually why people think what they think. And i i that's what ended up becoming the book. But you're absolutely right. The reason I decided to put myself through the agony of

you

trying to write a hundred thousand words on this topic was was actually the of course, the politics of it, which is

In a highly polarized context in which we find ourselves when I don't know where where where Philip, where are you based? Sorry, you're based

Germany.

you're based in Germany, which is having a great time when it comes to radicalised politics. Bruno, you're based in London, which is having a for out you're in France now? Originally French. Exact Yeah, exactly.

Yeah. France. Yeah, I'm originally French. I don't know, even London. That's it.

And and I'm based in Italy, which has currently got an ex army general who is trying to take I I mean, you see my hand gesture? That that

Yeah. Yeah.

Mm.

that's where

That's where they're heading. Yeah.

Yeah, for your podcast

listeners, that was a fascist salute, who's trying to take us back there. that we're we're seeing this radicalization. And I came across so many very worthwhile attempts to fix it. You know, in Germany, for example, I think it was die Zeit, got was it die oder spiegel which got people on other side either side of the political spectrum to have coffee together? And then

Mm.

The Guardian picked it up as well. And I spent a bunch of sort of

academics and activists in the US trying to work out how you bring people back in the middle. And none of them really seemed to work because they were they were premised on this, on a kind of a liberal notion of manners. Like it's important to speak, to let the other side speak because

Mm-hmm.

because because you sort of have to, because it's manners, because that's civility. And it it feels like a poor argument. It felt like the argument needed to be made again. And

Because of course you should let the other side speak because it's because it's manners. But sometimes if you really, really disagree with the other side, you're not better off shouting them down. There are certain things which we should be shutting people down for. and so as I discovered more about the kind of

more about.

tremendous spread of our opinions, political opinions particularly,

Mm.

and I discovered more about the way that we truly think, which is in conflict with others, it seemed to me that you could make

A case for pluralism on almost evolutionary grounds, on the grounds of deep human progress, but deep human progress going forward with evidence of what deep human progress going back tens of thousands of years also brings us. Because it turns out that reasoning is a team sport. It's collective. We think better in conversation with our opposite number. And so that as a

Mm-hmm.

message to bring to a f fracturing, polarizing world feels powerful because we're essentially arguing for liberalism, or some version of liberal pluralism, on on biological evolutionary grounds.

The dots are connecting very naturally for me as well. I can see how you go from what we see on social media being maybe a mirror of that polarization, why would people seem to have a very exaggerated view of the danger of other people having different opinions. And I guess you're trying to de-dramatize this by showing that actually difference of opinions are very natural. They come from differences in...

Like you mentioned earlier, agriculture, genetics, all factors that are outside of our control. I guess what I'm curious about is the, when do you see the demand for pluralism? in the, because that's, I had a couple of years ago an appetite similar to you. Like when, when Chat GPT came out, I thought one of the wonderful use of Chat GPT among others could be maybe can depolarize the world.

by just helping people having debate conversations, getting people to the same platform where they could exchange both sides of the argument. And then quickly, the fear I had, and I'm sure this is one you've studied more than others, is that it's not really the way internet works. People tend to go to a place where they know they will be with peers, they tend to go with their tribe. And often, it's often to the point where the best conversation on the internet

tend to happen in a very political platform, know, like fans of Donald Trump's in the platform would actually talk to each other in a way that feel like a healthy debate. But if you try to put everyone in one place, and as I suspect you were trying to do, yeah, there seem to be less appetite for debate. Can you maybe walk us through kind of what was the optimism behind the platform? Like what was the, I guess the business model, if you will, like the reason why you expected people to come?

debate if they were not already part of the same tribe, if that makes sense.

So what I what I hoped we might be able to do would be to trick people into coming and discovering new ideas. What I my

interesting.

the basic model was as follows, which is I tried to build an encyclopedia of all opinions, collectively built. And we figured Wikipedia gets 18, at the time, at least something like 18 billion uniques a month,

Mm.

because people are typing stuff in, et cetera. I did I did deep sort of Google search traffic analysis

Mm-hmm.

and worked out that people ask Google.

really interesting questions about everything all the time. And w my calculations were if we get the if we get to be the number one when people ask, you know, does God exist or is the monarchy a good idea? Or, you know, what do we think of euthanasia? Or or or from the big and obvious kind of high school debate questions all the way down to the latest stuff to, you know, and br breaking politics and the rest, breaking news. I figured we'd be able to drive enough traffic there for to then surprise them when they when they came. So if we could become if we could rank top on

Google search for complicated questions of this kind, which we were able to do. then we would by design, and this was my co-founder was a designer, we would be

Hmm

able to kind of move them through alternative perspectives, surprise them, challenge them. And we did on some level.

I'll see.

So there's two platforms. One is called encyclopedia of opinion.org, and the other one is called parlia.com. And

Mm-hmm.

encyclopedia of opinion.org really does look like, you know.

A Didol fantasy for the 21st century. And Palia was and my kind of like my 17th iteration, which was to say, okay, fine, just tell me who you are, tell me the kind

Mm.

of person that you are, vote on all these statements,

Mm-hmm.

and then I will show you how you correspond to other people. Because here the idea was okay, fine, don't argue with me, don't sit down, don't hammer out complicated theses and antitheses and hypotheses and whatever it is. Just tell me what you feel.

And I'll show you stuff which surprises you. So Bruno, how old are you?

I don't know if I can say that publicly, 42. Yes, I'm 42. Yes.

Yeah.

Are you a are you a millennial? You're a millennial. Okay, so you're an old millennial.

so what I would say, so you would sign up to parliah.com and you'd say, I'm a male, I'm an I'm an old millennial, I own,

Whatever. Yeah.

I make this amount of money, I am, I have, I'm married, not married, I have these kids, my politics leans

Mm-hmm.

left or right, etc. etc. And I would show you the idea, was that I would show you people who you were not. You know,

Yes.

Gen Act, a Gen Z woman.

Who lived in a different country who agreed with you? I'd show, hey, it turns out that actually Gen Z women on this topic, you know, 50% of them agree with you on this thing. I would kind of try and break your identity silo. one by showing you people who you were not who agreed with you, and showing you the people who you were like who disagreed with you. So that was that kind of attempt. But but yeah, your initial question is is also a backhanded insult. And I'm I'm starting to learn your tricks here, which is you know, what was the optimism? What was the optimism for this?

platform and it turns out was very misguided.

I

think we're getting close to a good transition towards the topic we were trying to focus on in podcast, which is the question of taste. So you've talked about how asking this question could help predict someone's opinion. And I guess one of your focus was political opinions, right? Like whether this president is good or bad or whether this policy might be preferable to another. Philippe and I have been investigating a notion which is

maybe similar to the concept of opinion, but have some slight differences, which is the question of taste. What we mean by taste is really this, I would say maybe a mix between skill and preferences, right? Like when we use the word taste to talk about someone who has a great taste in music, right? There's an aspect of, they spend a lot of time deciding what they like. They have like a deep appreciation of what they hear. They can maybe,

recommend it to others, maybe they can write music. There's a set of skills and capabilities that goes around it. Maybe I'd be curious to maybe start hearing your thoughts on, I guess, what are the type of things that predispose opinions, like political opinions that you've researched, and how might the same mechanism influence tastes, say, taste in food, in music, and so on?

That's fascinating, yes. I realize that your your rendering of the word taste is a French rendering of the word taste, right? So you can say of somebody, Il a du goû, elle a du goût, they have some they have taste.

Mm-hmm. Mm-hmm. Yeah.

yeah, exactly. In front in English you sort of have to say they have good taste.

It's right. There's a nuance. Maybe

Yeah, yeah.

there's another dimension that we don't need to spend too much time on it. right now, we're talking about taste also because it's trending in the Silicon Valley. And in the AI industry right now, everyone's talking about taste. They tend to talk about it really as almost objective capability or it's like a non-subjective skill. They talk about taste as someone who can predict if a

specific UI basically is going to sell a lot of subscription to a product. Or they talk about taste as can you sell clothes? This is another version of taste. I guess we're closer to a more maybe European definition of taste, if you see what I mean. So with its own mix of

I hear you.

skill and subjectivity. Philippe can also interject if you.

Tudy, you go first, because I have a question in

All right.

a different direction.

Go.

Okay, gotcha. Yeah, yeah, no. So so

I mean that there's a massive amount to unpack here, and I'm really interested in your view on this stuff. But yeah, just from some interesting parallels, let's start with the literal one, which is that one of the weirdest pieces of of of data that I discovered is exactly what I mentioned earlier, which is that a higher correlation, a higher number of taste buds on your tongue. So what's called greater papillary density on your tongue

Mm-hmm.

predicts conservative politics. So

No.

Which is which is kinda kinda kinda crazy. I can explain where that comes from, which is it's part of the it's part of a what's called a conservative neurological complex, which is which is tied up i i in great in in in all sorts of ways tied up with disgust, tied up with with being hyper attendant to threats and risks, etcetera,

Mm-hmm.

etc., which predicts a kind of more conservative approach, insofar as you have a neurology which is more aware of threats,

Mm-hmm.

whether it's on your

d you know, i whether it's foreigners cr crossing the border or whether it's threats of rotten f stuff on your tongue,

Mm.

which predicts a particular kind of political leaning. But so so yeah, what do I what what what can what can my discoveries in the realm of opinion tell us about other things in taste? So so one,

one.

genetics, absolutely.

Mm.

Genetics has an influence on the kind of things that we like. And I I found some crazy papers suggesting that you could actually track interests in contemporary art back to certain kind of gene clusters. It's almost certainly not true because the the but I just love the idea that people are running in that direction, right?

Yeah.

and it seems to me perfectly plausible that at some point we really do discover gene clusters which correlate with greater interest in, say, abstract art rather than figurative. I don't think that would be difficult to.

to to imagine.

Mm-hmm.

So genetics play a role. genetics play a role in musicality, for example. So when you're talking to about taste as in skill,

Mm-hmm.

we know that musicality is very deeply genetic. So you take and

Yeah.

and the reason you know that is twin studies. So it turns out that identical twins are more alike in their musicality. They're either

Mm-hmm.

both musical or both unmusical than fraternal twins. So you

The skill element clearly is tied up to a whole bunch of of of of genetic stuff. And then when it comes to n t tendencies or pref preferences, then yeah, the things that this is n none of these things are surprising, but turns out that the context in which you were brought up, the people that you hang out with, the kind of contagion effect that is all social interaction really, plays out across our opinions

So thank you.

just as it will across our across our tastes.

There are other interesting things about tastes, which I know I'm sure you guys have looked at, which is that tastes are there's something about childhood and adolescent, which adolescence, which has a very serious impact on taste, doesn't it? We we weirdly much less so with my world, which is opinions and thoughts, possibly because kids and adolescents are not thinking that hard about politics or ethics or whatever it might

Maybe, yeah. Yeah, that's right. Yeah.

be. But there is something about we've all had it, you know, the experience of childhood deeply imprints preferences in a way that

That's right.

Way that

we kind of give up. I mean, I know this. I moved to Italy a few years ago, and my wife is Italian, so I've been speaking Italian for 25 years now.

Mm-hmm.

But I'm half French and half English. And I speak better Italian now, I think, or more fluent Italian than I speak French. But for me, reading French is so much richer as an experience because every word has texture, it has smells. Croissant for me, I'm

Yeah.

immediately I'm four years old and I can smell the butter and it's trickling down my hands.

Focaccia, which is just as good, still is still just a word. Doesn't not childhood for me. So yeah. anyway, there you what? Yeah.

It's not the priorche. Yeah. It's not the prior from if from

Yeah. Yeah.

the memory. Can I ask you a question on this? this is more like an exploration in a certain direction, take it where you want. if I think about the individual opinion or also taste, there is a in some sense there there are two contrasts, right? One is the assumption that I'm I come basically from a free position to my current opinion, which somehow also implicitly suggests I've chosen it. I've op I'm kind of optimally arrived, and the economists

There is a, there a context, but I wanted to say something that I come to appreciate my current state. Somehow all the businesses that I've chosen, I've kind ultimately arrived at,

would say.

but you are just coming from a different place. Now, if you paradoxically, it's actually the same as what you described to think more, like more about where the health of my patients actually comes from.

Versus you are contrasting this is this is coming from deeper. Now I think paradoxically, it's actually an invitation, what you described, to think more and introspect more about where the hell do my tastes actually come from or my opinions.

And I think there is a bit of a parallel also to this kind of discussion on free will and also Eastern, you know, meditation traditions where if you are suddenly realizing, my God, there is an ape in my head,

And I think there is still a parallel between the current and the current and also the search for meditation. If you are talking to me realizing, my god, there is a cave in my head.

you know, you might be thinking differently about your decision making. What's your

Your your take on this introspective side of the the connection you make?

That's such a beautiful question. And thank you for framing it that way. And I and I haven't actually framed it that way to myself in a very long time. But but yes, of course, that's correct. I mean, so many things are popping for me now. One,

think the public is now one.

my I have two kids both have been through adolescence at this point. And one of the things which was so striking for me was the two of them, possibly because a culture of therapy is in the air, or TikTok, or or they're just more mature than I was at 17.

was that both of them were able to see that consciousness was some well they had a they a they had a kind of a a a what am I trying to say? an unstuck view of consciousness. They could see that they could see that their minds were at work in a way that wasn't them. And

Mm-hmm.

I remember when I when I heard them when we started talking about this, it like, my God, it was 40 before it this kind of crossed my mind. So so I think it's and I think it was very helpful for both of them

Going through that weird and rocky period, but I I fundamentally deeply agree with you, which is that if there is, if what you're starting with is this quite uncomfortable notion that a great deal of your positions, whether it's you know your preference for strawberry over chocolate ice cream or your preference for totalitarian right-wing totalitarianism over liberal democracy, if if you start with the understanding that a lot of that

Is outside of your rational control. You're you're f you're forced to think more more, I think more deeply about Well you're forced to think more technically about two things. One, whether those opinions are really yours, two, whether they are right. They it kind of it asks the question in a different way. If you've come to your opinions from a random perspective, you hold them hopefully slightly less hard. You hold them in a slightly less absolute way. And there I hope lies

way, it's like you're in a round space and you're like, oh, I'm calling it a round space. And there I know my

goal is to say, well, it's a round But it also does a different thing. It's a beautiful thing. I if I've ever that. I usually do most of the

all sorts of space for intellectual humility, but also then exploration. What it also does, I think, is it it it should force you to think about your opposite side in a slightly different way. And here's the thing, the piece which is most interesting

Yes.

to me about the work that I've done on this blasted book, which is that

You know, when when you're arguing with somebody with whom you disagree on a subject which you know something about

When you're arguing with somebody, you're talking to yourself.

And they come to a different conclusion than you. It's very difficult to take them at their word. It's very difficult to believe that they're

Yeah.

sincere. You think, hang on, I've done the work, I've done the reading, I've done I've I've I've really thought about this, and they've come to the opposite conclusion. They must be lying, or they must be stupid, or they must have not

Hmm.

done the work, or they must be somehow morally corrupt, or or they must be bought, you know? And this key realization that there are

That there are profound differences between us. That actually that you know, your right-wing authoritarianism and my liberal pluralism or the other way around is thought, but it's also predisposed. It stops you thinking of the other person, the other side as in as corrupt, and it stops, it starts you thinking thinking of them as different. And that allows you to trust the sincerity of their arguments, to trust that they come to the arguments sincerely. And that may be the most important thing for a conversation.

And that allows the truck to up. The truck will come to the ice rink and that may be the most important

You have to trust that the other person is sincere in the way that they approach you. You can then disagree, but if you're assuming at starting point that they're lying to you, then there's no point in talking. So I think maybe that's the thing which is most interesting to me.

part. I that's

It's also an invitation to exploration, right? Because if you have the empathy realizing, then you are also interested and curious about learning why has that person that taste. And not like they have that or they have this opinion, full stop. Let's go home, right? following up on this, so maybe we kind of take the next step to empathy, also

into the expector. You to realize then you are all these years about learning. Not like they have that or the after the opinion. Let's go home. Following up on this, so maybe we kind of need to take an example.

to this kind of color collaboration. I think if you in the political realm, but I think it's also true in in in other realms, also related to taste.

I think if you in a public political realm, but I think it's possible in other realms also really takes

some very compromised to the best. It's almost like if you're compromised to giving up on your position. Can you maybe give us a bit of a framework of compromise? What is a compromise?

The word compromise has a bad rap these days. It's almost like if you compromise you're giving up on your position. Can you maybe give us a defense given your framework of compromise and what is a good compromise?

I am the most unfashionable thing on the internet, which is I'm a centrist dad. And and therefore that it's like it's my it's my shtick compromise. But maybe before

I am the most popular person on the internet. And I'm the most popular person on the I'm And

Let me perhaps start with a whole bunch of things that let me come in to come at this from a different way, Philip. Sorry. thus my the s my standard argument is that we are not able to break out of our own consciousness unless we encounter the consciousnesses of others. so that's the kind of the psychological piece. And that psychological piece I think has got merit in its own right, in so far as if we are curious, this is the thing we have to do. We owe it to ourselves, regardless of society, regardless of progress, regardless of anything else.

It's a pleasure. We should encounter the ideas of others because otherwise we're stuck in who we are. So there, I think, is a that's a I think that's an argument from that's a personal argument. It's a psychological argument, it's an argument for intellectual bliss or whatever you want to call it. but at a social level, I think there's, you know, the scientific method sort of sort of demonstrates the point that actually we we we do think better in conflict with others as well. And I've got a whole chapter in my book about which which which hopes hopefully shows.

well, I've got a poll countered by both of them. It's so postedly shown

that the back idea comes from a lot people. So there's an argument for conflict of interest. I'm not sure that that's happening. I'm not sure that's But hopefully what happens

that that the best ideas come from the conflict of minds. and so there's a there's a there's an argument for conflict itself. I'm not sure that that necessarily means compromise, but hopefully what happens

is okay and so because I think compromise in people's head is like, would you would you compromise over slavery? I mean do you compromise over the Holocaust like, okay, I want to kill six million Jews. Well how about just three? yeah you you

Wow, yeah

So and I think so there is a there is a there is a very strong case for both being wary of

where

compromise for compromise' sake. But

This is exactly

the sorry, just to maybe interrupt, this is exactly the kind of bad rap that I mean. I mean it actually the positive that we are creating something new out of a compromise. Right? That it gets better because we are doing it together and we

I mean, it's the positive. We are creating something to out of compromise. That gets better because we're doing it together.

come up with solutions that no one had before.

Exactly. But those don't that feels like the opposite of compromise. Because that feels like that feels it feels like

In g when great conversations you have the sense of two people building simultaneously. They come with different bricks. But there is this sense, I mean, if you're if there's a sincerity of argument, then I don't have to give up my point to you. I'm not handing over kind of agency or cognitive agency to you. if the if the pairing is sort of matched and if we if we're both bringing things to the table, which is what all great conversations feel like, then you're kind of building a castle in the air together.

Yeah. Yeah.

and and that feels that feels like the s sum of it's you're building something which is more than the sum of its parts. it does require a kind of intellectual honesty, and intellectual honesty is sometimes agreeing that you're wrong or that the data doesn't back you up, or that somebody has a more interesting or more ethical or more wide or more experienced or or more rational take than than you do. That feels like the give and take of argument. It doesn't feel like

compromise. That feels like a admitting when you're wrong. So I feel like maybe compromise is exactly what we shouldn't be doing. Maybe compromise is

Mm-hmm

has a bad rap for a good reason, because actually when we're really working well together, it's not compromise, you're actually building something new.

Yeah.

Maybe I'm using also the the the the word in the wrong way. What

I'm essentially referring to is this idea that two political parties they crash and they can't they can't pr propose a solution basically that is seen as constructive. I think that might be the the better way to frame it. But I see I see where your point, maybe the w the word is the wrong wrong direction. if you Yeah. Maybe if yeah, so

Maybe, yeah.

On politics, Philip. Sorry. Yeah, yeah. No, I was just gonna I yeah.

you go on, Turi. Sorry.

Yeah,

I would like to say maybe to finish a bit on the political side, what in your mind is the goal of depolarization? What is the end state of, say, the left and the right in the US talking to each other better? Is this finding policies that two thirds of people agree on? That's often a framing that centuries could possibly focus on. are you actually satisfied with the idea of

l'alternance as they call it in France. Sometimes the left is in power, the right is in power, it's just more respect between the two sides. What is your goal of depolarization?

Well, I mean so so so personally and politically, but so personally, as a centrist ad, I I tend to be for middle policies, and they're very unfashionable. So it's you know, part one of your answer.

Yeah. Yeah.

but I'm also very happy for those not to be the case. The thing which is more important, I think much, much more important, is that you end up with this alternance that you're describing,

Yeah. Yeah.

but in a culture of of of mutual r respect.

And and the conviction and the remembrance, the memory that the other person is fundamentally on the same side as you, or they they come to the same question of how to improve the lot of the country. They come to that question with the same sincerity that you do, even if they come for it at it from a very different perspective. Because you're familiar with this notion of the of the infinite game, which is

Which which is the games you continually want to keep on playing and democracy is this infinite game. You never want anybody to win democracy. If anybody wins democracy, democracy has lost. And so and so I think it's this business of making sure that we're shoring up. And I think probably the core message of my ridiculous book is that we need to be doing a little bit better job of protecting the spaces which which guarantee the infinite game. Because I think the infinite game, the conversation between up opp opposing alternate sides.

is evolutionarily how we were designed for and how we how we adapted to reason. And it's better for society as well as being better for ourselves at a personal level.

At some point,

If I quit so sorry.

suggest to switch also to AI topics, but go

for it.

Yeah, I I'm

actually about to tr to to transition into it

Perfect.

through the back door. so the the framework you set the idea that parts of how we think, how we form taste is coming from a biological then also cultural perspective, which is I think arguably moving very slow relative to other things. now we have AI, we have I think many other trends that are still at least seemingly accelerating our time.

the evolution of what happens. How do you see this? How do you see change, positive change happening in the face of well, maybe we you know, our we as humans are just moving slower and as cultures we are also moving slower, but the environment is basically exploding very fast.

see that, I really think that our human rights are lower and our culture is also put lower. The environment is also put

It's question I would ask you. You both spend a lot of time thinking about, I mean, engaging playing with AI in a way that I I do as a consumer, but

Damit.

not as a

not as a not as a builder in the way that you guys do. Can I can I just drop a a a little positive framing at the beginning of this, which comes not from not from my

I'm professional. a professional. I might drop a little positive training at the beginning of it. I'm not a I'm not

a I'm professional. I'm a a a not not I'm I'm I'm I'm not I, yes, know, keep it down, I'll be on the rocks, I'll break my heart, but, but, but, but,

work on the book, but from work as a media entrepreneur and investor for the last twenty years or so, which is that I and here, you know, shoot me down, tell I'm wrong. You'll break my heart, but but do please, which is that my

my senses that of the of the of the media technologies that we've seen over the course of the last twenty years or so, you know, ending with the last one, which is this explosion of social media, which has been monstrously detrimental to, I think, not just our, you know, our adolescent's mental health, but also the health of our politics and the

type of conversations that we have.

I wonder whether actually the rise of AI isn't the first technology which actually privileges accuracy of all the things that we've seen in the last twenty years. Insofar as social media privileges outrage, privileges, you know, knee-jerk, visceral, violent reactions of one form or another. You either want people to agree with you furiously

Another surprise of AI is the first technology which acts as a It's first, especially with Privilege graphics, Privilege is a feature that controls lot of reactions to a particular type of

or disagree with you furiously, or you want to show them people that they disagree with and you guys all agree about it.

just from a content perspective and I work in news

Thank

predominantly, it feels as if the business models of AI are really are really the predicated on on accuracy, so far as the fight between Claude and Chat G PT has been

I think it's more of an AI aspect. It's better to it on my app, I to make sure it's more of an AI aspect.

around not hallucinating. I I wonder whether I mean is this a tremendously s simplistic view of this, or can I can I hold a tiny bit of optimism?

So

there's a there's definitely truth to that in the sense that if charge pt start misquoting wikipedia or start making up fake facts, know, fake numbers, it's bad for the brand, right? It's not it's not when it comes to factual matters. You can trust that all the companies will optimize for.

What they call the very the verifiable domains right when it comes to mathematics when it comes to coding when it goes to anti-clobidic knowledge They will have all in the incentives to answer truthfully But in my mind, is the intelligence question, Like this is AI as a source of Answering technical questions right in

Mm-hmm.

as a replacement for Google as a replacement for Wikipedia. There's certainly value there

When it comes to most objective domains, this is where obviously things can go wrong, right? There's one fear, for instance, which is AI is going to be even more sycophantic than social media could be, right? So if you use someone's social media really fall into a rabbit hole, whether it's, know, looks maxing, whatever terrible...

Mm.

There's so many mental health issues that can be reinforced, it's just you like...

joining

this tribe of people with the same problem as you or all these echo chambers, all these problems we're very familiar with, there's an argument to say that AI could potentially be worse because it's going to be even better at guessing what you want to hear. It's going to be even more capable to mimic your words, mimic your style so that to build this rapport, it can be basically a very manipulative force. And on that front, the business model is not great.

These companies have raised so many dollars that they will have to return at some point. The battle for building an addictive experience, building rapport with the user is going to be fierce, just like social media optimized for screen time. There's definitely

Hmm.

theoretical reasons for concern. I know you talked to lot of people for your research on your book.

If I remember correctly, you've talked to some people who are quite cautious of social media, for instance, having all the negative impact that you've already mentioned. But I remember you also had a guest who actually said, maybe we sometimes overestimate how much this technology matters because if we blame extreme conservatism on Fox News, maybe we...

maybe this is more a mirror. Some people would say that social media, Fox News, they're just telling people things they already believe to some extent. So maybe I'd be curious to hear your thoughts on, I don't know if you have a stronger opinion on this yet, but how would you go about deciding whether or studying this topic or studying whether AI is being just a simple mirror and maybe it's fairly innocent in the same way that...

it's entertainment sometimes, it's just flattering, but in a way that is not too problematic, or whether AI is actually this sort of force that will modify, that will radicalize us.

Amazing. Okay, yes. So again, this is pure hunch, but maybe, you know, maybe that's all we've got, guys. Maybe all we have is hunch. I

Yeah, at this stage I think it is, yeah.

mean, not just about AI, but also about everything. but so just at a at a hunch perspective, and there's a there's a bunch of data that backs me up on some of it. I've always thought that the blame social media for all our political ills was a disastrous and you know it was just a mistake and wrong.

And

Mm-hmm.

You know, if you look at Fox News that you just you just cited, Rupert Murdoch did not invent a media company for an audience that wasn't there. Rupert Murdoch figured out that there was a massive gap in the media discourse and filled it, and filled it spectacularly well. And there's an argument that actually what we've been seeing over the course of the last X amount of time has been people moving into a space, not radicalizing it necessarily, but allowing it to express itself. And my

That's that's my hunch. That's so I would say eighty percent of my gut tells me all this stuff is already there, it just needs to be needs to be surfaced, and then then some media companies surface it, some technologies surface it, some cultural figures help surface these things. and so I think that's why I probably put the bulk of the weight. There is no question that that new technologies like new discourses, like new people.

can also pull things in different directions. And that I think that's but so you you asked me for my hunch on my view on this, and it's

Mm.

not a view, but it's a hunch. So that's where I would that's where I would put myself. And then otherwise, yes, what's it like having your permanent doppelganger? Your your doppelganger who's constantly reinforcing permanent in your pocket. I have I mean I know this in the most rarefied of all cases, but one of my oldest friends is an English professor and he wanders around with chat GPT.

discussing with himself the merits of mostly architecture. He's an English prof, but he talks about architecture like he's talking about 19th century novels. And ChatGPT has now figured out how to talk back to him in those in in in that way. He is a pig in shit, as the expression goes. He's blissfully

Yeah.

happy, he's finally found the perfect counterpart to talk to him about, to talk with about architecture. That doesn't seem to me terribly nefarious, although he can get boring.

But but yeah I suppose you you guys are thinking of what what it does in terms of politics, but what it does in also

Mm-hmm.

in terms of notions of self, what it can do to mental health and the rest of it. W d is there data? It I mean, do you have a is

It's holy

there something that we can look at which you go okay, what this is just like X in the fourteenth century or twenty minutes ago with do we have parallels?

It's early, unfortunately. Obviously, the big labs are starting to release some data about how much they think AI was used in some ways or another way. But some of it is very difficult to interpret. For instance, we know Gen Z is now using charge-gpt to make a pretty big decision about personal life. They're using it to decide, I date this person? Should I break up with this person? They used to go on Reddit. Now they just ask the same question to charge-gpt. So maybe it's not such a big difference.

The problem, we've had similar issues with, you know, after the last US election, there was a, well, I say the previous one, there was a lot of question about how, yeah.

Hang on, you mean you mean you mean twenty twenty four? You d you

mean twenty twenty four, the one that Trump won, rather than twenty twenty, the one that Trump won. Yeah.

Yes. Yes.

There's a lot of questions about how much Facebook might have influenced people's political opinion. there's a large quantity of researchers who try to scientifically resolve these questions. How much does spending time on Facebook influence your opinion and so on? And it's consistently the results are non-conclusive. There's so many of these studies end up saying, we don't know, because

It's never a purely controlled experiment. It's never properly A-B tested and so on. And there's just so many other factors, so many correlation and so on. So I think, unfortunately, we're going to be in the gray area for a while. It's going to take a long time before any data really tells us, I think, really the impact of these tools. Is that your read as well, Philippe, maybe?

I think as of now, yes. I think that that's I mean, and in general these kind of empirical studies are obviously extremely hard and there are also lots of interests behind it. So it's it's it's very hard to get basically to an objective point of view from the outside. maybe one thing that I would like to explore a bit l this is tying a bit back to what I said before. So the realization that maybe not everything is optimal, what as a as opinion you hold as an individual.

Then you go on a quest quest of understanding where does it actually come from an introspection of what does it actually mean. I would argue that we have technologies, some of them are called books, to help us actually explore this, basically, right? We are, you know, ke putting ourselves in the shoes of a character and understand and start thinking about this process as well. Could we maybe explore a bit, from your opinion, the the role of AI that or what could be the role of AI in a similar way and but different?

I would argue that we have technology of some of the historical books. How about that we explore that we are putting ourselves into the character and that something about the process of... Could we maybe explore a bit in your opinion the role of AI or what could be the role of AI in a similar way but a

different way?

How would you so so so I just wanna make sure I understand the thrust of this question, which is we go to books, the minds and imagination of others, to try and find ourselves reflected, to feel less alone in the universe.

Sorry, I'm not yet. But I know you're cool. The mind might take a while. And I

And

maybe understand ourselves and also understand other perspectives, right? To understand motivations. This is this opening of not everything is fixed, but there is actually a reason. They might even change, they might come to a point, etcetera.

really understand how it feels. And I also understand the fact that I do understand what you're waiting for. This opening of not everything is big, but there is actually reasons they might change if I come to this point.

Yeah, yeah, yeah.

Yeah, yeah, I I could couldn't couldn't agree more. I think one of the one of the sadnesses is that with this I know you guys I'm sure have seen this, but there's been a c massive drop off in people studying the humanities at university and to yeah, and I have it turns out a couple of friends who are English professors and both of them

Mm-hmm

suggest that actually they think what's happened that the number of people who stopped studying literature at university, they all gone off to study psychology.

Mm-hmm.

And

that's a really interesting move because what it does is it says rather than okay, it means that you're you're clinicizing the experience of the world. you're clinicizing the feelings that you have on the hope

Mm-hmm.

that somehow you'll be able to diag diagnose and fix what you've got. Whereas maybe what you've got is, and Philip here, I'm gonna need your help, one of those very long compound German words, which is, you know, anxiety about being in oneself or not really under whether it's

sign and does sign or vel I don't know what all these words are but but but you know, the the the experience I kind of felt like Kafka was helpful to me, age fifteen. And I think probably way more helpful than trying to work out whether I had, you know, some version of and I social anxiety what you know what I mean? So so I think that there is an enormous

Yes, I agree.

amount lost. And I have a parallel here with a great friend of mine who's who's she's a historian of science but spends a lot of time with neuroscientists. And she is

Always very impressed by what neuroimaging can do and tell us about the ways in which humans are and the brain works and everything else. But she's also like, dude, just read poetry. I mean, we were way more accurate about the

I'm for it to stop.

human condition in the twelfth century when the troubadours are trying to explain what love or yearning or pain or

about.

sadness was than we are now who go, It turns out that your hypothalamus goes flickety big when So, we've

Yes.

got these brilliant technologies which are telling us really,

Hmm.

really, really obvious things that, you know,

That we've known for a thousand years. So I think we're in an interesting moment there. But your question, I suppose, is is there anything we what is qualitatively different about a conversation that you're having with AI than a conversation that you're having with an other, right? Whether it's Kafka

Mm-hmm.

or Maupassant or Goethe or Shakespeare. so question back to you. Do you think that the models now are able to render that otherness

in a way that is going to be valu val valuable to to to to the interlocutor, or to to Bruno's point, the kind of the the lean, the approach here, is always going to be so sycophantic if we want to be brutal about it, but kind of mirrory that we'll never be able to have that alterite.

That, yeah,

I think there is

yeah. I think there is a step. Yeah,

go first.

sorry, I I because I I think I'm I want to make an intermediate step. I think there is there are several steps in the the way you can use it. And I think one thing we already discussed also with other guests is that if you're using it as a sidecar to the things you do anyways, it can be extremely useful. Like if you're reading again, Madame Bovary has an example, which is, you know, for a a white male, maybe easy not not so easy to understand in the context and so on, but if you get it, it's so much a it's such a deeper experience on understanding w

I'm standing.

where does it come from, right?

And I think there could be a similar actually also for collaboration, that you're using it as a as a sidekoff for creative interactions. And we actually had a guest on kind of developing a tool in this direction. Moving on to as a full partner now. Honestly, I find it just annoying and boring at some point. That's my experience.

I think it looks very very nice to taste again, right? So one of the group of people

Yeah.

who is the most disappointed by AI today is writers, right? who have fallen into this trap of saying, I'm going to use AI to write my blog post tonight. And they generate something which in the moment looks maybe competent, you know, like the word look fine. And then the next day they look back at what they published the day before and they're very disappointed. They realize that's very bland, very sterile. It's really lacking something, right? And maybe there's an...

simplistic way to think of it, saying, it's just shallow or maybe it's just too polished or something like that. I think it's really some sort of emptiness, Like there's something missing in what comes out of the AI. And typically this can be traced back to lack of a personality, lack of taste, lack of something kind of unique that would make the AI have something to say, right? So that maybe that's a more kind of a romantic, kind of artistic view, right?

When a musician is communicating something about themselves in the music, you can feel it. But there's something often I find in the interaction with AI where if it's a technical domain, I'm asking you to solve the math equation, it's great. If I'm trying to use it in the same way that I would use a person as a source of inspiration and learning, it's extremely disappointing. There's something lacking. I'm curious first, is this something...

That resonates with you roughly, this difference between the technical and the more tasteful, I guess.

Yeah. Yeah, yeah, completely.

Completely completely in every way and and

I feel like we're all getting better at re at at hearing AI. I mean, I don't I don't mark academic papers and that's a very particular kind of language, but my God, I can tell when, you know, a board report has been sent to me from one of my CEO companies or whatever it is, and I'm like, this has got the that thrusting, optimistic short sentence rubbish

Yeah.

of of kind of a kinda kind of a kind of AI. We we can we can hear it, but I I've also got you know I've got a n a novelist friend who's just who's has who's spent a lot of time trying to work out how to use it and and used it with

real value as a sidecar to himself as he put

Mm-hmm.

together elements of what's ended up becoming what I think is a very beautiful little autobiography, autobiographical essay. And then quite quickly dropped it when he was writing a novel. He's also published an enormous amount. So in terms of learning, the AI was able to use an enormous corpus of his material.

Mm.

But even there, I think found it found it frustrating and has now and has now has now dropped it.

So yes, technical, super helpful. On the other, is there some nineteenth century romantic idea that beauty is a is an external absolute, an abstract thing which exists and and we aspire to it by virtue of of being human? I I don't even think you necessarily need to go there. I my my hunch with with

My hunch here is that we have this, the only advantage we have really over AI is that we die. And it's which is another way of saying we have all the things which are wrapped up with dying, a kind

We have to be very honest with you. have to be very honest. We will be honest. This is another way of saying we have to be But wrap up the talk.

of a sense of urgency, you know, desire, desire played out in that context, a sense of narrative, a sense of a sense

Mm-hmm.

of of failure and success, etc., which infuse everything that we do and and and and is therefore the thing that we

we most connect within other people. And here I think is the last point that I'd like to make, which is I think we've all been dancing around and you've gested to it very clearly, which is which is the is there a qualitative difference in the pleasure we get from interaction with the other that we don't get when we're interacting with a version of ourselves. I've had really interesting conversations with OpenAI, in fact, with the ChatGPT about friendship. I to teasing it out. Like hey La Boessie, Montaigne,

rend d versions of friendship, male friendship, the difficulties of framing male friendship, the total absence of any literature on male friendship. There are novels about women, but there are very few novels about male friendship because people are so nervous, I think, of bleeding into homosexuality, etc. There are there are buddy movies. There's

Mm-hmm.

the three musketeers, but they're they're groups. Then it's never it's never one-on-one, always too threatening.

Mm-hmm.

So I've had an interesting conversation there, but and that that was but I wonder as I think back on it now,

The pleasure that I got from those conversations was a sort of an informational one and an and I was super impressed by the what what ChatGPT was able to do. But is it the same pleasure of discovering the other? I mean, I sound like I sound like a postcolonial theorist, I sound I sound like Edward Said, but psychologically, do you guys feel that? Do you think there's something there? Am I being wishy-washy?

I've always preferred going to a lecture and listening to a professor in a domain rather than reading a book. I always had a bit of that preference from say learning from

Yeah.

someone maybe charismatic, if that makes sense, know, or

And I think also just in terms of attention, I find it easier to stay engaged and curious if I'm hearing from someone who disagrees a little bit with some of my views. you're not keeping me entertained as well, right? The rough edges can keep me entertained. So yeah, so I find AI boring, basically. Like a long learning

Yeah.

session with an AI gets boring compared to the more human experience. Is that also your experience, Philippe?

Same. I think for me it's just basically a pool a a tool that really focused on the information, it makes it is perfect and you can do things that are just unimaginable, also from a scientific perspective. But for a conversation that is an exploration of yourself, I feel like it's it's just lack lacking depth and it's getting annoying and boring.

All right, so how do we end up on a non-boring note? Because we've taken quite a lot of your time, sorry. Maybe to look back a little bit to a topic you've covered a lot in your book. Would you say the concept of taste, as we've touched on a little bit, is this more a left-wing concept or a right-wing concept?

Ooh, hang on. So let's talk about taste again. So it's the idea of taste like goo, like having good taste. Like good taste.

Yeah. Yeah.

Yeah. And I'm going to even explain why I'm asking this. But basically, when reading your book, I noticed you highlighted how the more liberal left-wing spirits might maybe care a bit more about the creative side of things, right? Like more trade openness and so on. Right. But the concept of taste often includes this idea of discernment and judgment almost, which maybe is more of a conservative kind of vibe.

a little bit. So would you in a more kind of centrist conclusion, if we can, do we need a bit of both to basically reach taste in a way? A very leading

yeah, so

question,

yeah, totally totally. So just we just do the see if we do the tour d'horizon of of of all that. Yes. all the data suggests that innovation and innovation across genres, I specifically music or art or et cetera, et cetera, tends to come tends to come from the left. There is you know, if you wanna render it neurologically,

Mm-hmm.

where that kind of seems to exist is in the what what I've looked at and certainly is in the

Anterior singular cortex, which is interested in ambiguity and getting stuck into complex, often contradictory ideas, etc. So there is a natural instinct towards innovation, which is the obviously a left-wing leaning. There is on the other side, the idea that there should be such a thing as taste is profoundly hierarchical and

Yeah.

and and d and deterministic. I mean, it's like it's so so th that is considerably more right leaning.

Yeah.

as a framing. and I and and w what we know about taste historically is that it swings all over the shop. And I think that the way that it swings all over the shop is enormously based on what where we happen to be sitting in a cultural or economic cycle wherever we are.

Hmm.

but still today I think my wife has she's an architect I think she has spectacular taste and and but she has there we have we have friends who sell extremely well in the art market which both

Okay.

She and I think are are monstrous. And therefore this so I think the idea that there is a a single taste is is is is is obviously rubbish. And what's interesting about taste is precisely what it tells you about the cultural and economic moment that you happen to find yourself in. So I that's what that's how I read it. But do you, Bruno, think that there is such a thing as good taste?

Only in the way that you define, right? So there's obviously you can, know, everything feels subjective until you narrow enough the group to the few people who agree, right? In which case you can start calling it objective. So if you, you know, it's

In music, that's this question. If you break things by genre, perhaps, you get closer to a notion of good taste in each genre, possibly. So this is where I feel there's useful about the word taste, basically, which is that there's a reason why the word is in our vocabulary. And there's also a reason why people talk a lot about it today. The premise of this podcast is we noticed a huge uptick in people starting to talk about...

Taste becoming the new thing that matters because now that AI is taking care of intelligence, the only thing left in a way is taste. So

that's exactly. Yeah, yeah.

that's where I see the stakes being high in understanding the mechanism behind it. Very similar

Mm, fascinating.

to you because I'm also in a similar centrist dad category.

Ha ha.

I have a strong hunch that we need some ideas from the left, some ideas from the right.

We're to work together in whatever system is going to understand or augment or taste in the future.

There's this there's a this great professor at Brown called Rose McDermott who simplifies it really radically. She says, you know, societies don't just need the left and the right because sometimes the left is has the most adaptive response to the circumstances

Hmm.

in which we find ourselves, which you might very roughly describe as opportunity circumstances. So when things are going well, we tend to become we tend to lean more left. When things are going badly, we tend to lean more right. we don't just need them occasionally for when.

the circumstances change. We need them both all the time. You win and as as she frames it, she like you

Exactly. Yeah.

need the right to defend and protect and you need the left to to to go to to explore and and to care. So

I like it, I like it a lot. Maybe this is a point to finish on. Thank you so much for your time, Turi It has been an absolute pleasure.

Loved c talking

Thank you very much.

to the both of you. Really, really interesting conversation. Thanks so much for having me on.

Thanks.

Thanks for watching. We've got more episodes in the oven, so please like and subscribe to get notified when they're out, and please check taste-bench.com to read about the taste benchmarks and the creative engines we're building towards.