Skip to content
S4E4 The Intelligence All Around Us episode artwork

Aug 15, 2024 1 hr 13 min

S4E4 The Intelligence All Around Us

from Invisible Machines podcast by UX Magazine Hosted by Invisible Machines

There are many lenses through which we can view the world, and that's especially true when considering the technologies associated with artificial intelligence. In this episode, James Bridle, author of Ways of Being, joins to explore a multitude of perspectives relating to our ongoing relationship with technology. James’ book explores the interplay between animals, plants, and machines in the search for planetary intelligence. James offers a broad vision of "thinking" systems that draws from biology, physics, and computation and spurs an investigation of what a corporation is and could become. Robb, Josh, and James have a dynamic discussion, trying on various lenses for viewing intelligence and uncovering new ideas about technology along the way. Learn about orchestrating conversational AI agents for your team: https://onereach.ai/ai-agents?utm_source=soundcloud&utm_medium=social&utm_campaign=intelligence_around_us_episode&utm_content=1 Get James' book, Ways of Being (Farrar, Straus and Giroux, 2022): https://us.macmillan.com/books/9780374601119/waysofbeing Learn more about James Bridle and their fascinating work: http://jamesbridle.com #AIPodcast #AI #ArtificialIntelligence #WaysOfBeing #PlanetaryIntelligence #Technology #Innovation #ConversationalAI #MachineLearning #TechDiscussion #AIAgents #invisiblemachines

Episode details come from the podcast's RSS feed.

Embed this episode

Ready to play

S4E4 The Intelligence All Around Us

0:00 1:13:30
of matches

Transcript Auto-generated, may contain errors

Welcome to Invisible Machines, a podcast exploring the expanding world of AI agents and conversational technologies. I am Josh Tyson, co-author of Age of Invisible Machines, a best-selling book all about the agentic approach to automation and the pursuit of organizational AGI. Each week, I am joined by Rob Wilson, CEO and co-founder of OneReach.ai. Rob and I wrote our book over the course of many, many conversations, drawing from his extensive experience with research and deployment of conversational solutions. We are continuing our conversation here with UX Magazine because attention to experience design is critical for the proper deployment of AI agents or invisible machines. We've talked a fair amount on this podcast about the many lenses through which you can view technology. On this episode, author and artist James Bridle helps us try on a whole multitude of fresh perspectives. Their book, Ways of Being, explores the deep interplay between animals, plants, and machines in search of a planetary intelligence. James draws from biology, physics, and computation to bring us a broad vision of thinking systems. Even if humans have failed to recognize or acknowledge them, animal intelligence, plant intelligence, and other intelligent natural systems

demonstrate complexity, agency, and knowledge. So it's possible that by better understanding them, we can change the way we see ourselves, our technologies, and our societies. Maybe we can even forge a more balanced and equitable world for all of our planet's many inhabitants. In our time with James, we cover everything from plant memory to the enduring relevance of so-called ancient wisdom to the ways that corporations already exhibit many of the key traits of an AI system. We are big fans of James' work, and we're thrilled to have the chance to explore a whole array of bold ideas together. A quick note, Invisible Machines is available wherever you get your podcasts.

You can also watch these conversations unfold on our YouTube channel, which you might consider for this episode as James shows us an oddly sized chair built with design help from ChatGPT. Enjoy this stimulating conversation with the brilliant James Bridle. Invisible Machines is produced for UX Magazine in partnership with OneReach.ai. Named a leader by every major analyst group, OneReach.ai's generative Studio X platform was built specifically to facilitate the no-code orchestration of AI agents. GSX is trusted by global giants like PepsiCo, PwC, Deloitte, and DHL, but anyone can experience the level of agent orchestration we talk about on this podcast through interactive demos at OneReach.ai. Now, on with the show.

Well, James, thank you so much for joining us on the podcast. Very excited to have you here. Thanks very much for having me. Rob, a pleasure is always seeing you. So, James, I was in my car this weekend and I was listening to an NPR segment that featured a poem written by an LLM. Those certainly aren't in short supply out in the world right now, but I was surprised to suddenly find myself sitting in a parking lot kind of mulling over the words and feeling quite moved. You know, as someone who's kind of written my fair share of poetry over the years, you know, there's part of me that wants to see this kind of thing as an affront, but then there's this other part of me that's kind of open to it. And I think it's because I sort of realized that this poem, like a lot of the generative output out in the world, is sort of a, you know, it's not really a shard of some separate technology. It's sort of like a kaleidoscopic vision of a big...

hunk of shared human knowledge. Your book, Ways of Being, is kind of a deep exploration of the symbiotic relationship between humans and machines. I was wondering if maybe you could talk a bit about the broader nature of AI and some of the hidden truths out in the world that it might help us see. Well, thank you. I mean, that's a pretty big question. There's lots of many, many different aspects to that. I've written a book on it. Possibly. Well, I've written a book on it. I mean, I've written a book about, you know, quite a, my own very personal, narrow take on this whole thing that we call AI. And I guess it's worth starting any conversation that starts with AI by qualifying that a little bit, saying what that is, trying to clarify what we're talking about, but also to clarify what I'm talking about when I talk about it, which is that...

which is still really hard to do right because on the one hand you're talking about like what actually exists right what made the thing that wrote the poem that you heard that is a set of statistical relationships embodied in a data set uh run through by an algorithm that's what it is um like that's a very complicated thing it took a lot of people a long time to build it took a huge amount of power it took a huge amount of data it did all those things but that ultimately is what it is um you know at the other end of the scale you have this thing that we call ai And in our head immediately there's robots and spaceships and like the whole panoply of 20th century science fiction and all the cultural stuff that kind of falls out of that. And all of our strange technological imaginings, plus literally anything that a marketing person can slap the words on. Like just so many possible references and ideas and hopes and fears and dreams because that thing, that cultural AI is also like, you know, is it a robot butler or is it like Skynet that's going to take over and kill us? You immediately have to like deal with all of those references at once. And so, you know.

For me, what I try to do is try, and all of them by thinking, but particularly when it comes to something like AI, is just try to remember which lens we're looking at this thing through. Like what are the set of assumptions that we're talking about? Which of those types of AI, species, whatever it is, are we kind of talking about? Like the one that actually exists, the one that like tech bros tell us might exist in five years time, the one that's existed in our head for 20 years, the one that will clearly never exist, but it's really cool to think about. But also, you know, to play with those lenses a bit. Yeah. To say, well, you know, it's kind of fun to look through this lens. Like when you hear a poem or a song.

that was apparently generated by one of these machines, like it's also okay to let go and feel what you feel, right? Because they're just words that have emerged into the universe by God knows how many strange processes. And they can have an emotional effect and you can disconnect that also from a lot of the other discussions around it and just feel what you feel. And so, yeah, trying to bring a few of those lenses to bear, sometimes putting one on top of the other, sometimes, you know, intently using the wrong one, you know, and just kind of seeing what happens. That's kind of my, that's my thinking about AI. I could go deep into like aspects of the history or my own thought about it, but I think that's a good place to start. For sure. I like it. And let me put on a lens.

We can put on other lenses throughout this conversation. Let's start with, I think, the lens I'm interested in most, which is humans have been creating tools, using tools to enhance our experience on Earth better, right? Whether those tools were dicks and stones at some point. And over time, we've evolved those tools, but some of those tools are very inert, like... like a rock that we chiseled into something. Some of it is like a horse, like another living being is our tool. When we turned our environment, that is the environment we live in, nature, we started converting it into all, making everything our tool, right? And sort of modifying our environment and trying to turn our environment into a tool instead of like just being in it.

um in the harmony that sort of exists we're trying to adapt and create our own ecosystem and and utilize every every aspect of the living planet around us to the point where you don't even recognize the nature in it anymore like an llm like it's it's still rock i think as you said in your book right it's still rock it was metal and um but you don't even you don't even recognize it anymore as rock and we've converted into into tools and tools to what end like as an llm as a tool is a pretty pretty fascinating tool but how how does it and why should it be treated any differently than any other tools we've created um or should it um i'll jump right into like trouble one of the sort of ways in which that can be described, which is by saying that there's no such thing as the environment. The idea of the environment is something that's kind of been built up through particular Western European ways of thinking about our place in the world. And it's something that even plagues what we call the environmental movement to this day. Because the thing is, whenever you talk about the environment, you're making this space between you as an individual, us as a species, and everything else on this planet. There is no such divide, right? We can't separate ourselves meaningfully in that way from everything that surrounds us.

And realizing that, and I could go into lots of examples about why I think that and why that's an interesting, useful thing to do, if it's a lens, if it's a truth, we can say. But what it does is it transforms everything from being like this tool use, as you describe it, which reinforces that distinction, right? It transforms it into a relationship. It's to say these are actually relationships that we have with things. And the second thing I trouble with is the notion of a stone as inert. Partly because my cosmology admits to the beinghood of rocks and stones and mountains and all kinds of things, and I do not consider a stone to be itself inert. But also because...

There's no such thing as the environment and there's no such thing as a separation between us and the stone. And so when we take this stone and we transform our relationship with it by picking it up, by wielding it, by chipping it, by coming into this relationship with it, it doesn't just change the tool, it changes us. Because we suddenly can act on the wider world in this radically different way. Suddenly we can bash things, we can smash things, we can break things, we can skin things, we can dig things. And our relationship to the whole world changes as a result of that because of the agency we have as a result of that relationship we've created with the stone. And of course it changes.

with the horse as well and with your cell phone and with an LLM in all these different kinds of ways when you come into a relationship with them. Many of those relationships are deeply extractive, domineering, one-sided. The stone is an obvious example. Most of us just think of the stone as a thing. Fine. It's fine to do that most of the time. Most of us think of horses as just, or a lot of people, I would say, think of horses as just things or sort of somewhere between things and humans. That's a huge problem because that's part of the reason why we completely despoil and poison the planet is because we see other living beings are somehow a bit lesser than us and so we can use them as tools in this way.

And something similar may happen with many of our technological tools, with things like LLMs, with things like cell phones, with things like the internet, with these tools that we consider to be higher technologies in whatever way. Because we don't think enough about the relationship that we have with them and the relationship with the rest of the world that is changed by our relationship. So what changes with everything else when we interact with this thing? And one example I would give of that is the stone and the horse. They're actually very similar relationships in that they're deeply physical relationships. They're embodied relationships. You may use your mind to think about what you're doing with the stone. Think about how you ride or break, to use a horrible phrase, the horse, in these kind of ways. But you use your body to do it. You're very obviously in and interacting with the world in a very broad sense in that relationship. What happens when we relate with high technologies is that kind of...

the corridor or the passageway or the line of communication with the world narrows down to this degree because we think for starters and most of the time we really are only using our minds like our eyes are barely moving like maybe a finger twitches to tap on the screen right and so much of the embodied nature of so many of our relationships falls apart and therefore our brain goes into complete overdrive and it becomes this like incredibly cognitive relationship which has a whole bunch of other effects but the main one of which that I come back to is what I call computational thinking is that that relationship is so intense with this particular way of thinking which is the way that machines like digital computers think is that it sort of overrides everything else and we start to think that maybe we're like computers and maybe the world is like a computer because it's such a powerful experience but it is really only one way of relating to the world it's totally different from embodied bodily physical relationships and it's also totally different from so many other possible relationships that we might have Yeah, it was interesting. You brought up horses and ecosystems because I was thinking too about a few things that I've seen semi-recently in the news. Like there's been developments with the technologies associated with AI that might let us communicate with whales or other animals, which is both fascinating. And also, I guess we better start crafting an apology in that language. But there's also like kind of like ancient wisdom, right? Like when you think about all the growing scientific evidence that shows how important the gut biome is and how complex it is and how it's almost like a second brain and how that might correspond to, you know.

paintings that you might find of chakras from thousands of years ago it kind of made me wonder if it's easy to think of that that time as sort of like a time where there was less knowledge but it was sort of like a different knowledge and it almost feels like we're in a pendulum swing moment where we're starting to come back to seeing that like some of what ai is unearthing is that that was there all along and now we're kind of having this deeper understanding so it almost has this effect of in my mind anyway like kind of connecting certain bits of ancient wisdom with kind of a different way of looking at it but hopefully not like having this separate relationship between uh things anymore Yeah, I mean, I'm sure you didn't mean it expressly that way, but I will push back very strongly about the idea of a kind of ancient or traditional knowledge as being somehow less knowledge about the world. There are different ways of knowing the world. And ways in which actually compared to our own ways of knowing the world are massively more broad, more expansive, more connected. Our way of understanding the world has actually been massively reduced. What we've done is removed so many of these ways of knowing and concentrated on really just a couple of ways of knowing. When I say we in this very hand-waving way, I mean like Northern Europeans, North Americans. We call the Western acculturated peoples of various kinds. It's a hard thing to pin down, but you know what I mean.

And so what's, what's... Yeah. So what you're talking about is this stack of lenses that I was talking about earlier. That essentially, you know, we've picked in the West a couple of lenses through which to see the world. Computational thinking, mathematics, binary computation is one of those lenses. The scientific method is another of those lenses. They're very powerful lenses. They've done us a huge amount of good. But when you only see through one possible lens, you are missing out on so much of the world. And this is so obvious from like all of the critiques of kind of technological culture, which we're all very familiar with. You know, the fact that when you try and put something into a database, right, you squash it in all these kind of bad, gnarly ways and things that don't get fit.

things that don't fit into those little boxes, they either get forgotten or they get like actively repressed, right? You get these kind of decisions made many times unconsciously because things just don't fit the schema, right? And so, so much of the world is left out of the models that we make of it. And that's how models work, right? Models are reductions of the world. And when you want to do something specific with that model, that's very useful.

But you must never mistake the model for the world. You must never think that the world is what science describes it as alone. The scientific method is one way of describing the world. And you can do this in so many multiple interesting ways. One of my absolute favorite scientists is Monica Galliano, who I write about in the book, and he's written wonderfully about her own work herself. And one of the things she's very famous for is her work on plant memory, which is just a fascinating concept. The idea that plants can remember. And she designed a series of very specific tests, very carefully controlled experiments that...

where she put these mimosa plants. Mimosa, those are almost like fern-like plants that when you touch them, they curl up like this almost instantly. They're one of the very few plants that kind of respond in human time that we can see a physical response happen to a stimulus. So they're really good partners in experiments. So she got a bunch of these mimosa plants and she put them on a kind of rail that dropped them like 10 centimeters onto a pad. And they'd curl up. And then she'd wait for them to open and then she'd do it again. They'd curl up. And she noticed that after a few times, they stopped curling up. And she's like, that's interesting. And then she put them aside, tested some other ones, same responses. What she discovered is she tested the same plants weeks later.

they still didn't curl up. If he did something else, if he touched them, they'd curl up, right? So they were responding to a specific stimulus and they'd learned that this stimulus was not harmful. It didn't require them to use up the energy it's required to curl up. So they didn't need to do it. So plants learn and remember.

And this is just unbelievably fascinating on a bunch of levels, not least because we have no idea how they do it. We have no idea what are the mechanisms. One of the problems of botany is that we tend to treat, historically, plants as machines. In order to understand them, they'd be cut up into ever smaller pieces until you get down to the cells and various structures within them, and you'd find the mechanism in there, the little solenoid or transistor analogue or whatever it was that produced this response. We don't know where memory resides in the plant, and yet the plant has memory. One of the interesting things that Galliant does is she treats plants as individual beings with sets of behaviours rather than as tiny machines. That way of seeing them from the first...

not how most botanists see plants, allowed her to see this other quality in plants and therefore designed an experiment that showed this thing. Now, it took her quite a long time to publish these experiments, basically because they were so wild. And there's an inherent conservatism in science and scientific journals, which is good because we want to make sure we're, you know, ticking all the boxes, et cetera. But this stuff goes through peer review. It was checked. It was repeatable. Eventually it found its way to journals. But the problem was really like, which journal, right? Because there's no journal of plant memory, right? Because that's not how we conceive of plants.

So the kind of institutional structures of how we understood plants also didn't allow for this new knowledge to easily spread because our structures of knowledge are so fixed. And the other thing that really troubled a lot of scientists is that Galleon is very public about having a shamanic practice, right? She talks to plants. She engages in rituals and goes on what are called dieta and various other things. You can read her accounts of it far better than mine. But short version, she talks to the plants and the plants helped her design the experiments. They were collaborators in her experiments in every sense of that term.

That really puts a lot of scientists' backs up because that's not science. And yet what's brilliant and beautiful about her work is it's peer-reviewed, it's reproducible, it's in the journals, it is science. So it doesn't matter if you don't believe the other stuff. But you'd be kind of crazy, I feel, not to like maybe think there's a couple of lenses going on here. And maybe there are different ways of approaching the world and different ways of meeting the mimosa spirit and different ways of seeing what all the strange beings around us do all the time. And we can choose how we encounter and meet them. And we can be generous with our kind of ways of thinking and understanding and understanding that there is no right way of thinking the world. No like single viewpoint on the world that has like puny embodied highly evolved apes allows us to magically have total knowledge of the world. Like we just, we have to keep building all these weird models in different ways and some of them will teach us something. And apparently one of the models that we're building at the moment is this thing that we call AI.

which at heart is a series of different models of the world that we poke in various ways to see what it does. And no one really knows what it does or what it will do. And no one really knows how to construct that model. We've got a lot of theories. We've tried a bunch of different approaches and the current like LLM thing is like our best shot at it yet. And I could say more about why I think that's interesting. But like everything else, it is just one model of the world and you can come at it from any of these directions. Yeah. I don't know why, but I got stuck on...

As I was reading your book, I got stuck on this concept of rock and silicon, going back to its core components that we do recognize, right? Not the ones we don't. And just imagining, we're all talking to a rock. We're sitting there, we're talking to a rock. But it's talking back. And we're like, wow, this is amazing. And how much that's changing our impression that it's still a rock. Now it's not a rock. It's something we're imagining. And then I go to that, I don't know why, it took me to the movie Castaway where he's talking to a volleyball because he's got nothing else to talk to.

And was there anything wrong with him talking to a volleyball? No. I mean, in that circumstance, that volleyball was very therapeutic for him. Maybe that's what got him through it. Maybe he didn't go crazy because he had a volleyball with a smiley face on it to talk to. And how far is a volleyball from a rock? Just somewhat.

more processed than a computer um but at the end of the day we're all talking to a rock and then asking ourselves like is this rock is this rock going to start controlling the environment or are we going to continue to control the environment like are they going to take over and control our environment for us and we're going to become the one who is subject to all the change or will we still be in charge um and and the question is well as long as it's a rock it's probably safe but i don't know it took me down this this this weird road of like Somebody was saying they had a debate with an LLM, and this was a pretty significant author, and he was surprised at how well the LLM debated him. And I kind of in that moment was like, are you impressed with the LLM? He's like, yeah. I was like, I think you should be less impressed with the LLM and think of it more as a reflection of how poor your argument was.

Because if an LM can argue well with you, then your argument is pretty well-known and there are pretty well-known rebuttals to it. So is it a reflection of how smart The Rock is or just more how... unoriginal your argument is um i don't know anyway i think i just took us down i think i'm just sorry i'm just thinking about like the tom hanks cluster which i think i've ever actually seen but i'm fairly familiar with the setup but i don't know what else he does in the movie but there is something deeply sad about a man finding himself completely alone in nature and only being able to converse with like the one other manufactured product from like so-called civilization that's reached in this place like i'm sure there must be some birds or some fish or like something because also when you mentioned earlier about um you know there's this development of machine learning systems to try and decode animal language and so on and so forth, which is really fascinating and might well yield some really amazing stuff. I was reminded of a possibly apocryphal story that someone told me a while back, which was when someone, an indigenous scientist, went to a kind of conference for machine learning and these kind of other language translations. And they said, look, we're announcing this huge prize, right, for the first team to be able to speak with a non-human animal using their language. And the guy just sort of put up his hand and was like...

Sorry, but my people have spoken with non-humans forever. We are in daily conversation with all the other living beings around us. We have shared language. We speak to them in word and in dream. We can talk to the animals. And you don't even need to go into shamanic practices over this. You can look at the honey guide, honey hunters in Africa who share words with birds and go and call out the birds from the trees. And the birds call to them using specific sound calls. We can already communicate with animals. So A, they renamed the prize or we wrote the prize. The first people to build an AI that can talk to the animals. Okay, that's interesting.

We're starting to build, like, these weird multiple relationships where humans are only allowed to build machines to talk to the animals. And why do we put machines in this place rather than thinking we can have direct communication with them? Which I will return to in just a second, Rob, as a possible, you know, answer to your thing. But, like, first of all, it's like, yeah, is this... Oh, I wanted to say about your friend who debated. That's the other thing. Can I move my computer? Look, I don't debate LLMs. But what I have been doing is I've been trying to help them build me chairs. Turns out...

well, they're really bad at chairs. So I had a bunch of scrap wood lying around and I asked, this one was, this is the ChatGPT chair. So I told it I had this bunch of wood and I asked it to build me a chair and this is what it came up with. Which to be fair is like chair shape and does hold your butt off the ground. It's weirdly big.

My three-year-old kid really loves it because it's basically a high chair. And actually, I'm like six foot and it's actually quite a good size for me. It makes me realize all of the chairs are quite small. But it was really... It was sort of touch and go whether it would work. You can read that on my website if you're really interested. But my point is, like, this guy was an author, right? Or a writer you said? And, you know, into debate. So maybe that was his thing.

The debating was like such a high standard for this thing to reach. And if that was what the LLM was specifically designed for, then great. And that's kind of what they are designed for, right? I mean, they're designed for conversation. As soon as you put them to other tasks, they kind of fail. Or they do interesting things. Like working together, we built a chair, but it produced a lot of weirdness and we had to work together to get there.

And I went into that relationship wanting to be in it together with the AI as something where we would use our respective skills. I would use my very limited carpentry skills. And the LLM would use its skills at measurement and efficiency maximization to use a bunch of scrap wood in the most efficient way, which was quite good. So we're trying to find these points at which we actually had a set of shared qualities that were kind of more useful or that kind of complemented each other. And that comes back to this thing about to speak to non-humans or to speak to to anyone else even other humans like through or with or alongside these things right that the fears that Rob mentioned about them kind of replacing this or overtaking us is because we have this idea that this thing is like above us in some way or might potentially become us as we are above all the other species and creatures we like to think right because we're the pinnacle because we're so like enculturated socialized into like obsessed with these notions of hierarchy that put like one way of knowing one way of being one body type gender sexuality race species like um carbon basis as like you know into some kind of weird hierarchy and it's just it's ah it's just not

to me it's not how the world works and there's like so many more interesting things to be done with so many of these relationships because relationships is what they are um if you if you can first of all identify where these like hierarchical barriers pop into things and then start to kind of dissuade yourself from them and try and root around them in certain ways and you know a good example is your assumption or the assumption of a bunch of scientists that either we can't talk to or communicate with let's say better non-human species and secondly that that is only performed through language right which is a real human obsession. Many other species have language, to be clear, including probably whales, fairly obviously whales. They also have a whole bunch of other things. And we're quite capable of communicating pretty effectively with other humans who don't speak the same language. So other forms of communication are also possible. But when you reduce it down to what an LLM is capable of, just as when you reduce an LLM, like AI, down to being good at debating because LLM is good at language, you're back in that tiny, tiny tunnel of what's not really interesting. Yeah, you're really hitting on something here. It did happen when I was reading the book, but it's being reprinted in a more significant way right now, which is...

So much of our perception of what we understand as a layman is based on these misunderstandings. As in like Darwin, that the fittest doesn't mean the one that fits into its environment best, but that dominates it best, which is incorrect. And this idea that if you look at species, that there's a top of the food chain and a bottom of the food chain, that it's linear instead of circular. And this is not the way we understand biology. This is not the way we understand these things. It's the way that it's been misunderstood. But as long as we see it as linear, as long as we see there being a top and a bottom, then we're trying to figure out...

where to put llms where do they fit in this linear thing and you're like well it's not a it's not linear it's not top and bottom it's circular it's much easier to understand that they're going to fit and not fit depending on what circle you're looking at the circle of what um and and this these misunderstandings have caused so much this idea that we're superior that that a tree will take over the whole forest if given enough resources both the tree has mycelium it has all this connectivity to right size itself to fit into its environment um in the right way and that's and that intelligence in itself is not hierarchy it's not the most intelligent species that's that's the most uh that's at the top of a food chain that there is no food chain but the idea that there might be a right size intelligence and that too smart actually goes extinct and that you look at the species that have been here the longest the organisms that have been they're not the smartest so maybe too smart doesn't make you at the top it it actually if you're measuring it from the longevity of how long a species will survive on earth maybe we're too smart

There's not really any such thing as smartness or even really intelligence as something that can be measured in a meaningful way that can be categorized or put into a hierarchy in that. There's like abilities that matter in certain situations. And, you know, humans possess a pretty good bunch of them. And in a bunch of other situations, we kind of fail completely. And one of the ways in which we obviously fail completely is in our ability to moderate ourselves for like the health of the global whole, the health of the entire ecosystem. And so you have to ask yourself, like, why? Like, what are the things that have crept into our evolutionary history?

that have, that bring us to this. And this is not, sorry, that's not also to get into some kind of evolutionary psychology argument about that. I think this is much more cultural. But, you know, and also knowing that, what are the, what aspects of that are being reinforced by the way we describe things in the present moment, you know? I just seem to repeat something that I sort of say, I think quite frequently in the book, which is that...

You know, AI is, if there's something interesting about AI, it's very simply the fact that there is such a thing as non-human intelligence. Let's bring it back to that, right? The more time you spend with these strange things, even these very rudimentary pieces of software that don't, we can argue endlessly about whether they're intelligent or whatever, let's not do that. Let's just say, here's a thing that acts in the world in an interesting way and that I can have a relationship with and presents itself as being intelligent to me. But crucially, thinks about the world in a very different way, right? That sort of has a different sense of number and dimension and knowledge itself and quite simply thinks in a different way. I mean, there's a...

One of the examples of a really interesting use of machine learning that I go back to quite a lot is... Should I do the chess one or should I do the nuclear fusion one? The nuclear fusion one, which is the Tri-Alpha Google experiment. Tri-Alpha being a kind of fusion startup. I don't know where they're at now. They were doing quite well a few years ago. They worked with Google to produce this, what came to be known as the optometrist algorithm. And the thing is, when you're trying to make fusion...

you know, in a lab where you're trying to develop a process for making fusion actually work. You've got kind of working theory of how it should, but you've got like all these lasers, all these mirrors, like a lot of, like a lot of energy flowing around and a huge number of different parameters that can be kind of set in each experimental shot, right? Way too many for a human to, to like really get their heads around. But like a machine, no problem, like laser parameters, fine, it's just numbers, right? But if you imagine like a problem space, right? Like this problem space is composed of all the different possible settings of this machine. It's a landscape, right?

And there are different ways of exploring that landscape. Something resembling a solution might be somewhere in that landscape. We know it's feasibly possible. What's the best way of getting there? Well...

The way that, you know, machine will get there, there's many different algorithms for doing it, but we'll do it in some kind of fairly methodical fashion, right? Left, you know, basic assumptions. And it will kind of step through or move around in this quite sort of methodical place. Humans won't. Humans will like, you know, humans have hunches, they have feelings. That's not, again, to prioritize them both machines. It's just a different way of thinking. You get a sense of something, an itch, and they'll move kind of in this direction. You will get a kind of problem solving pathway that's kind of more of a wonder, more of an exploration. What happens if you...

They work together, right? But these two different ways of thinking the world work together to explore this problem space differently. Like me and Chatty Petty exploring the space of chairs. You come up with some weird stuff, but you also go to places that you wouldn't have gone to otherwise. You perhaps get there faster or you go closer to the kind of edge of a problem. Humans don't tend to go towards the extremes, but maybe the answer is out there. The machine has no such qualms. It will push the human in that direction. The machine maybe is just methodically chunking here and the human is like, let's go that way. And what happens is you get like a different approach to the problems because you have two different minds at work. And I could talk a lot more about what happens when you have multiple different types of minds at work. But the thing is that like machine thinking is different to human thing.

And if there is more than one way of thinking, there are infinite ways of thinking. Right. And so what is interesting to me about AI is that just at this moment that we were all obsessed with this like different way of thinking, we're starting to notice that we are surrounded on this, you know, European, we, by all of these different ways of thinking that see the world in so many different ways that have been here all along, many of them from long before we were here, that have all of these different approaches to the world that we have something to learn from. And I'm fascinated why we seem to need to construct these little toy machines in order to come to that realization. That's another conversation perhaps, but it's quite a realization. And it's something that, you know, that I, that gives me, that for me is what a relationship with AI gives me is access.

not even to a particular way of thinking, but just another way of thinking. And my desire is for more ways of thinking, not better ways of thinking. Again, not to hierarchize, but to hybridize these things. Right. That's exactly what happened when I read the book. It opened my eyes to the fact that I think what you're sort of saying is we've muted understanding our relationship with our environment. For whatever reasons, we've muted a lot of that. And that our relationship with things, with living things,

inert things who cares um it's something worthy of understanding because it it it literally is at the heart of our the environment that is going to like you said we'll create our tools but then our tools become our environment and then shape us as mcclellan would say it's not even just interesting to understand it's like vital to understand because our current relationship with other beings is going to kill us and quite a lot of that so yeah it's a critical understanding in order to come to yeah and and that it's it's not just understanding the thing which is what we focus on but it's understanding the relationship between the thing and us and each other that's It seems to be like this blind spot for us. And LLM's are just another thing to understand a relationship. But as your point, we become suddenly obsessed with our relationship with that one thing. Why that one thing? Why not everything? Why not plants? Why not? Those are just as interesting. And now, does this open us up?

Is that a good thing that it opens us up to understanding that at a core level and starting to look at, like, yeah. It would be a good thing if it happens. I don't think it's a medical product of doing it. Like, it requires conscious thought and attention in order to kind of interstitial awareness and thinking towards that position. That just happens to be the one that I'm proselytizing as kind of like this kind of interesting direction to go. But also, it's not without precedent. Like, we have a history of kind of looping back and...

We build the way that we think of the world all the time, right? You go back to, you know, the late 19th century and everything like hydraulics and pneumatics. And you also get like early chemistry and physics that theorizes the world as the universe, you know, at least as like vapors and kind of phantasm of the air. You get like a pre-germ theories of disease where it's all just these kind of fogs and myths moving around. And you get these kind of pneumatic model of the world that's both a lens, a scientific way of thinking the world. And it's also like the science of relationships of technologies and all these kinds of things.

Um, you know, we, we, we move through these different ways of thinking the world and that even happens in smaller cycles within different technologies. Um, you know, in the, um, uh, the, you mentioned the mycelium in the forest, you know, um, when, um, uh, Dr. Suzanne Simard and others in the, in the late nineties, started publishing a lot of their work on, um. uh mycelium in forests uh the way that these interconnected fungal symbiotes um allow trees and other plants to share not just nutrients but also information amongst themselves that these are communication networks nutrient carrying networks um these kind of incredible yeah information sharing networks among many other things um you know when they first published that research in the special issue of nature um on the cover they they they coined the term the wood wide web this lovely phrase that's where they came from um you know and yeah and the thing is that um The mycelium networks are not the same as the internet, right? It's not servers and cables. It's much more the living threads that are also information bearing and becoming machines, all kinds of stuff. But the scientists knew that people had a mental model of the internet in their head at this point, right? They knew that if they wanted to communicate the real wonder of these mycelial networks that we were coming to understand that have been below our feet the entire time, they needed to kind of like...

you know, put in a good metaphor, right? To something that people had an existing mental model of, which was the internet. And that wasn't even the, like, but where did that come from, right? Because actually... Like those scientists, many scientists in many fields were also the first people to be connected to the internet in the kind of decades preceding that, right? In the 1780s, it was scientific institutions, it was universities that were first connected to the internet. So they were some of the first people to be given that mental model, networks that didn't look like this, networks that look like this, sorry, like weren't conceptualized before the internet, which grew in a very strange way that no one really planned and wasn't fully understood how it worked. They just had this growing notion of kind of interconnected distributed networks that formed the early internet.

And then scientists went out in the world, they found that those networks already existed all around us and they hadn't really seen them before. But in some way, I think, and some scientists have spoken about this, they were more capable of seeing them because again, we built these kind of models of them, just as perhaps we're more capable of seeing intelligence in some form out of the world because we built these little AI tools. And when, you know, they really started studying the internet, they really did discover that it's what's called a scale-free... Uh, network, the, the scale free is like a term from, from network topology to describe these kind of networks that have, you know, nodes of different sizes, different numbers of inputs and outputs that you can kind of swap around, you can drop things out and it will still be capable of growth and change. Those are adequate models of, um, ecological networks as well. Um, and, uh, those can be applied to natural networks like Mycelium as well. Not, not the actual.

working of the thing but it's topology the way that it connects the way that it works at the larger ecological scale and ecology itself is the science of relationships it's the science of content there's no like There's no such thing as ecology as its own thing. There's only an ecology of other things. And every discipline discovers its ecology in time. And this has happened throughout the 20th centuries. Various sciences, in particular, the hard sciences to start with, had these kind of ecological turns, where biologists and zoologists at the beginning, botanists died realizing, like, you can't just study a penguin or, like, a thistle, right, without considering the context in which it exists. Not just the context of its species, the context of its landscape, the context of its kind of food web, its watershed. All of these things matter.

And technology feels like the one field they're still struggling to discover its own ecology, to like account for itself within this larger sphere. It should. It has the grounds to do so because it's everywhere. It's spread through the world, as we've noted multiple times. It's built out of stone and silicon and the long dead bodies of creatures that died billions of years ago, right? Literally, physically, like everything else we made. So it's all for the world itself. So what does it mean to take that ecology seriously? So much of that, almost all of that thinking at present moment is intentionally captured by large corporations that do this stuff for money. And that's a shame because it's both fun and important to think about outside those contexts.

Yeah. Because no one else is really going to invent the weird AIs that are really going to teach us about the nature of the universe and what we're capable of and what consciousness means if they're doing it for, you know, to sell these spreadsheets. It's not how the world works. But they have a bit of a stranglehold on our imagination at present. They are the ones who are describing the thing that we'd like to talk about in more open terms. And what I'm trying to do is just like prise that conversation open a little bit. Yeah. I, yeah, I like that section in your book where you talked a little bit about the motivations of corporations and there's like a lack of alignment between, between that and, you know, just.

a sort of positive, I guess, outcome. You can put it that way. Yeah. It made me sort of drill into that, like, sort of the why of that to go like, well, what's the company and where does this motivation comes from? And, you know, shareholder value. And you realize it doesn't have, it's not the company itself that's the problem, it's its motivation and this motivation.

to serve something that doesn't belong to it. Like there's the employees and the customers that are part of the ecosystem that are part of that company. And then you have this sort of external source that is this primary beneficiary of it. And so the idea is this external source exploits the company and the employees and the customers. And you go, where does that come from? It comes from an idea that shareholder value is the number one objective of a company to drive shareholder value. But if we shifted it to employee value,

and customer value and shareholder value. Or ecological value. Yeah. Earth value. Does that shift it? Like, is it the company or just its value system? And is that original value, shareholder value, even a good one in the first place? Like, it's based on falsehoods anyway. Like, shareholder value is short-term thinking. And so, as you look at the longevity of companies, it's gotten shorter and shorter as we become more obsessed with shareholder value. Can we... Because I like the idea of company as a way to organize people towards something. I feel like that's...

If you just take that part, like let's organize people towards a common goal, it feels good. Unless the goal is wrong. Yeah. I don't know. Corporations are only one of many models for doing that. Right. But yeah, I mean, what you're talking about is, you know, well, the idea that I first encountered from, I think, Charles Stross, the science fiction writer, probably has bases elsewhere, of, you know, corporations as entities, or rather corporations specifically as AIs, right? That if you want to know what an AI actually looks like in another form to like an LLM, a corporation is a good example. A corporation is an artificial intelligence, right? It's a...

autonomous being essentially talk about this autonomy that has like sensors and effectors it can receive information it can send it out it has most of the like qualities required for like a life form um it also has a brand it has a brand identity it has a personality Yep. And it has legal rights. It has legal standing. Corporations are people under law. So corporations have legal standing. They have huge resources. I mean, if they're successful, they have money, they can spend it. They can choose how to do that. They can shift politics. They engage in politics. They lobby. Most large corporations have more political power than most people through lobbying. Personhood under the law is an incredibly powerful thing in our society. Not all people have personhood under the law. And certainly most animals don't. Other non-human entities don't.

ships, states do, and corporations, right? So these all can be sort of entities. And if you look at the landscape of the planet right now, like what dominates? Corporations. More so than nation states. So we already live in a world that's dominated by a form of artificial intelligence. And we know it's a form of artificial intelligence that does not have our particular best interests at heart. So hand-wringing about like whether open AI is going to like break the planet is really missing the point to my feeling that we already have these large intelligences stalking the planet, killing millions of people and poisoning the waters. So how much worse can it get? There are however many other forms of organization.

outside the corporation that can be doing this. And in doing so, create very different forms of, forms of assembly, forms of politics, forms of ways of doing, but even within the sphere of AI, I can give you nice examples, the indigenous AI movement that brings traditional ways of thinking about knowledge and relationship to the construction of AI systems. For example, there's amazing work being done in New Zealand by Maori groups who have built their own machine learning powered translation engines for, specialized in Maori languages. That also creates different forms of politics and community as a result, because the results of those, the common knowledge that is generated in the form of community knowledge, in the form of like transcripts of radio shows, the wealth of the culture, the translation of knowledge into and out of their language, that belongs to the community. That power stays within the community and empowers that community, both through the skills they learn, but through the cultural wealth that is generated and the propagation of that culture. And it comes from a different way of thinking about what we can make with these things if it's not controlled by a...

Particularly way of thinking, which in this case is this corporate way of thinking. The AI that we're talking about all the time is corporate intelligence. It's the intelligence of the corporation. And it's where so much of this stuff about hierarchy comes from. We know Silicon Valley and most of the dominant US tech culture emerged from World War II and the military complex. So it already had a lot of that kind of weaponized dominance built into it. And you layer capitalism on top of that and you get...

AI that just really wants to beat us at chess, right? That is the idea of what's the most intelligent thing is, is the thing that beats the human at something. Like what a pathetic idea of what is intelligent than something that just wins at something all the time. Whether it's chess or driving or making money, like even the most diehard top-hatted monopoly man, if you sit them down, must agree that that is not the only thing that counts as intelligence. And yet we spend a lot of time and money and a lot of cultural time, like debating whether like winning at things is the best form of intelligence.

That's so huge, right? It's being the wealthiest man in the town. Taking that to its full extent is where everyone has nothing but you. Who wants to live? You don't want to be anyone in that town. Everyone's got nothing but you. You won. You know, like, great. I mean, that's how a certain section of the... Yeah, quite a lot of people on the planet nevertheless behave that way. Right. I don't know if you've seen the Adam Curtis docuseries, Can't Get You Out of My Head.

But he spends a lot of time kind of exploring, I guess, a lot of like the classification that went on during the 20th century, kind of at the behest of imperialism and colonization. And that really is just sort of pointing out how separate everything is to this notion of a self and how that kind of gave rise to individualism. And that sort of makes me think that, like, I guess even going back to the poetry thing I brought up earlier, like in some ways, in a very immediate sense, these AI technologies kind of force you to confront, like as a writer, right? It could write, like, it makes me feel less special, less like an individual. And I think that could even connect to sort of like the corporate existential crisis that's looming. But there's also sort of this, it's sort of like, if it really can kind of point out things that we've been ignoring all along and to a greater...

a sense of connection between everything that's kind of floating around on the planet, bumping into each other, it kind of threatens the whole idea of how life is to a Western mind. Like we're conditioned to kind of think of things in these little boxes where they make sense. Whereas like we were talking about indigenous cultures, like that is often more of like a circular view of life and the world. So I wonder if, I don't know if there's a question in there. Do you get the sense? No, I mean, it's really worth pointing out that actually it's not even Western. Like it's just like a very, it's a popular, I guess. And I don't mean as in, I just mean like the most common view.

Like, most of us still live in a Newtonian model of the world, right? Most of us still understand the world on a day-to-day basis with our minds, right? As a Newtonian universe, right? It's over 100 years since Einstein completely blew that apart. Einsteinian physics is really hard to understand. The quantum world is really hard to understand. But it's not impossible. I don't want to do that. I had the privilege a few years ago of listening to Karen Barad speak. If you're not familiar with Karen Barad's work, I highly advise you to check it out. She's a poet of physics and much else. I had the privilege of being in a room where she explained quantum physics through speech. She told us about it. And for like 20 minutes or so, I was...

like it was there it was hanging in the room above us and then it like when she stopped speaking it just it's hard to understand this universe and yet this is the universe we live in and it's not just physics either like biology gave up on the notion of the individual almost 20 years ago there's a very famous editorial published in i think it was in nature by like several of the leading editors like serious heavyweight scientists saying like we can't really hold this thing called the individual together anymore it doesn't really make sense right because like we now know that like We carry around two kilograms of other beings within and on the skin of our own bodies. And when we remove those, we die, right? So how are we individuals if we require the constant multifarious presence of all these other beings in and on us? If you change your gut, makeup, your intelligence changes. I don't really buy IQ tests, but it's possible to measure when something's up with your brain. And when you mess with these organisms, the way you think changes. Biologists are trying to work out, how do we talk?

about bodies about beings um except as this set of relationships that we have right and and so we live for now until the next Einstein or Niels Bohr in a quantum universe. Sorry, in the Newtonian universe. We live in the quantum universe. We still think of it as the Newtonian universe, right? And we still live in this like 19th century Darwinian universe, even though people like Lynn Margulis and many, many other scientists have spoken for a very long time about how much more complicated, just biologically, physically the world is than those understandings, right? So it's not just...

It's not science. It's not us. It's just like the speed of our attention and awareness, the social structures, family structures, everything that we're acculturated into that takes a very, very, very long time to shift at the cultural level. Of course, you can go and sit under a tree and have a conversation with a tree now, and the tree will speak to you if you listen. I'm entirely serious about that. And you can change your own mind like that.

So these things move at different speeds and in different ways. The universe is immensely strange. And we have a choice over which lenses we use at any one point to have our own relationships to it. And that is, for me, the thing to just always be kept alive in all of these conversations. Yeah, and I think there's an element to this sort of thread alongside, by the way, literally, and that is we are...

very addicted to stories. You know, the way we want to understand the world is through stories. And when someone comes up with a great story about mycelium and the internet, we all jump on and it just, we tell each other and the stories kind of, you know, go viral and we learn collectively. When it's a boring story or too complicated a story, you know, it just sits, doesn't get known. And I think that's, it's interesting, like to your point way earlier, which is like we choose, we pick and choose what what to learn, what to take in. And that shapes our understanding of it. Like, why do we pick the things we pick? Because we like the story? Or because, you know, it serves us? I don't know. I think I got lost there. But the idea that AI...

And the conversation of, you know, what is happening? Are these tools for corporations? And the corporation itself is the problem. And now they have more powerful tools. And that's what we should be scared of, the corporation itself with better tools. I ended up going, as you were talking in my mind, to this place of is anthropomorphizing something else or anthropomorphizing something. And then considering it more important than ourselves, the real problem, when we think, sacrifice yourself for the company. Sacrifice yourself.

for this entity that's not real. Is that the real danger? Whether that something else is a machine or an AI, like you said, a corporation. You get what you want. If you're compromising yourself for something non-living or something else, then it's the something else that's not going to compromise for you. Is the company going to compromise for you back? Is it its symbiotic relationship or is it asymmetrical as you pointed out? And that seems to be what the company's objective is not symbiotic. It's asynchronous. You serve the company and the company doesn't care.

It would be lovely to think about what a truly symbiotic assembly would look like. And again, I would argue it simply wouldn't look like a corporation. There's plenty of forms of community. There's plenty of ways of going about making change in the world that do not conform to the ideas of corporations. And they exist within human society and they exist within non-human society in all of these kinds of ways. In the book, I write about the examples as diverse as the decision-making processes of honeybees, the way in which hives of honeybees perform these incredibly complex...

Consensus. Political. things in the form of dance, right? They do the little, what's called the waggle dance, and the bees dance their vote in these complex political assemblages in order to make choices collectively. It's very similar to the way certain political systems work. It was similar to early Quaker political systems in New England, but it's found in the kind of Swiss cantonments. It was found in Native American Iroquois Confederacy and various other forms, forms of direct participatory democracy that are found within human society and within non-human society that are not mirrored within the larger structures of our national democracies at present.

We have so much to learn from these other structures. And so we can diagnose much of what's the problem with certain current arrangements, the form that most of our contemporary industries and therefore the way we make and think about things or think about them and then make them is essentially extractive and reinforces existing hierarchies, increases inequality and prevents the kind of the real efflorescence that might occur through like a broader diversity inequality. We can diagnose that. We can see it. We know it's the case. And where we look for, you know, the question then has to be for me, how do we change? Where do we look for models of that change? Other forms of community, human and non-human? And how do we bring as many of those into dialogue with one another as possible? Because as I say, the thing about the hierarchical thing is it means the levels aren't really talking to each other. You have the humans making the AIs, worried that the AIs are going to supersede them, talking down to the people below them all the time while throwing all the other species, like, out of the window. But the examples that I continually return to are the ones where of mutual.

aid, like working together with an AI to build a chair, right? Or in order to build a community to see what happens when different things work together rather than thinking of everything in terms of hierarchies that leads to extraction and inequality. There's for me a very clear roadmap there that's possible for us to follow. Yeah. As you evaluate other models, how much does decentralization and centralization do you think play into that evaluation criteria? Like how, is it a factor or is it just a characteristic?

What centralization means is you privilege, ultimately, one way of being and knowing the world over all other forms. Because power accrues to one thing, but the way of knowledge that's accepted flows down from there. It becomes the only acceptable way of gaining knowledge and power. That's what centralization means. And it leads to, like, a world full of inequality and, in fact, boring one, but also one that ultimately doesn't function in ways that are interesting. Sorry, it's such a silly example, but it's the one that I'm thinking of today, which is that someone, I saw someone on the internet talking about the fact that, I think it was there in San Francisco, that a stoplight had gone out. And so the Waymo taxi in front of them stopped, right? Because it didn't know where to go. And it's a funny thing in that context, and we have a million examples of that. But, like, where I live, like, there's at least three stoplights I pass every day that have been out for months. And, like...

The possibility of even conceiving of a driverless car, the robot car, another of these great science fiction imaginings, just doesn't apply here, right? And it's not that different here. I live in a European country. We have mostly the same road signs and stuff, but it's just not going to work. And it's so silly and boring to imagine a world that has centers in that way, right? That produce all of this fluff and noise and power that affects everyone else and doesn't work.

in 99% of the rest of the places because it's repeating the pattern that humans have also been doing for like a good couple of thousand years now of centering the human at the center of the world and thinking a world that works for humans that's where everything else is tool or resource like makes the world better and it demonstrably doesn't right it's bad for the world which is bad for us right to put us or anything else at the center is whether it's a way of thinking or just a model of any form to elevate in that way like ultimately fails. And we live constantly in these various states of failure. We're living in a massive state of failure now. We're almost always living in a state of failure of like the particular model that we're in. Paul Virilio, one of my favorite philosophers, said, you know, the invention of the ship was the invention of the shipwreck. We know like this is what happens in the afterlife of the thing. What we have to think about is like, okay, cool. That's what's going on. Where do we go out and find a new, weird, strange thing that changes that set of relationships that moves away from this thing in like a deeply radical way? Not just, you know, the new hype thing, but like the new way of thinking that we can learn from.

pay attention to and ultimately change with awesome yeah well this was a great conversation we could keep going as we've demonstrated all day um but yeah i really enjoyed here i'm afraid and uh thank you and i think that was that was a great conversation i really appreciate you coming on yeah thanks for spending time thanks very much indeed for having me It's been a pleasure. All right. Well, thanks for hanging out with us again here on Invisible Machines. Don't forget to like and subscribe wherever you get your podcasts. Always remember that you can watch this podcast, too, on our Invisible Machines YouTube channel. Head over to uxmag.com. We have a podcast page there where you can explore our back catalog and you can look at episodes centered around popular topics. Many, many thanks to our incredible production team here in the U.S. and in Ukraine. Elias Parker, Michael Lipinov, Allison Harshberger, Ola Slessarenko, Vera Pakhodkov, and Dorena Moskovchuk. Till next week. Awaki.

Frequently asked questions

How long is this episode of Invisible Machines podcast by UX Magazine?

This episode is 1 hour and 13 minutes long.

When was this Invisible Machines podcast by UX Magazine episode published?

This episode was published on August 15, 2024.

Is there a transcript available for this episode?

Yes, a full transcript is available for this episode. You can read the complete transcript on the episode page.

Can I download this Invisible Machines podcast by UX Magazine episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.