The Right to Be Unpredictable: AI, Algorithms & the Cost of Conformity
Full transcript
Sean: All right, one week later.
Andrew: It's amazing how time flies.
Sean: That just felt like a couple of minutes. Boy, oh boy, you know, it's that time of year, you know. Yeah, what's happened this last week, Sean? Oh, so much has happened since last time we recorded that I can barely think of a thing.
Andrew: And of course, the proof, the watermark, if you listen to the previous episode, is look at the backdrop, it's changed. That's right.
Sean: And not just because we did it in a hot changeover, you know.
Andrew: And we happen to be wearing the same clothes.
Sean: I forgot to change that.
Andrew: That's a bit of a giveaway.
Sean: Yeah, well, it's all right. We're quasi-uniform. You're still in your summer uniform, no black t-shirt.
Andrew: Yep, no, my sabbatical uniform.
Sean: The sabbatical uniform now, that's right. And yeah, so we're back at it a week later.
But I think, so one thing that was funny, again, just thinking about the last episode, is I love that we can start with something that's like, here's the news item from the latest and greatest AI thing and end with, what is language? We're like, we're like, we're like, and squid circles. That's right.
We don't need, I mean, I don't have, is peak humanity here yet? How would you know? Right.
Andrew: So, of course, once we've all got our sort of brain implants and we're telepathically communicating, why read anyway?
Sean: Well, you know, there always is this thing for me where I'm always like, you know, again, who doesn't love a story around the search for extraterrestrial life and all this kind of stuff? And always people being like, well, if they were out there, we would hear them. I'm like, no, you wouldn't. Like, if you caught the right crossover at that right exact moment of when we could, I mean, even if you look at human history, right?
Like, our, you know, the original radio waves and things that are out there that will be going on for infinity, you know, in the cosmos, great. Current things, we don't broadcast like that anymore. Right.
Right? Or if you're like the equivalent of the ant walking on the sidewalk, you don't know what Wi-Fi is.
Andrew: Well, plus the fact that the arrogance of saying, well, if they're out there, this is aliens, we'll hear them. You look at species on Earth. How many species are we still discovering that we've never discovered before? So things live on Earth that we have never, ever seen or experienced before.
Sean: Yeah, and how much of the oceans are unexplored and all this kind of stuff?
Andrew: So to say, you know, we can't find stuff on Earth, but of course, if there are aliens out there in the vastness of the universe, of course we'll find them.
Sean: Of course we'll find them. But, and again, I always think about this too. I'm like, yeah, but we can't even understand what like dolphins are saying.
And then you're somehow, you're gonna be like, well, if there's aliens out there, surely we'll know it. And like, oh yeah, okay. Because whatever.
Andrew: Somebody must have written that story. The fact that we are utterly ignorant of the intelligence on earth while we're looking for intelligence off-Earth.
Sean: That's right. We're like, well, you're like, good luck. It's like the condescending pat on the forehead, you know, like, good luck.
Andrew: The dolphins say, hey, hey, we're here. And say, you know, there's just nothing here at all. No intelligence anywhere.
Sean: They're like, we've been telling you this whole time. You fools. What are you doing with like
Andrew: Douglas Adams had it right all the time.
Sean: The whole time, right? Like, what are you wasting your time with? Cars and buildings and clothes. Like, this is fun, man. Just swim around in the ocean.
Andrew: Yeah, so what are we talking about today? Sort of dolphins and stripping off. Tell me about this episode.
Sean: So this episode,
Andrew: we're talking about the right to be unpredictable.
Sean: Oh, yes.
Andrew: I have no idea where you're going to take this. It's because I'm trying to be unpredictable.
Sean: Because you're being unpredictable. But I think the concept here is really taking a look at this, which I think ties in actually kind of nice to the episode we just recorded about the watermarks and again, how that watermark starts to either intentionally or unintentionally start to conform acts of writing, behavior, all this kind of stuff. So this idea of sort of in this world where everything seems to be optimized around predicting what's next, our behavior, our responsibilities, all this kind of thing, this concept of as, which is kind of in some ways, we can also extend this into a Futures Improv place. I don't know if we can quite tap the gold of the squid stuff. But this idea of does like things in the future, like unpredictability, does that become a right, a responsibility, or purely a luxury in having the freedom to be unpredictable in an age where everything is being conformed into?
Andrew: So with that, I think we ought to sort of do the intro and then you can say exactly what you mean about unpredictability. Well, because this could get really interesting.
Sean: Yeah, sure. Okay, so let's take a quick break and then we'll jump right back in.
Voiceover: Initializing waveform.
Sean: Modulating signals for transmission.
Voiceover: Exploring the possible, probable, and preferable futures.
Sean: Welcome to Modem Futura. All right. Well, we're back after movie magic. We're back. Well-rested. I'm Sean Leahy, joined by my co-host extraordinaire, Andrew Maynard. I almost sounded like co-horse.
Andrew: Co-host? Hey, now that would be unpredictable. Suddenly you're a senator. This is going to be one of those episodes, isn't it?
Sean: I think so. The punchiness is set in, right? When is lunch? I don't know.
Blood sugar is getting low. Unpredictability goes up. You know, we'll see.
But anyhow, that means you are listening to Modem Futura, the show that explores the intersection of technology, society, and the possible, probable, and preferable futures, asking deep questions like, what will it mean to be unpredictable in the future? And me trying to get through that while trying to figure out ways I can try to be unpredictable ends up not being unpredictable at all and very predictable.
Andrew: So actually, I'm waiting for you to be very, very predictable here and refer to Jurassic Park, which has got some beautiful lines about unpredictability.
Sean: That's right. Yeah, well, so anyway, yeah, now that I'm like, well, now, see, now, this is part of that. Now you've created the loop where I'm inside the loop. So now you have influenced me to no longer be truly unpredictable. Now my response is, in fact, predictable.
Andrew: But tell us about what you mean about the right to be unpredictable. How important is it and what's eroding that?
Sean: Well, I think the concept here about the right to be responsible... Unpredictable. See? Well, now, did you predict that?
I saw what you did there. But the idea to be the right to be unpredictable. Coming back to this idea that, you know, in an age, especially one in the artificial intelligence age where every our behavior, again, we've talked about this, I feel like almost countless times at this point of there is almost no way to escape the touch of or influence of AI across our sort of modern daily lives.
But this idea, too, that the constructs that we create for ourselves are, in a way, putting us into this conformity of boxes. And those can be literal, right? So the way that we split up the land into countries, into states, into municipalities governed by rules and laws.
Andrew: It's something that humans do. We organize the world around us. We organize society.
We organize our own lives. We're a weird creature in that we both like predictability, and yet we rail against it. Yeah.
And the interesting stuff comes out of unpredictability.
Sean: Yeah, absolutely. Well, I mean, even things when you consider, like, why do I have to have a first and last name? Yes.
I mean, it's understood. Because if you didn't, every time I said, sure, I'd have a thousand people saying, yeah. Me?
You know?
Andrew: Yeah.
Sean: But again, so pulling that, so I mean, again, Using some of those like sort of fun and ridiculous things, but also pulling that down into, you know, again, as we see more and more use cases, the way more and more platforms that we're engaging, relying on are building profiles of our behavior, expecting us to behave in certain ways. And then also, as we've talked with other tools and technologies, again, pulling in AI, because that's a really easy one to talk about, is then essentially surrendering to the way, oh, I should do it that way instead of maybe the way that becomes natural or intrinsic to me.
Andrew: And that is where I find this really interesting. So, and again, I have no idea where you're going to take this, which is the unpredictable bit, because I intentionally wasn't prepped for this. But if you look at sort of the history of humanity, it's this really interesting dynamic between the predictable, which we need in order to sort of prepare for the future.
Also an unpredictability, because unpredictability is where the innovation and the creativity comes, where you begin to find new ways of creating those futures. If everything was predictable, it would be static. You wouldn't be dynamic.
You wouldn't be changing. So part of our superpower is this unpredictability. And yet all of society seems to be built around minimizing unpredictability.
So you've got that. And then you've got AI that has been trained on primarily unpredictability. So you see this in how large language models respond.
They try and squeeze everything into a predictable box. And that is now affecting us. So you talk about used futures.
Effectively, the AIs themselves are beginning to impose used futures on us where we are not being given permission to be unpredictable. They force us into these categories.
Sean: Yeah. And what's interesting, too, is even some very low-level use of those AI systems, too, What you don't have is it prompting you to like, well, think about it. What are the other ways you can think about it?
It's like, oh yes, would you like me to create a rundown for this? Would you like me to put that into a PowerPoint for you? And you're like, well, okay, sure.
But like, but I thought I was trying to like learn.
Andrew: I think there ought to be a rule here that every AI response has to have the word boring in it somewhere. So would you like me to create a boring email for you? Would you like me to create that boring PowerPoint for your boring pitch?
Sean: Like, come on, Sean, that's boring. Can't you think of, I have the world's knowledge at my fingertips. Can't you ask me something better?
Andrew: And I think people don't understand this to the extent that they should. So if you think about large language models, they've been trained on a corpus of past human writing. So that establishes and captures the norms, the predictable.
And they try and squeeze our futures into that past. So an example that I think we've talked about this, but I've also written about, you ask almost any large language model to give you a profile of someone. And they will strip out anything that they do not think fits a professional profile.
So if you're an interesting and unpredictable, a whimsical person, all of that goes. All you get is the formal business-like stuff because an AI thinks that that is what it's supposed to do. It has no concept of this wider sense of what it means to be human.
And they've been trained that way. So now it's trying to pull you back to a conformity that disallows unpredictability.
Sean: Yeah. And I feel like it's really interesting when you think about it like that too, because again, that's exactly, you're losing so much of those pieces and it's forcing you down into those like boxes, again, that are pre-described. And then of course it runs all these questions, right?
Like, okay, well, where'd that come from? Who decides that? But also this idea too, I mean, again, even in the tagline of our show, right, is when we think about futures, one of the things is, this is why it's plural, right?
Is because we have possible, probable, and preferable futures, which are, you know, there's more you can stuff in there thanks to the Voros cone and all this kind of stuff. But the idea is that there is this unpredictability of this element. But I think what is interesting is the way that we treat unpredictability to equate uncertainty.
Right, yes. The fear of uncertainty. And the almost, like, the desire certainly, but then sometimes, you know, again, finding ways to justify, you know, further actions to this as people want a, especially those who are enjoying things in terms of you have power, you have money, you have influence, right?
So if you're winning the game, so to speak, you want a linear progression of what is to continue to exist because you can control it.
Andrew: Right.
Sean: You can understand it, you can control it.
Andrew: But even there, so, yes, Yes, people want control over the odds with the future. But even people in power or people that are comfortable want to ensure that life continues in a way and along a trajectory that they want. That involves change.
So there's got to be a change from where you are to where you want to be. But you want to be able to control the odds of that change being what you want. So that's where the predictability comes in.
But it's not about being static or not resisting change. It's about having a certain level of control.
Sean: Yeah. Yeah. I think, I mean, it's really interesting to start peeling. Again, this is one of those things where you throw these words out and you're like, well, yeah, you, and then you just, you can just start peeling them back.
Andrew: And yet, so this is, you're going to get me going on this because it is so central to not only my work, but my philosophy and how I approach things. The more you control things, the less you're able to see different possibilities. So go to those sort of different types of futures. sort of preferable futures, etc. If you are so controlled that you think you know what sort of future you want, you lose sight immediately of the spectrum of possible futures.
Sean: Absolutely. You completely humiliate those futures down. And when you do that, what you, so it's interesting too, like even just from like the futures talk and perspective is what you're doing is you're basically putting in these black swan breeding grounds, which is, So that's a whole nother thing, right?
This idea of black, this concept of these events or things that you can't see happen. Yes. For, and the only way you notice them is in hindsight and they have massive impact and stuff like that.
But these breeding grounds of black swan events become the result of exactly, of forcing yourself to not be open to what those possibilities could be.
Andrew: Right, right. And so you get to the idea of fragility as well.
Sean: Oh, absolutely.
Andrew: So predictability at some point equals fragility.
Sean: Yeah. How could we have seen this coming? There was never a possibility that we could have had, you know, fill in the blank, right?
For usually these are big disasters or something. But, you know, and then that kind of comes back into this area of like, well, when you intentionally or even just accidentally foreclose down your receptiveness to all those different possible futures, you are creating a very fragile and dangerous system in the sense that it is like, and then we'll talk about, this is where we can invoke the Jurassic Park. But like, when you've done that, you have completely left yourself ripe for disruption.
Andrew: That's right.
Sean: Yes. Whatever that looks like.
Andrew: And I, so talk about unpredictability. I had no idea that we were going to go in this direction, but this ties in directly to at least one strand of what I'm looking at on my sabbatical. And it's the idea of how do you equip people to thrive in a future, not just of AI, but transformative technologies that are changing everything that we thought was constant and reliable at a speed that we've never seen before. So it changes every week, every month. And the reality is that if you rely on predictability and constancy, you end up in that future exactly as you said, where you create a breeding ground for black swans. So how do you help people relish this idea of unpredictability and serendipity as part of that? How do you release them from the constraints of conformity so that they can actually thrive in a future that we're building.
And it requires a completely different mindset. So if you do not allow unpredictability, you effectively sort of sentence them to failure in the future we're building.
Sean: I think so. And I feel like people on a daily basis feel this tension or this struggle even in some very, what at the surface level might feel like very like minimal ways. But even things like prediction engines that, you know, algorithmic prediction engines, right? When you're like, that's not me. You don't see yourself in some of those.
Andrew: Right. Because they're aggregated.
Sean: Right. And you start
Andrew: Based on past data as well.
Sean: Based on past data or based on behavior. And you start seeing these cracks or like, I mean, I just, I'll share just a personal anecdote for myself is I get really frustrated because it feels to me like a lid or a box. And you're like, no, no, no. I want to see what's available.
Don't show me what you think I want to see because I get actually like really, I mean, you can imagine me late at night in my own office at home being like, right, you know, having like a little mini rant because I get so frustrated because I'm like, no, I want to have access to what is possible, not what you think it's you're being gatekept in ways because of past behavior, because of these things. And again, even today, right, this has become somewhat of behavior that we have learned from these systems, right? Like, I mean, if you're a Netflix subscriber, how many times you'd be like, no, no, no, no, you've just ruined my, you know, you can't watch that on my profile because now everything you get now is tainted from that.
Andrew: I know, it's based on that, yes.
Sean: Or I was having this conversation with my kids the other day because something we like to do is as a group, we'll watch funny videos on Instagram or something. And I'm explaining to them, like when we stop and watch these funny animal cat videos, I'm like, we will see nothing but, and I'm like, watch.
Andrew: Prediction is based on past behavior, and it doesn't understand the nuance of being a human.
Sean: No. And you're like, what I want is — and this is fun for me — in those social platforms, I really love seeing artistic expression, right? I want to see that cool painting, that music, the thing, or whatever. But what'll happen is you'll watch a cat video, and then it's all cat videos. That's it.
Andrew: That is it. The end. That is it. And it's amazing that we're still there. Yeah, it felt like we were there about 20 years ago. So remember, even 20 years ago, you remember sort of, you looked online at, I don't know, I use an example that we've had. You needed to buy a new toilet.
So you looked at toilets, you bought one, and then every single advert was for a toilet. You don't need a hundred toilets. But we're still there.
Sean: It still happens. And they're just lame. Because exactly, you'd find something, you'd buy it, and then a month after you bought it, you'd be advertised for the same thing you've already bought.
I know, I know, yeah. You're like, there's only so many bathrooms in this house. But then you extend that to almost every aspect of society.
Andrew: And there is an outlier here that I'm going to get to in a second. But certainly sort of in professional circles, there is this very much feeling of regression to the mean of conformity. So everything pulls you back away from exploring and being creative and innovative to doing things the way that they're done.
Even more so when it comes to education, We are dreadful at this. We had this idea that there are certain ways in which you do things with education and with learning, and you are not allowed to go out there. So simultaneously, we say that we want people to think critically, we want them to be creative and imaginative, but we put them in a playpen that is all about lack of permission.
A playpen says, you cannot do this, you cannot do this, you cannot do this, you have to do what we tell you in a certain way. And it basically robs you of that permission to be... Yeah.
What was the word? Unpredictable. Unpredictable.
Yes, yes, yes. So these systems are built to basically inhibit us, even in education, that claims to sort of allow this. The outlier that I was going to mention is, if you look at entrepreneurs and founders, so if you are starting a new company, you are basically predicating that on the unpredictable.
you're predicating it on your ability to do something, to think of something, to envision something that nobody else has. And the only way you can do that is to allow yourself to think way outside of the playpen bounds. And we know that that succeeds.
And so you now then go to universities who say, we've got to train everybody to be an entrepreneur. We've got to train everybody to be able to think like this. But you have to do it in a formal setting and a certain way with certain ways of assessing you.
and you're not allowed to color outside those lines. The two are incompatible.
Sean: No, absolutely. As you're talking, I was looking at it because I want to make sure I get the name right because I feel like it's gone by different terms in the past. But that's exactly it.
Like when we think, and again, there's like, I want to draw a line between education and schooling, or learning and schooling, right? Just to make sure that, because those, while you can learn in school and all that kind of stuff, they are not the same thing. But exactly that, like there is this, what we want are people who can have that, they can, where is that creative?
Where does that innovation, that creativity, that, you know, these new ideas, they come from critical thinking. Yet we also invented like the B.F. Skinner teaching machine, which is like a crank, right? So you must learn this, is the answer there, crank.
And so you can form this down into this capability where it's like the way then that we assess, did you learn this thing is in this little box? Did you do the thing in the box? the right way as it's predicted that other people will do that to achieve the same outcome.
To demonstrate you have the cognitive capability.
Andrew: You want that predictability to sort of say, we have put this student through the machine and this is the cookie cutter that comes out of the other end that we can describe and we can sort of understand.
Sean: And then in worse to that, right, as you can see then quickly how this sort of maps at bigger and bigger scales through a society is that now you are a unit. That unit can do what? We have now a prediction measure that you come out with some diploma, a degree, a specialization.
You are now a unit that can predictably provide what kind of output. Now you go into this box. So now we've moved you from the output machine of this box to this box, where now you're expected, again, predictably to perform in some fashion.
Andrew: Right, right. And of course, the irony with that is, how many times have you heard people say, we need problem solvers? We need people to use the cliche, think outside the box.
And yet we are training people not to be problem solvers at the same time as claiming we are. It's just a different box. Right.
It's just the problem solvers come out and they're the... You're only allowed to problem solve in certain ways that are based on sort of past precedent.
Sean: They're like, don't worry, we've got a box. That's the problem solver. We'll put that problem solver in that box.
Andrew: It's like, no. And it doesn't work in the world that we're building. Maybe it did work 20 or 30 years ago where things sort of evolved at a slower pace. But in the world we're creating with technologies that are transforming everything, you need people that are agile, that can see possibilities that you simply cannot see if you've been shoved in a box or a playpen.
Sean: Yeah, no, absolutely. And I think this comes back to something. I mean, I've been on this kick now for at least 10 years, which is about this conversation.
And I've probably mentioned it, if we were to do a word search on the transcript of our podcast, I've probably mentioned it a bunch of times, is this idea of what we really want is a performance of understanding. right that in that so what is that it's not a necessarily a performance in the way that you might think it's not a it's not a concert or a thing right but it is a way that you can demonstrate the depth to which you understand the the situation the problem what the your uh sort of creative or critical thinking might be on how to solve or move or do a thing but you get to define what the what that looks like. That's right. It's not a, and so for people, if you're, if people are struggling with what that means, it's like, that's not taking a test on like a math exam of like, you know, here's the formula, here's the paper and a pencil and go forth and demonstrate. It's maybe using real world scenarios. Like how do you, like a classic example would be asking students, for example, of how mathematically to come up with the, the volume of a cone. You can do that with a formula. You can sit there on a piece of paper. You can do a multiple choice. You can also pick up a traffic cone and describe how you can determine the volume what's inside that cone.
Andrew: It shows that flexibility. And of course, in education, we talk about this both in terms of formation. So this is not sort of how you do in a test, but how going through the process, you have formed your ability to make sense of the world around you and to apply that sense to new challenges, but also in practice. So if you look at how we train doctors, if you're training as a doctor, you don't just do an exam and then you get put in front of a patient on your own. Yes, you have to learn the knowledge, you have to go through the exams, but the formative process is through experience. And the test is whether you can successfully help patients or whether they are not helped by you. At the end of the day, it doesn't matter how many tests you've sort of gone through. It's whether your patient actually gets better or not. And the same with engineers. So in those professions, we know that. It is not book learning at the end of the day.
it is that ability to problem solve in situations that you've never ever been trained for before.
Sean: Well, and I was saying, as you were talking about engineering, I'm like, we have a long history of examples of why that's so dangerous. Right. As you have those, that's where you have those collapses, those things that were That's right. You're like, well, the math works. And you're like, well, but you didn't consider the human factor of like, no, there will be 80 people on that balcony instead of the four it was designed for because of people.
Andrew: So we know this, and yet we continue to persist in saying that sort of you need to form these sort of predictable pathways to a skill set, which is rigidly set, rather than teaching people to problem solve and adapt to situations that they've never been trained for, never experienced before, but they have that formation that allows them to do that. It's incredibly hard to put a number from an assessment on that, but you show that they can do it through a matter of trust, because you trust the system, but also the proof that they're actually thriving.
Sean: Yeah. Yeah. No, I think that's really interesting. And kind of taking it in a slightly different direction, too, is, again, kind of on this tale of unpredictability versus predictability is also, you know, again, as these technologies continue to evolve and, you know, converge and disrupt faster and faster and faster, right, you start getting in, at least my brain went this way a little bit, too, is sort of moving into that sort of the Philip K. Dick sort of Minority Report. Again, using that as an extrapolation of an extreme, but with this idea of like, okay, so if we surrender to an extent to these, you know Deterministic models.
Andrew: Deterministic models. That say we can predict everything.
Sean: Yeah, and you can start to see where this starts to get really dangerous because we're seeing some of this propagate in all kinds of places. Now, humans have always had measures of gatekeeping and stuff forever. But now with tools like artificial intelligence, also then being, but not just like someone using those tools, those tools being used to look at behavior, to look at social economic status, to look at where you're born to determine your statistically probable outcome of where you end up. To then also to pre-essentially pass judgment on what your role in society could and should be, right? Starts getting into this really dystopian.
Andrew: It does. But it's amazing how many people actually believe this. Even people that are academics, who believe that if only we had enough data and big enough models, we would be able to predict everything.
And it scares the life out of me. Because, yep, so if that's true, it's effectively, to me, the death of civilization, because there's a dynamism in civilization. Because we live in time, we're either evolving and growing and changing, or we're static and dead.
And to evolve and grow and change, you've got to have a level of unpredictability. It's what our biological heritage is based on.
Sean: Right. I mean, that's how you get a new trait.
Andrew: It is. And it's how you discover new things. You're unpredictable. You have this sort of random walk that occasionally bumps into something different.
Sean: That's right. I mean, we won't have time to talk about it today because this will open up a whole other can of worms. But that's also where you start to bridge on this entire incredibly weird and wonderful field of the adjacent possible.
Andrew: Right.
Sean: is through that unpredictability.
Andrew: How do you hop over into that adjacent area?
Sean: And some of those things, I mean, again, looking back at things that have been instrumental in the formation of the advancements in science and technology of human endeavors, we're all born out of these adjacent possible things. Like glass, like forming, you know, compressed silicon into glass to turn that into lenses that allowed us to look into the cosmos and to the microscope. Like these are the electric light. I mean, all of the things that we do, I mean, these things that are
Andrew: Yes, and you can say yes, once you understand the science, that's predictable. But until you get there, there's a level of serendipity that leads to those first initial steps. Yeah. I mean, it's incredible, right? When you think about those
Sean: giant leaps. But back to your point too about that, the death of that piece is like, I mean, again, when, yes, we can talk about some fancy things, we can throw in Futures Improv, but you can also feel this today. Again, whether it's your Netflix, you know, recommendation engine, or the controversy of the Flock cameras in the United States, right? That are, these are basically advanced traffic cameras that are...
Andrew: But not just traffic. I mean, the indications are that they're trackers for a bunch of things.
Sean: Yes, including, and there are some that, including using different technologies that will sniff Bluetooth connections to just be like, well, who do you connect to? Who do you talk to? But again, the danger there too is, again, kind of going into that road where you can see this, again, this used future of Philip K. Dick's Minority Report is this almost like panopticon level of surveillance of where are you going?
What are you doing? And from that, what can we guess? What can we predict?
And that leads into, again, this idea of, you know, what does a, as a scenario, right? What does a future look like where we are relying on these AI precogs
Andrew: to determine like, oh, this person actually,
Sean: the way they're driving suggests they're going to, what, commit a crime? They're going to endanger someone. So therefore, we should take action now to prevent that from happening.
Andrew: And of course, this is when I teach from Minority Report. This is what I teach. The precogs, if anybody has seen the movie Minority Report or read the original short story, they are a fantastical idea.
These genetically sort of modified sort of humans who can sort of predict the future. And yet the analogy is AI that is trying to do exactly that through precedent. And again, it sounds attractive.
If we could predict the future, surely we could build better futures. But counterintuitively, it doesn't work. The more you predict the future, the more you put walls around what is possible.
Sean: Yep. And again, you start... And then, correct me if I'm wrong, in that story, wasn't that one of the sort of plot pieces is that the precogs saw different futures?
Andrew: Occasionally, they saw a minority report indicating that things couldn't change. But actually, also the flip of it was, once you have a system that is built on prediction, if you are an unpredictable human, you can basically work the system.
Sean: Yeah, because you become part of the loop, right? Right, right, that's right.
Andrew: So as soon as predictability is there, it is brittle. And if you're a smart, unpredictable human, you can actually break it and use it.
Sean: Yeah, which is, again, but also that fascinating thing. And there's also this element too of, right, like you can see a system at a societal scale, right? So you're entering this from birth.
telling you statistically, oh, you were born in this neighborhood from these parents, therefore your best outcome in life is going to be this. And maybe that's not something that's desirable, right? And so even just being handed that future is going to impact the way you think about yourself, the way you think about your environment.
Andrew: Then you look at it in today's society. So we wouldn't have been having this conversation probably even 12 months ago with regard to AI. But AI is now at a point where it is deeply influencing how people think and how people behave and how they see their futures at every level.
So if you think about the conversations that individuals will have with AI about life choices, it is giving them advice based on predictability, based on sort of past data. So it is creating a playpen around you as you ask it for advice. It may feel liberating.
It may feel as if you've got this superpower that allows you to sort of make informed choices. But actually, it's basing those choices around a very limited set of possibilities. It is constraining you.
And it is trained to do that. Yeah, absolutely. And just as an example of that, so I can't remember whether we've mentioned this before, but with the Future of Being Human Initiative at ASU, we now have this online knowledge base at beinghuman.fyi.
Sean: We need to talk.
Andrew: We do.
Sean: We'll mention it here, and we should come back at another time to do a deep dive on that.
Andrew: It's been such an interesting experiment, because this is a really deep database developed in conjunction with AI using Anthropic's Fable. And it is unlike anything that I've ever built before or seen before. But in the process of this, we had to get the AI to understand what we do at a really deep and complex and forward-looking level.
And it kept trying to constrain us into what is normal, sort of what is normal predictable behavior. So if you're an academic, these are the things you do. If you're in business, these are the things you do.
These are the things you do not do. You don't play. You don't sort of use imagination.
You don't use sort of Futures Improv because those are all unprofessional. And it tried to keep on constraining us into what is considered to be professional in a very old-fashioned way. And when we were doing this, we had to push hard to get the AI writing this knowledge base to understand that at the core of what we do is being unpredictable in very, very profound and important ways
Sean: because that is the secret sauce of what we do. 100%. And that's what's really fascinating about it too.
And I'm sure there is countless number of people out there who feel that some ways. They're like, I don't fit. I don't fit the mold.
I don't fit the thing. It's like, yes, and, but, and a lot of times that and is the secret sauce of what makes you so amazing and interesting and good at your job or whatever. But that's not what, that doesn't make it to your little bio sheet.
Andrew: That's right. And actually, Futures Improv is a great example. So if you're asking an AI to sort of create your LinkedIn profile, almost definitely wouldn't be there because Futures Improv feels like a hobby, not a professional thing to do. Apart from the fact the way we use it is profoundly professional.
Sean: Oh, yeah, absolutely. And also, yeah, I mean, near and dear to my heart as well because it is so much. But again, it misses that.
And even just that behavior, even watching something, And something is lowbrow, if you will, as just like just watching LinkedIn. Right. Like just the way it has changed over like the behaviors and the way like from me, and again, maybe this is the algorithmic poison that I have to swallow.
Right. Is that when I look at it, every post is the same. Yeah.
It's the same cadence. It's that conformity to the mean of what, and I'm sure most of it It is predictable and boring. It's so boring.
And again, oh God, I could go on. Because it is, it's also like, and no one reads. So when you put something real in there and you put this thing out there and you're like, yay.
And people respond to like the blip on the thing. They don't actually read the content. I know.
And then you're like, well, why did I bother? I should have just, going back to last episode, like I should have just had Claude just write me something and then who cares?
Andrew: So we were talking about this the other day between us. I mean, the irony of LinkedIn, where you can put up a post saying, hey, we've got a really interesting episode coming up about X, Y, and Z, and you'll get thousands of people liking it and sharing and everything. The actual episode comes out and you say, hey, here is the episode.
Here it is, man. Absolutely flops. Nobody's interested.
And you're like, what the heck? This makes no sense. The thing with the content, nobody is interested in.
Sean: Yeah, the content, people don't actually want to read it. They don't want to read it. They don't want to listen to it. They don't want to watch it.
Andrew: Or the algorithm decides that this isn't of any value.
Sean: You know, what's of value? And then you put out something like, I forgot to tie my shoe today. And it's like, oh, 10,000 hits or whatever.
Andrew: It's just ridiculous, right?
Sean: Oh my goodness.
Andrew: But even that, so coming back to unpredictability, it shows that the system is stacked against it. So even with LinkedIn, maybe a trivial example, but you put something which is unpredictable on LinkedIn. It doesn't conform to what people expect. It doesn't go anywhere. Predictability gets you the attention.
Sean: Yeah, it really does. And there's a whole economy around that of gaming and YouTube, all these things that are algorithmically driven. it's the same thing as like, what, what is the secret sauce to get you picked up? Um, but it does, and it does, and it rewards, you know, those things. And then you start changing your behavior based on that reward mechanism. So even if the intent was like my podcast or my newsletter or whatever it might be that you're trying to share is unpredictable, bold and crazy and wild. And you're like, well, how you do it? Well, I can form to the algorithm ways to do it because that's how you get your views and your hits and all this stuff. And this is where it hits home really hard.
Andrew: This idea of the permission to be unpredictable. And of course, you've got to ask, well, who gives this permission? Where does it come from?
And I think partially that permission is permission given by the algorithms that rule our lives. Permission is given by the communities that we're part of. Permission is given by the expectations.
And if you're in an educational establishment, a university, ideally, permission is given by the establishment to not only be unpredictable, but to have that celebrated and validated rather than penalized.
Sean: Yeah, absolutely. And, you know, I mean, I guess you could say, yeah, I'm sure some unpredictability would be bad. Not all unpredictability equals creative.
Andrew: There are got to be sort of social expectations here. But, you know, the little stuff like sort of, I don't know, arranging your office in a way that actually brings joy and actually allows you to be creative. Absolutely.
Sean: Or that's conducive to your workflows and stuff like that, too. Because, I mean, there is a long history of the relation to space and to productivity and things like that.
Andrew: I feel actually, just thinking about this, I need a sign that says, predictably boring. Unpredictably, yeah. Unpredictably interesting.
Sean: Yeah, or something. Just start wearing t-shirts that, you know, be unpredictable or something. I don't know.
Andrew: But again, it comes back to sort of startups and founders. They thrive on the unpredictable. If they're all trying to do exactly the same thing, nobody would get anywhere.
Sean: Yeah, no, and I think, you know, and again, that holds true with lots of elements too, where it's like, you know, and again, going back to, you know, just picking on Future of Being Human, you know, if you go back and look at our sort of core elements around these different things of, you know, intentional, you know, serendipity, radical creativity. It's hard baked into the system. All of these pieces are meant to keep, I don't know, I'm trying to think of it now, to keep the water boiling. Yes. That these are all ways to keep it fresh, keep it new by looking. Personally, from my own perspective, when there's no more change or advancement,
Andrew: it's boring. It is. And it's stultifying.
So one thing, an image keeps coming back with these conversations. And it's a bit of a flip on the image. So you remember the 1984 Apple adverts where the Mac came out.
Oh, yeah. So you had the gray masses, and then you had the colorful athlete with the hammer sort of running down.
Sean: Yep, the big sledgehammer with it. And then the later one with the iPod that was digitally added.
Andrew: Right, right, yes. But I feel everybody seems to think that they're living in a wild, creative sort of world. They're not.
They are the gray conformists. Yeah. And somehow sort of we've got locked into this idea of thinking that because everything is advancing so fast, we're not the conformists.
but this is the gray world. And we need people to be able to have permission to be unpredictable in it. Yeah.
And bring a bit of color to it.
Sean: Yeah, absolutely. I'm trying to figure, I'm trying to, all of a sudden I'm like, trying to think of the, I'm looking at the language because what's that, the line that's in there that this is for the, oh my goodness, this is going to drive, I should have, I wish I would have.
Andrew: All I remember is something like.
Sean: If you would have been more predictable, I would have gotten this.
Andrew: All I remember is the line about sort of 1984 not being like 1984 or whatever.
Sean: Yeah, yeah. Oh goodness, I need to try to find this I think it was, this is for the crazy ones, right?
Andrew: It's where you do things on the fly. Yeah. But as you're doing that, I do think that people need to start thinking more seriously about the AIs we're creating, the establishments that are being created around them and relying on them, and now becoming the gray conformist masses.
Sean: Yeah. Yeah. No. Yeah. This is probably, I probably shouldn't. No. I probably shouldn't read this and we'll get taken off here. But basically it was like this opening line to Apple's Think Different campaign, which was leveraging that, right? Which is this Think Different, again, really espousing that message of
Andrew: Can I just say something? So the number of times I've invoked the Think Different campaign and somebody corrects me and says,
Sean: that should be Think Differently. That's the whole point. It's like, no, you missed it.
Like you fool. But again, going back to this concept of that think different sort of advertising campaign, which embodied that idea that this is for the crazy ones. This is for the people who think differently.
That's exactly it. This is for the misfits, the rule breakers. Yes.
Again, not criminally. Yes. Within the confines of our agreed society.
Andrew: It is the permission to be unpredictable. And actually, that whole campaign was giving people permission to be unpredictable.
Sean: And I will run this risk because I will defend it to the last breath. You can accuse me of being an Apple fanboy all you want, but I will stick to this, some of those guiding principles of the technology is meant to be the bicycle to the mind. Yes. The whole point of that.
Andrew: The jobs quote, yes.
Sean: And the whole point of this, and it's interesting that, you know, again, yes, massive, you know, what, $5 trillion company or whatever it is now. But also that at that core to it was that creativity. It was this spark of being the color to the gray monotone.
Right. That it is, that is what makes the difference because that is where you get those new ideas, the breakthroughs, that's where the new art, the expression comes from. And I would argue again, as we ask this question all the time of what will it mean to be human in the future?
Is it better? I will put my predictability on this, right? Is that it better come with it, those joys of being creative and unpredictable.
And unpredictable, absolutely, yes. And finding those new things, those new discoveries that could be personal, that could be in your family, in your work groups, all the way up to the societal global level of being able that the technologies, the systems that we're developing and designing and putting in place should not only allow, but intentionally provide the opportunities for that to happen. because without that, you do.
You end up marching towards these dystopian hellscapes that we've been warned about for decades, going back to 1984, for example, right? I mean, what's the opening line on that? It was like the boot on the neck.
You know, this imagery of like what the future feels like
Andrew: is a combat boot on your neck, right?
Sean: Like that is not the future that we want.
Andrew: There are days that actually I feel using Claude, it feels just like that. And so this, again, is a conversation for another day. The days that I find Claude and almost any other AI absolutely insufferable because it feels like the boot on my neck telling me how I should be conforming and thinking.
Sean: Yep. Like, no, put it this way. You're like, no, no, no. I want it like this.
Yes. No, what you need is like this. You've got it.
Here it is.
Andrew: Like, no. So the bottom line is everybody listening, you have or should have the permission to be unpredictable. And just to sort of wrap up the playpen idea that I've been playing with, because again, this is central to my work, we need playgrounds. We need playgrounds where you have permission to experiment.
And it's not limitless experimentation. So a playground is designed so you can experiment without causing serious harm. There are constraints there, but within those constraints, your imagination is given the permission to run free.
Yeah, yeah.
Sean: No, that's wild. Well, one anecdote I have on that too is something that was, again, it was a cultural sort of shift that I experienced living in the Netherlands. Yes.
Walking by a playground. So, you know, in the U.S., again, you know, I grew up for the most part, you know, I grew up in the Midwest and all this kind of stuff. And, you know, so my vision of a playground is fixed to whatever these, you know, equipment catalogs for schools were in the 1980s and stuff like that.
And then I would just remember walking by a schoolyard in the Netherlands and their playground was like wood. Right. I mean, like planks of wood.
That was it. Like no, no, no, like plastic prefab like slide. It was like wood and like stumps of different sites.
And the kids would just build stuff. They would literally make their own like equipment out of the wood.
Andrew: It's the same principle as at Christmas, when you give your kid a toy in a box, they play with the box. Oh, yeah. Because they want to be able to have that environment where their imagination can run away rather than being constrained by what you want them to do. Yep, absolutely.
Sean: And even with toys, we should talk about that some other time with toys because that was a big thing that I was really into back when my kids were really little, was being very careful about the types of toys that we got in the house because of the closed or open nature of the toy in terms of like, you know, is this something that like the use of it was prescriptive versus something that was like, use your imagination. Cause again, humans have been playing with toys as long back as we have archeological evidence of humans. So, um, yeah, cool stuff.
Andrew: Yes. Yep. Yep. To be continued.
Sean: Yeah. Dot, dot, dot on that one. Oh man. See now to be truly unpredictable, we should have just like mid sentence stopped the show or something, you know, but, um, yeah, but actually, um, there is unpredictability and
Andrew: unpredictability. So let's leave with this. This idea of constructive versus destructive unpredictability. Oh, I like that. And we'll come back to that. But there's definitely that
Sean: framework where constructive unpredictability is where we want to land. Yes, absolutely. And so, yeah. So if you want to do us the favor and be predictable, sign up. No, no, no, no. Give us the data. Be unpredictable and go and sign up. Or leave us an unpredictable comment.
Andrew: Live a little.
Sean: Do something different. Listen to that inner voice, you know, as long as it, again, is legal in your country or wherever you're listening to this. But I think that would be kind of fun. Yep. And I will also throw, so, you know, again, do us a favor, share the show, do the things, sign up, subscribe, like, do all the things that you can do to just, again, it's talking about predictability. It helps those algorithmic machines help hopefully find people who might be
Andrew: And it brings us a ray of joy as well. It does. Knowing that somebody actually values us, it actually, it means a lot.
Sean: You know, there are days where we also need validation
Andrew: from total strangers. You know, and there's somebody listening thinking, you know, a couple of dudes, so predictable. Yeah, you guys are, you guys just need to stop.
Sean: Like, okay, well, you know, you can go ahead. Please tell us that then. I don't know, go for it. But also, you know, again, if you don't like signing up through that traditional methods, if you want to avoid the algorithmic chains,
Andrew: you can also Just go to modemfutura.com/signup. And it's just a little one-liner, which will end up giving you an email each week saying there's a new episode to listen to.
Sean: Yep. And I will throw in one final plug because you mentioned it. We'll also, if you fancy just an interesting dig and tell, because we have not talked about it at length, is try a different website.
Try beinghuman.fyi and just have a go. And that's all I'm going to say. Just go there.
Go down the rabbit hole. Just go there, have a go. And, you know, hey, maybe that'll change what you respond or what comment you leave or whatever.
But anyway, yeah, Andrew, fantastic episode as always.
Andrew: As always, yes.
Sean: Unpredictable and fun. It was predictable how unpredictable the conversation was.
Andrew: It was constructive unpredictability.
Sean: We're living what we're talking and that's what it's all about. But yeah, so anyway, we'll catch everybody on the predictable next episode of Modem Futura.
Andrew: We'll see you then.
Sean: Demodulating signal. Modem Futura is a production of the Future of Being Human Initiative at Arizona State University. Be sure to subscribe on Apple Podcasts, Spotify, or wherever you listen to your favorite shows. To learn more about the Future of Being Human Initiative and all of our other projects, please visit us on the web at futureofbeinghuman.asu.edu.
Voiceover: End transmission.