The Spatial Web and Active Inference A.I. with Denise Holt

A Transition Mindset: On Verses, Active Inference Intelligent Agents & the Spatial Web Tech entrepreneur and content curator Denise Holt speaks with Andrea about Denise's support of Verses AI, following the development of the Spatial Web. Denise recounts her journey into technology, most recently influenced by her friendship with Dan Mapes, one of Verses' founders, and the company's collaboration with Karl Friston to advance AI through active inference. She discusses the significance of spatial web technologies, including HSML (Hyperspace Modeling Language) and HSTP (Hyperspace Transaction Protocol), and their potential to revolutionize how we interact with digital and physical spaces by enabling intelligent agents to operate within our internet, expanding it into a 3D digital twin realm. The conversation explores the implications for privacy, data ownership, and the integration of diverse technologies, projecting a shift towards a more natural and immersive internet experience. Denise also touches on her personal motivations, aspirations, and the optimism she holds for the future, shaped by technology's ability to bring about societal improvements. The discussion concludes with Denise outlining her efforts to educate and build community around these emerging technologies through her Substack and live learning sessions.

Sign up at substack.com/@deniseholt: https://deniseholt.substack.com/

Transcript:

A Transition Mindset: Denise Holt on Active Inference Intelligent Agents, Verses & the Spatial Web

Denise Holt: [00:00:00] The space itself becomes a domain. And the modeling language is what's used to program those spaces and it can program all kinds of things. The ontology for the programming language, it takes into account activities. It takes into account credentials.

It takes into account, All kinds of aspects, multidimensional aspects about those things and what's occurring with them over time in that space and with whom and who has access who, all kinds of things like that, like if it can be measured. In a physical or digital reality, it could be programmed using the hyperspace modeling language.

 Hi, Denise. I'm so glad you're here today

Denise Holt: hi, Andrea. Thank you so much for having me. I'm excited to be here.

Andrea Hiott: I am really excited about your work, I have to say. Sort of found it through my collaborations with [00:01:00] active inference groups. And I really love what you're doing.

The podcast, the Substack, so, I really want to know how you got here. How did you get into this?

Denise Holt: Well, Verses AI they're leading this whole, effort with the active inference methodology being used for AI. They've got Karl Friston as their chief scientist.

Andrea Hiott: Not a bad chief scientist, I must say.

Denise Holt: Right. So, um, one of the original founders, Dan Mapes, um, is a long time friend of mine. And so I've known that he, when he and Gabriel Renee started versus, back in like 2017 ish, um, I knew about it, I knew what was happening. So I've been kind of following along in the shadows watching because I knew what the vision was.

I knew what, and I've let it unfold and grow and expand and, Um, they wrote the book, the Spatial Web back in, in 2018, I think it actually was published early 2019, but, that laid out this whole vision. And [00:02:00] then they started, um, collaborating with Carl and, it, it all just expanded from there.

So, about two years ago Talking with Dan, um, we hadn't been in contact for a little while and, we were catching up and he was telling me that Versus had just gone public up on the Neo exchange up in Canada and that the platform was going to be launching to the public within like a year and a half ish, that was the projected projection of time.

And I was like, Oh, okay. There's a transition mindset here. People know about this. At the time I was really heavily working in the web three space and technology for, my adult life

Andrea Hiott: I've

Denise Holt: always been kind of a tech nerd. And like when, um, Back in like 19, I want to say like 1998 ish, I was using Photoshop and I started really getting into these 3d graphics programs back then, like Ray Dream Studio and [00:03:00] Bryce and all these things. But rendering time was.

So that was actually one of the big motivators that really got me diving into technology was, I was using these platforms.

Programs on more of a creative side, but to be able to actually make them useful with a home computer, I just started building my own computer. I think

Andrea Hiott: that's what I'm trying to figure out about you a little bit what's the driving motivation, because when I listened to your podcast, the spatial podcast, I think. It feels to me that something is motivating you more than just that you want to, be a podcaster

Denise Holt: yeah, so no, and I have a lot of, so what I've done, I mean, I have my podcast, but the way I've done all of my writing around this and, um, the podcast and,

Andrea Hiott: Yeah, you're a writer first, we should say that.

Denise Holt: I write a lot better than I speak. Well,

Andrea Hiott: you clarify things very well. We'll get into it a bit, but this, Deep learning versus spatial web active inference, um, AI. I mean, I think some of your writings on that clarify it really [00:04:00] well, but we'll get into that, but anyway, yeah.

Denise Holt: Yeah. So what I've been, what I've been building is a media vehicle around all of this. And it's because I have a bigger vision with it, , and a lot of it has to do with the technology that's coming out with versus in their platform. I really see that when their platform comes out and it enables anybody to be able to build one of these intelligent agents, these active inference, intelligent agents, um, within our internet because that's what's happening with their platform. Our internet is expanding into this 3D digital twin space. So it's going from the worldwide web to include all entities and. The programming language makes context into all the things in the spaces and, relates them to activities and it's all credentialed and permissioned and everything that informs the ai that coupled with the internet of things, sensors, everything that's gonna give these active inference agents [00:05:00] the real time information to learn from their perspective within the network.

All of these agents are going to be aware of each other. They're going to be aware of the network. It's going to, I feel like we're about to see this explosion of development. Yeah, it's

Andrea Hiott: almost, I mean, it's we,

we should probably back up a little bit because, when you say the vehicle, you want it to be a vehicle.

Do you mean sort of a place where people can come and start at as a beginner or any place

Denise Holt: yeah, and so at this point, yes, education, I feel like is the most important aspect because, there's a lot of people that are building amazing technologies. And one of the biggest problems, especially with the web three technologies or the extended reality technologies is that they're all just disparate technologies.

It's really hard to, um, bridge that interoperability gap. And that's what The spatial web is going to do. So it's going to do that. And then also empower everything with AI. So to me, it's [00:06:00] exciting because all these people that are building these amazing, um, products within those spaces, I feel like they need to know this is coming because it's going to solve a lot of their problems.

It's going to expand their capabilities. And, I feel like it's going to usher in this, um, kind of augmented reality. Mixed reality existence that we've all kind of envisioned, but the foundation hasn't been there and this is bringing.

Andrea Hiott: So like, let's try it for people who maybe even haven't ever heard of spatial, the spatial web yet, or web 3.

0 even because there's a wide variety of different people who are going to probably watch this, but, so what do you think of as 3. 0 or web three and web 3. 0, maybe you can help? me and others understand what that means to you.

Denise Holt: So that's a really important distinction. So, um, to me, web three refers to the technologies themselves, right?

The, um,

Andrea Hiott: so you mean like augmented, blockchain, crypto,

Denise Holt: [00:07:00] blockchain, crypto augmented reality, virtual reality, internet of things, um, all of these. Technologies that, kind of enhanced that web 3. 0 is simply the next evolution of our internet protocol. So, um, the worldwide web that's web two, mobile with the worldwide web was 2.

5. And, we are about to. Evolve that into web 3. 0, which is also called the spatial web. And what's interesting is versus actually created the protocol, but they donated it to the public because nobody could own the internet. It's just that the protocol had to evolve into, into this 3d digital twin kind of space to where, like it had to become a spatial protocol.

It had to take us away from a library of pages and documents like the worldwide web. Um, it had to take us away from, [00:08:00] the concept of the internet has always been this decentralized vision, but it's impossible with the worldwide web because all the transactions take place. Under the umbrella of whoever owns the domain and the domain is the website.

So it's a centralized transactional process, right? So you don't, it's self sovereign identity is impossible. Um, security is really difficult. The worldwide web is the most. unsecured environment to even try to be using all these technologies in. Because, um, the only permissions you, the only permissions you're granting within the World Wide Web is, do you agree that they can have all your information and data or not?

There's no nuance to it at all.

Andrea Hiott: How long will you pay for whatever service or?

Denise Holt: Right, right. So, um, What's really interesting with the spatial web protocol, which is HSTP hyperspace transaction protocol and HSML hyperspace [00:09:00] modeling language. Um, so Verses they developed the protocol about three and a half years ago, donated it to the public, to the IEEE, which is the largest core standards body in the world.

They develop core standards for. anything that's like a public good as far as a technology that's going to be used globally.

Andrea Hiott: That's wonderful. What is the protocol though for people who have no idea what we're talking about? What's like, maybe we should even, let's just think like, let's think of a little narrative of, so we have, we'll skip the invention of the internet and so on up to kind of the fact that it's a part of all of our lives and we're all using it.

We're all embedded in it. We all sort of take it for granted that this is. What happens like this is just the way this kind of technology works that you go to a two dimensional or whatever, depending on how you're looking at you go to a page, and that's what the web is. It's, that's HTML and all these different coding languages are writing these things that end up just being pages that you visit.[00:10:00]

Um, and there's always a medium, someone owns the service usually, and you have to kind of go through them. So that's kind of what you're saying is, you're basically handing over your data and rights in this way. And we've just assumed this is the way it is, right? This is. Yeah. How it goes. Yeah. And what you just said is that, and I think I've heard or seen it somewhere, maybe you wrote it or Dan, or I don't know, someone that instead of the page in this whole way that I just described, we're going to start thinking about space itself as the interactive, um, environment.

I wish I had the, I guess like instead of programming a page, you're going to program the space. I think that's what the quote was. Yeah. Yeah. That's a completely different. Flipping weird, way of thinking about this, thing that's now become normal for us, which was also very weird when it first came out, the internet.

So, that's what this, like you just talked about, um, HSML, which I think is so cool. Even just the name, hyperspace, modeling language, oh my god, as someone [00:11:00] who knows a little bit about programming, it's like, it's very exciting. But um, for someone who doesn't know anything about like what this means, how would you just, you've done it in your writing and I'll link to it, but just, how would you kind of introduce them to, you What, what's going to change?

What could change? What's happening with SpatialWeb?

Denise Holt: So you know, instead of having website domains, all entities become a domain, so people, places, things, um, they become domains that are programmable.

Andrea Hiott: And you're not going through a medium anymore.

Is that right?

Denise Holt: Yeah, the space itself becomes a domain. So, um, and the modeling language is what's used to program those spaces and it can program all kinds of things. Um, the ontology for the programming language, it takes into account activities. It takes into account credentials.

It takes into account, um, All kinds of [00:12:00] aspects, um, multidimensional aspects about those things and what's occurring with them over time in that space and with whom and who has access who, all kinds of things like that, like if it can be measured. In a physical or digital reality, it could be programmed using the hyperspace modeling language.

 The spatial web talks about being natural. And, and the way you just described that, that sounds like a human life that kind of develops and you have a history and a trajectory, which we don't have in deep learning.

Andrea Hiott: You just have a, or programming, it's a totally different way of thinking about it. Um, does that make sense for you? Because I think a lot about how we each have these kind of paths in life and so on and so forth. And, um, something like a deep learning AI has. Has not anything like that. Um,

Denise Holt: right.

Yeah. No. And what's really fascinating though, is that you're programming these attributes about the physical life, [00:13:00] but also about a digital realm too, right? So, that means that everything then becomes a digital twin. And then you can run simulations with the digital twin. You can do projections. You can, um, you can Run simulations to find out all kinds of things, tweak parameters and stuff.

And, it's going to, um, really give us a lot of control with this simulation aspect of it. Businesses will be able to run real time, real simulations based off of real data and real things. And, Um, make better decisions we'll be able to address things like climate change problems and stuff and run simulations of our planet and really be able to see what we need to do to what degree to get the results we want.

When you think of smart cities, um, city planners will be able to run simulations and really find [00:14:00] out what's going to be effective, what's going to, um, make a difference for the population, for, systems within the city

Andrea Hiott: in a way it's, um, giving us better predictive processing skills, by sort of, yeah by having this twin simulated layer.

But I guess what I'm trying to envision is like, do you, okay, so do you imagine I could just walk into my garden, and, Simulate what I might do there in that space, or would it be more like similar to something that we're used to where I would go into a VR environment or even something just normally like a computer screen and simulate, what's the difference in the way that everything is linking together?

Denise Holt: I think what we need to really take into consideration is that, um, A lot of things are going to be evolving over the next, even just the next couple of years,

trillions of sensors are coming on board over the next 10 years, and those are all sensors that are going to play a part within this, network.

As far as AI is going to be able to interact with [00:15:00] it, with the information coming in off of those sensors, um, it's going to be imperative that we have AI to help us parse all this information.

And then devices are going to evolve. I mean, we've already seen the Apple Vision Pro, and you know what that kind of does.

Um, There's going to be, different types of devices that evolve once, it's like with any technology, once there's new capabilities, then people start finding how you're going to integrate with it,

Andrea Hiott: Some kind of wearable, perhaps, or some kind of, I mean, we already talk about, this, the biological not implantation or something, but more.

No, but like,

Denise Holt: The, you already see stuff like with the little retina things, all kinds of clear

Andrea Hiott: implants and stuff for Parkinson's. I mean, it's, there's plenty of examples. Um, but I guess what I would like to try to clarify before we talk more is what are the kind of main Goals of spatial [00:16:00] web, or maybe not even goals, but priorities that are different from the deep learning world, because we've already tried to mention them a few times, but

Denise Holt: so what it enables is a completely different type of AI than deep learning because it, it enables distributed intelligence at scale so the big difference is obviously you've got deep learning and then you've got the active inference and they function completely differently.

There are two different methodologies, but what the spatial web does is it gives this, um, common language for all of the AI to the intelligent agents to communicate with each other across that network. So it's. It enables a network where you can have multitudes of these intelligent agents that are all learning from their own frame of reference within the network.

People will be developing these intelligent agents. Humans will be developing these intelligent agents and imparting their [00:17:00] own knowledge into them, right? So it's going to preserve. All of these, all the diversity of knowledge throughout our world, um, you've got deep learning, a monolithic, machine that is trained off of just, trillions of pieces of data that are, intended to, to get it to Output something that seems appropriate or matches kind of your expectation of what you're hoping to get it out of it.

Right. But all it's doing is reformulating what it's been fed and spit it out. Right. It can't come up with new information. It can only draw off of the information that's there.

Andrea Hiott: Right

Denise Holt: now in the spatial web with these active inference agents, they're learning based off of real time data.

They're learning based off of the real world that is constantly evolving and with the way active inference works, with the action perception loop and they're actually really learning. [00:18:00]

Andrea Hiott: Yeah.

Denise Holt: They're able to, they're able to adapt with all that new information. They're able to then share it with each other, learn from each other.

So then you have this distributed knowledge that grows into this ever evolving collective intelligence. And that's really the difference.

Andrea Hiott: Yeah, very well said. I guess, um. So with something like deep learning, you, there is a programmer who's putting in the information and yes, we could, we can train it on tons of data.

We can put all the data in, but it's still, once you put it out, then it's trained on that data. So you're going to have to do a whole update, like chat GPT for whatever you have to keep updating. Um, but what you're saying is this this, um, AI, a spatial AI would be more about the space itself continually updating itself, much the way the real world already works.

I mean, that we all sort of, um, dynamically interact and update one another's, um, Markov [00:19:00] blankets or whatever. But is that kind of the

Denise Holt: Yeah. Yeah. when you think of, our human knowledge, it grows because we have a diversity of intelligences that are, sharing with each other, but also pushing back on each other and challenging each other.

And, that's how the information gets tested and then what's right sticks. And then it grows, um, it's the scientific method, right? Yeah. Yeah. And that's really the difference. And that's why, when you just have one brain, it's not going to grow its knowledge beyond what it knows.

How can it, it can be introduced to more information that then is in its bank, but those machines, they're, they cannot adapt to new information, just like you said, they're because of the way they're trained, they take a lot of adjustments to, to to [00:20:00] get it to actually, um,

Andrea Hiott: Yeah.

Denise Holt: Yeah.

Andrea Hiott: Or

do whatever we want it to do. Write a paper.

Denise Holt: Exactly. And they're useful. They're very useful, but they're tools. And I think what we're going to see is that they're not going away, but they will also come onto the network as tools, just like every other Yeah. Every other technology,

Andrea Hiott: I don't know if you've seen this.

I wonder in your experience. I also wonder when you first started, like encountered AI and I don't mean using it, but thinking about what AI is and, um, becoming part of the creation of it or whatever, but

Denise Holt: well, I've been a nerd for a long time. Okay. So you always, yeah. I was reading Asimov and Heinlein.

Oh, okay.

Andrea Hiott: That's great. Yeah. We were all, a lot of us start with sci fi. Um, yeah, but I guess, um, I wonder if you've seen this, like with people who aren't nerds about AI, I'm not, I've become more so, but when you just are like, in your normal life, a normal person doing their normal job who doesn't need to think about AI.

AI [00:21:00] every single day. But now most people do because it's, it's suddenly everyone's thinking about AI or they see AI doing art or they realize AI can write their paper for them. It's changed right in the past few years in a big way. Oh yeah,

Denise Holt: I think so. Language moms. Really with chat GPT, that, that really brought it to the surface for most people.

Yeah.

Andrea Hiott: And I guess people seem to imagine that it is learning in the way that you described the spatial web? It's it doesn't seem that it, um, people really understand that it's actually very language based in terms of this kind of, that you're putting in the representations and that's all that can, it can give you back.

Denise Holt: Yeah. I mean, people definitely misunderstand what's actually occurring there because it's, yeah. It's a parlor trick.

Andrea Hiott: Yeah, it is. I mean, it's, it seems magical if you don't know what's going on.

Denise Holt: It's really good at making people think that, oh my gosh, it must be thinking because, I asked it this and it gave me an answer.

 But, but the reality is, is if people really [00:22:00] understood that, the way these language models are working, it's not hard to recognize, you give it enough training data. It's not hard for it to start recognizing the patterns, I mean, we've got 26 letters in the English alphabet, and, you find that certain letters tend to follow each other, certain words tend to follow each other, and then you train it on specialized data like legal data or whatever, then it starts to, learn patterns that are of things that are said over and over. So you give it a prompt, it's gonna search through its database and go, okay, well, this It's probably, advice and sound appropriate, but appropriate and accurate are two different things.

Appropriate doesn't mean that it's understands anything, um, so if

Andrea Hiott: you ask me a question and I just search on the internet and tell you what I find, I don't have to understand it at all. It's kind of the Chinese. thought experiment in philosophy where you're just, it's just take it.

It is, um, the way that language works, if it didn't have all those very repeatable [00:23:00] regularities and patterns, it wouldn't be language. So the machine as it's built just learns those and then can reproduce them, but spatial web. How is it? I mean, I know it's going to use AI. So maybe we should talk about that.

What's like, what do you see? How would people begin to understand that? Is it more like how they're actually imagining the current AI?

Denise Holt: Yeah. So the thing is these active inference agents, they actually can, um, do causal reasoning, right? They the way it works within the free energy principles, it's constantly trying to, um, it's taking in sensory information.

And it's trying to predict what was causing that, right, so that it can then minimize

Andrea Hiott: surprise, I guess. What's your like one, one sentence idea of the free energy principle?

Denise Holt: Yeah, I mean, basically you said it, it's, taking in sensory information and then, acting on it to Make better predictions,

Andrea Hiott: yeah, to make [00:24:00] better way.

Yeah, I think we can it's generalized, but we can just for people who don't know what haven't heard it if there is, that it's pretty simple. It's what we're doing. And when we were moving through the world, we don't want to it's what we're doing. We want to develop certain habits and patterns so that we can continue moving through the world, so that we can exist.

That's what, we learn how to walk, we learn how to talk we learn all these kind of, um, ways of being in the world, making our way in the world. And you can think about that as the body minimizing, surprise and just aligning with the world in a certain way.

Denise Holt: Um, Daniel Friedman, he actually gave me the most simplistic example, which I've been using with people because actually it does kind of give you that understanding, he was like, okay, say you have a wooden table, right?

You think it's wood, um, and, or it's wood, but you're wondering, is it rougher? Is it smooth? You have to touch it. You have to actually act on it and find out. And then that gives you more sensory [00:25:00] information and Oh, it is smooth or it is rough. Now, right. But it's that whole, um, curiosity by acting on your environment so that you can get more information to, gain that understanding.

And. Yeah.

Andrea Hiott: Great. So the spatial web would be using a kind of, it would, instead of deep learning, which where you're just training the agent on on data that pictures or words or whatever, um, what's happening with the active inference? So it's

Denise Holt: taking in real time information through sensory data in like IOT sensors.

So you have cameras, you have robotics, you have devices, all kinds of things that are feeding real time information, um, into the network. And then you also have all of the, um, HSML that has been programmed into the network with providing context for all of the entities in the spaces and their [00:26:00] interrelationships with each other.

So, So when inactive inference, when you're taking in these, the sensory information, you're measuring it against what, to be true, the model that you've built of your world, right. And it starts out, as an infant, it's just your mother, and then you start opening up to the room around you and the other people in the room.

And, as children evolve intelligences, but You know, it's that expansion of our own world model by this constant interaction with our environment, and learning in that way. That's how these active inference agents are going to learn. They're going to be constantly taking in real time data through sensory input, measuring against what they know to be true, which is the the model they've developed from all of the context.

Baked into all the things in the world around them. Yeah. Um, they'll be operating from their own frame of reference. Like, whether it's like right now I'm [00:27:00] here, I'm sitting in my house, and so this is my frame of reference. My frame of reference expands on things that I learned, things that I've studied, things that I'm, my own programming, basically.

Yeah.

Andrea Hiott: Yeah, I mean, this is what I call, like, a way making or I think of as navigability in terms of just that, throughout our life we're born and we go through all these experiments and when we do develop, Um, you could call a model, um, even though it's not like a static thing that's sitting inside your body.

It's more your body itself in the interaction ongoing, um, but you develop a way of being in the world. So as to keep being in the world. And so I guess what this is going to do is help in a way extend that. So it's almost, um, we almost have to kind of rethink what it even, like, it almost feels weird to call it artificial intelligence now, even though of course it's in that, um, um, Yeah.

Denise Holt: To me, when it like with, cause my whole channels are, spatial web AI, but to me just because of what you just said and I mean, [00:28:00] if you listen to Gabriel Rene or, any of the versus team, they're like, we're doing natural intelligence, not artificial intelligence.

So to me it's like autonomous intelligence,

Andrea Hiott: that's good.

Denise Holt: I feel like that's what AI is going to be transitioning into actually meaning when it's referring to this. And you'll learn about the autonomous intelligence systems, um, yeah. Is it

Andrea Hiott: still human centric though? Because I hear the word natural used a lot and it's almost like a natural versus artificial and of course with Carl's work and active inference, I mean, it's literally building from a living dynamic biological system. So it's, um, but I wonder, is it. Like, do you really how do you think of that? Do you see that as a kind of a dichotomy that, um, or is it blurring a lot, this idea of natural and artificial?

Is it, what is the natural part?

Denise Holt: Well, the natural part is because, the way these intelligent agents learn is

Andrea Hiott: The way we learn, the way life

Denise Holt: learns. The same way that [00:29:00] biological systems learn, according to active inference, which is based on the free energy principle, the free energy principle is how neurons learn.

Um, so, that's why it's termed natural intelligence because it's the intelligence of nature.

Andrea Hiott: Now, um. Which is another confusion because deep, the deep learning, people think it comes from the brain, but it's actually not. Not really the way that life works. It just has the name like that. It's just,

Denise Holt: yeah, no it's engineering.

It's inspired

Andrea Hiott: by. The

Denise Holt: need of engineering.

Andrea Hiott: Yeah.

Denise Holt: Versus science, right? Right. So, I mean, that's really the difference. What's really interesting is that, um, Jeffrey Hinton came from University College, London, Carl Friston University College, London. At one point they worked across the hall from each other.

Andrea Hiott: That's crazy.

Denise Holt: But you know, Jeffrey Hinton

Andrea Hiott: was

Denise Holt: focused on, um, an engineering approach and Carl was like, no, I'm,

Andrea Hiott: yeah. And like you were saying [00:30:00] before, it's. It's not either or. I mean, we have this AI, traditional AI, deep learning. It's pretty amazing that we can use it to better understand things, to save ourselves time and energy.

Denise Holt: Um, for as far as like, um, the only approach to AI, it's just not sustainable energy wise yeah, that's really interesting is versus just gave a demo a few weeks ago, Demonstrating what they, so they've been, doing these AI benchmark tests, that are like really important within the AI industry because they know something.

Andrea Hiott: So like looking at,

Denise Holt: The

Andrea Hiott: Go

Or arcade games and stuff.

Denise Holt: Yeah, like Atari challenge. Yeah, and what was really interesting is, um, in one of the benchmarks, which I think was playing Pong right now, the fastest, best deep learning machine could get to human level performance within two [00:31:00] hours. They did it within 12 minutes, and they did it at like 1 percent the model size, and they did it on a standard laptop using a standard graphics card.

So, that's

Andrea Hiott: it's real.

Denise Holt: Yeah, that's a huge difference too. That means that, um, that means that All of the energy resources required for the AI is distributed.

Yeah. Among our devices we're already using, right? Um. Yeah, so. Much more like

Andrea Hiott: the body learns, because you can't, yeah you have to learn in that kind of a, at that kind of a pace, too.

As you're talking, I'm realizing it's spatial temporal. I mean, there's a whole time, um, paradigm shift happening here, too.

Denise Holt: Yeah. It's exciting to me. And that's why I like, um, because I think maybe because I've seen like the vision unfolding and I've kind of knew it was coming, right.

[00:32:00] Yeah. To know that, they did it and, it's even bigger than anybody imagined. And it's just, It's exciting to me because I've known what this, the potential of this could be, I, and I think, I think once people start to see really clear understanding of what sets their technology apart from what is happening in the, um, the deep learning world.

I think that there's going to be a lot of light bulb moments this year, where people are going to start really understanding, Oh, this is, Dan may say, he, he refers to, um, the, this collective intelligence within the network as a, um, nervous system for the planet, I mean,

Andrea Hiott: it's pretty interesting.

I think we, I mean, we already, gosh, it's already been some time, but I really, um, I really think it's important to talk about that idea of transparency, that idea of energy [00:33:00] efficiency. You talked about it a little bit already, but this that people could own their own data. Um, there's all these things now that we assume, as I was saying, we have this kind of system, and we assume it has to be this way, that we have to give over our data, that we have to go through a middle, we don't really own our domains, all this we just assume, is the way it should be.

Also what you were just describing about the efficiency, the time we spend so much energy and, it's not transparent. All these things. I guess when I was asking you about what are the kind of visions of spatial web, I see there's a bigger, there's a philosophical vision too, and a value based, and we haven't mentioned that word yet, but there's a value shift, right now we take for granted that we just are addicted to our phones and our computers, but those are algorithms.

Those are. Things that are, those are, that's on purpose, um, doesn't have to be that way. Right. Isn't this opening up like another way for people to imagine how that they could use technology?

Denise Holt: Yeah. And I, what I think it's going to do is it's going to [00:34:00] integrate itself, um, into our daily lives a lot more naturally, um, because it will be, we'll be interacting with sensors and interacting with all of these different, um, elements of it.

And we'll have, so everybody's going to have their own. Active inference agent personal assistant that's going to be able to help, parse all this crazy amounts of data that we are, that we encounter, it's going to be able to, um, help us navigate all of this information according to our own preferences and desires to, right now, like you said, we're all subjected to these algorithms that are predatory.

Andrea Hiott: Yeah, they're built to be. I mean, that's the way they're built, because that's how they get attention and money.

Denise Holt: And we're going to have a lot more control over our data because we'll be able to, um, we'll be able to decide how and when it's shared and for how long and with who. And, [00:35:00] um, people will still trade data for, um, access and different things like that.

But the difference is you'll actually control a lot more of the nuance with it. And you'll be able to decide, and there will be a value exchange that actually rewards you.

Andrea Hiott: Yeah. I mean, it sounds wonderful, but how, I mean, I can imagine people listening and they're thinking, this sounds great, but I don't see how it's going to happen.

Denise Holt: Protocol and the way the protocol works. Right. So the way H HSTP hyperspace transaction protocol, um, Every all entities, right? So people, places, things, spaces, every entity has an identity and, and uh, it becomes a domain. And so is it kind of a

Andrea Hiott: blanket sensor that just goes over the whole earth or something?

I mean, I guess I'm trying to understand how,

Denise Holt: but there's going to be a registry, right. And the, um, to be able to access domains through activities, [00:36:00] um, It will require credentials, right? So, um, credentials that, can prove identity or ownership or, um, authority or, um, I mean, all of that, right?

So, So

Andrea Hiott: there's more transparency, not less.

Denise Holt: Because the control is in your hands, it's in the user's hand. So, um, I will be able to control who can have access. I'll be able to go to the doctor and tell the doctor like, okay, you can access this information, this. Guardrailed amount of information for this long.

You can put an expiration on it there. The spatial web enables, um, zero zero trust architecture, right? So, um, because it's going to be powered by a I within it too. There will be able to be con confirmation of certain, um, aspects of data that, where you don't even have to reveal it to a source. You don't have to reveal the [00:37:00] sensitive data, but it can be confirmed within the network.

Oh, so it completely protects your your private data. There, there's so much privacy protection. One of the things, um, that Dan told me, a while back that when he and Gabe were really contemplating, um, how to do, what this protocol would need to be, they really wanted to solve the three main problems with the World Wide Web, which is like hacking, tracking and faking.

Right? So, That's great surveillance. Um, the whole hacking and,

Andrea Hiott: um. Hacking, tracking and faking. That's.

Denise Holt: Yeah.

Andrea Hiott: Good way to

Denise Holt: say it. So it is it's secure and it truly is decentralized. And so, the power then is in the hands of all of the individuals, everybody within the space, your AI agent, your personal agent is going to be able to know all of your details and know you.

So well, but that will be [00:38:00] guardrailed, um, it's not going to be accessible unless you give permissions and it's nuanced permissions. It's not just, Oh, you can access me. It's like, right now that's what we have. You can access or not.

Andrea Hiott: Yeah. Or you just give your data away.

Denise Holt: The blanket statement.

Andrea Hiott: Yeah. I think that's, um, I mean, I think people just assume that's the only way it can be too. And it's kind of amazing to. Realize how much we just sign away all of that stuff for whatever the next five years or forever or whatever And this is more like a disappearing message or something where you can Share it for the amount of time it needs to be shared and then it's not there anymore And that's possible with this technology.

That's clearly um,

Denise Holt: yeah And then when you think about like, even what that's going to do for, for business and for like big enterprise organizations, one of the issues that enterprise organizations have right now is, um, a lot of them are saying to their employees, don't use. [00:39:00] These LLMs don't use chat GPT.

Don't use them. And the biggest reason why is because of the privacy concerns, an enterprise organization has all this proprietary data that's opening their data up to a third party that is not part of their organization now, the spatial web. You can guardrail any aspect of the data because you can guardrail the domains, you can guardrail the data as domains.

And so when you've got an enterprise organization, that enterprise that can be within the network. And guardrailed off to where the capability is still there to build agents within that guardrailed part of the network, it can learn off of all of that data, it can access the network, but it can protect all of the data within that enterprise, right?

That enterprise can decide who in their organization can access what [00:40:00] data. That enterprise can decide what data is allowed to be accessible. Well, by the public or what parts of the public or memberships or anything like that enterprise, most enterprises have, several disparate data lakes, right?

Where they've got data from this part of the company, data from this part of the company, and there's no way to join those data,

Yes, like

Andrea Hiott: those silos you were talking about before.

Denise Holt: Now with this. The data is all good. They can build these intelligent agents that can share the data, learn and they can create, they can gain insights that they never had before in their companies, but it all stays proprietary.

And it's not

Andrea Hiott: private and transparent, weirdly.

Denise Holt: Yeah. Yeah. So it's. It's really interesting. I mean, I think when people really start to understand what's happening here and what's being built and they start to really think about the implications of it, they're going to see that It's it's pretty powerful, and what

Andrea Hiott: do you see when you're thinking it's powerful?

[00:41:00] What do you envision? Like, what would be your kind of one, one scenario that's makes you, I don't know that's happy or good or for the future once people do realize this, because it is a lot to think about all these terms and it's hard to get your head around, but like, how do you imagine that this could, in your own life or your, Someone's life that you love changed.

How would it change their life for the better?

Denise Holt: Yeah, so, it depends on how, like, near or far in the future you want to go. Um, although when I say far in the future, to me that's just like five years. Okay, let's go there. Five years, but

Andrea Hiott: Let's go five, ten years, best case scenario. Yeah, five,

Denise Holt: 10 years. I mean, honestly, I feel like it's going to really shift the way we, um, experience life as humans.

And, I think that it's going to enable us to shift in all kinds of ways, um, the, all of this technology, it's going to, [00:42:00] usher in this realm of abundance. I feel like, It's a transaction protocol. It's blockchain agnostic. All blockchains can work within it. All, um, I feel like we're already seeing a shift away from fiat into digital currency.

And I feel like within the spatial web, then we're really going to get to that value exchange Position right to where, it, our idea of what we've had for money, I think will completely shift so, um, I feel like everything that we, um, that we understand about the way we interact with each other, the way we interact with technology, it's all going to, um, I think it's just going to change into a way, like I, I know a lot of people think, um, when I talk like this, that it's, um, it sounds very utopian, but, um, I definitely see that there can be so much good in this.

I think that [00:43:00] right now, as a society, we've been so used to kind of scratching and clawing for existence. There's been this competitive nature because of that, right. Um, we. We have a society that operates off of ego and, um, there, there's just so much, um, it's funny.

So maybe this just popped into my head, but the, I forget who was, I think it might be in like C. S. Lewis's, like some of his writings and stuff, but it was a distinction between aspiration and ambition. And it says,

Andrea Hiott: Ambition. Okay.

Denise Holt: Yeah, aspiration is high. Um, ambition is just higher than someone else.

Like with ambition, you're just trying to be higher than the next person. But, um, but aspiration is high, right? So, the ideal, it's the ideal that you're going for your own personal best. And I think that we're going to We have the potential to kind of shift [00:44:00] as a society to where we're kind of like looking more outward.

We're more community focused. We're more, we're less, we're less in that vein of just trying to get better than the next guy or get the, the next thing, to higher up to, we're, I don't know. And maybe I don't even know. As you're talking,

Andrea Hiott: I'm thinking about how technology is such a big important part in so many people's lives personally, but also work and how it does depend on this kind of judgment.

Um, likes, follows how you look on Instagram or whatever, I mean, we've really gotten stuck in a way with, um, what the algorithms have the parameter have parameterized for us. What's that?

Denise Holt: It's that reward function,

Andrea Hiott: the dopamine and all that. But just also that we, I we I wanted to talk about this idea of a sense of self.

Um, you talked about it with, um, Mal I think, but. I'll just have to link to it, but there's this, um, there's also this the [00:45:00] way you were describing the agent or the AI or the, um, the technology, it's kind of going to develop a bit of a sense of self because it's biological, but because it's interacting with us, it's also opening up how we can see ourself in the same way that the old technology, the stuff we're still using now, which is great, but also has really limited in a way.

um, the ways we think of self. So when you said you see it opening up our experience of the world, it made me think like maybe we're gonna look around ourselves again and notice the space we're in, um, and start thinking of self differently as more spatial instead of clicks and likes and images on screens and all this.

Denise Holt: Yeah, I would definitely agree with that, I think it's going to, to that point, I think it is, it's going to be able to enable us to interact with technology more in the world, yeah, it will take us, [00:46:00] it will take us from being, um, tethered to the devices.

Andrea Hiott: Which is, yeah, do you experience that in your life? Are you also? If

Denise Holt: I leave, if I don't have my phone, like say I accidentally leave the house, which never happens anymore.

It's like, Oh my gosh, you feel naked. You feel like a part of you is missing. And I know everybody feels that way. Your battery dies on your phone. Oh my God.

Andrea Hiott: Yeah. Have you experienced any kind of, like, disconnection from space, the world? Because I keep coming back to this idea of natural and I start, I wonder if this is also gonna be, um, an advocate, so to speak, for other beings and forms of life, um, in a way.

Have you ever thought about that, like, or in your own life, do you, have you experienced some kind of way in which your focus has gone more toward technology and could open up ecologically through something like this [00:47:00] technology?

Denise Holt: Well, so to me, it's not all negative because, in that vein, I actually, so I live in South Lake Tahoe right now.

I've been up here since the beginning of COVID. Oh, wonderful. Um, when I came up here, it was with my ex boyfriend and we have not been together for the last two years. So I've been here by myself, but all my friends are like in the Bay Area or Southern California or the rest of the globe.

Right. So, um, and because I was up here for the first two years with him, it wasn't like I was out trying to make friends. So, um, It feels kind of isolating up here. But you know, I've also been working really hard for the last two years. And so the solitude I've been embracing it and just keeping my head down and building and stuff.

But, a lot of my social needs are met through technology right now. That's how I'm keeping in touch with my friends, and, we do the same silly stuff as everybody else, just send each other memes and all kinds of stuff. [00:48:00] But those are little touch points of connection and, I think they're important because you're not just sharing something, you're sharing things that, we'll find like a common laugh or a common, heartstring that you're pulling together, that kind of stuff, it's beyond just sharing, benign images, but, um, but, and, like Facebook and stuff like that really, That helps me feel connected with the people in my life, even though I'm not near them.

So I think that's really cool. Now the way this technology is evolving, yeah, it would be pretty cool to be able to like jump into a space and feel like I was in the room with my friends and, we could all just like gather like that or, like I've thought for years, like Why is teleportation still not a thing?

All those sci fi books made it seem like we'd be having it by now. Yeah. Yeah. Maybe it's just a different kind of teleportation [00:49:00] that we're gonna, so,

Andrea Hiott: yeah, that would be exciting. I think it's great you brought that up because, um, I mean, technology is wonderful and it's not either or, and we do connect very deeply through technology.

Um, it's not that we. In a way that it's opened up ways for us to connect and, um, I don't know, do you think that's kind of the point of it in a way?

Denise Holt: I do. So, um, it's interesting, um, because I have friends that they get, they feel very pressured by technology. They feel, um, I think that a lot of, I think a lot of it has to do with the fact that we're all individuals and we all have different psychological makeup.

We all have different experiences and traumas and different things like that, we're emotional. So I know some friends that they feel a lot of insecurity and a lot [00:50:00] of, um, pressure from social media to where it really deters them from wanting to have a whole lot to do with it, because otherwise then they kind of get stuck.

Yeah.

Andrea Hiott: Yeah.

Denise Holt: And, I've always kind of approached social media differently than that, because to me, I've seen it more as the positive sides of it of like, Oh, great. Like w when my kids were young, Family members were always wanting me to email them pictures. Oh yeah, right. Now you can just check my Facebook.

It's a bulletin board. It'll let you feel like you're involved. You don't have to email 400 people, you here for me, we can all keep in touch. That's true.

Andrea Hiott: Wonderful thing.

Denise Holt: So, like to me, I've seen a lot of positives. I, like even with like Facebook, with the Facebook memories and stuff, like that's cool to me, I love seeing stuff, even like things that you maybe wrote as like little, like posts.

to, 10 years ago now. Right. [00:51:00] And it's like, Oh, my past self is, giving advice to my future self, because that's like with me, aspects of it that. Maybe nobody even considered, but, um, it's kind of unfolding that way. So I tend to like, look at those parts of it and the positive sides.

And then also too, like when I first got a smartphone, I turned notifications off on everything because it killed me like the, I was like, I can't be interrupted every second of my day. Like I, I just can't. So, I think part of it is taking control of the technology and making it work for you.

Right. Because to me, that was too much of a distraction. And it's like, okay, well I'll check my email every 10, 15 minutes during a work day, I'll do these things, but I don't want it dinging me constantly because then it'll drive me nuts.

Andrea Hiott: Yeah.

Denise Holt: Um, so.

Andrea Hiott: I did the same. And as you're talking, I'm thinking maybe there's, because I know a lot of people who are really familiar with technology do these kinds of things and don't take it as seriously as, [00:52:00] or don't get as sucked into the, um, the aspect of it that where you're comparing yourself and feeling bad, but those are real spaces for people, but do you think there's something about having Yeah.

Yeah. Thought about technology, worked with technology for so long it's given you a way to, um, have that kind of distance or space to be healthy about it and enjoy the emotional side of it without the judgmental side? Um. Or is it just you? Like your personality or?

Denise Holt: I really don't know. Um, because it's really kind of hard to say because I've actually had kind of, I, I'm definitely somebody who embraces change, and a lot of people have a fear of change, right.

And I think that comes down to the free energy principle, that fear of that, uncertainty, a lot of people that, that creates this sense of fear because they're, uncomfortable with it.

Andrea Hiott: You like the surprise, maybe, a little bit.

Denise Holt: Yeah, to me, in my [00:53:00] life, that's where I've found possibility.

Possibility lies in the surprise, and, it's kind of, like in physics Things could always go one way or the other. So, um, but, for the last probably, I would say probably at least since like 2008, 2008 ish, watching what was happening with technology and seeing things that were unfolding and things that were starting to really wake up.

um, with these emerging technologies, I have just felt like lucky to be alive right now, to be able to witness this and, you can't stop any of it. So, find your place in it then you actually have this opportunity to make a difference and, so I've been so preoccupied with that mental state that I think it lets me embrace all these things in this way,

that's very beautiful.

Andrea Hiott: Yeah.

Denise Holt: I think people maybe who aren't that entrenched in that kind of a, excitement over it [00:54:00] all, um, then maybe they get preoccupied with other aspects of it, and, um, it has a different effect on their mental state.

Andrea Hiott: And there's something about understanding how it works.

Even when we were talking about the deep learning and understanding that you're just being fed a bunch of stuff that was trained, like once you kind of understand that, it does take you. It does give you, at least for me, I guess I'm speaking for myself, it's easier to not take it so seriously and to understand it's, it's, it is just what it is.

And maybe you can, um, use your own agency to decide how you're going to use it. But.

Denise Holt: Yeah. And you can kind of laugh at it at that point, like, yeah,

Andrea Hiott: you can, like I,

Denise Holt: There was a meme that I saw, and it was probably a couple years ago now, but it was like, my favorite way of shopping is just to call out what I want and let you know.

Facebook serve me an ad for it, . Right, right. That's great. And it's that, it's like what you're talking about. It's the understanding of the underlying of how it's all working, so

Andrea Hiott: it allows you have a playfulness with [00:55:00] it. I think that's very healthy. Um, yeah. Well, so I guess to end, I want to hear, about what you've been working on all these years.

I want you everyone to kind of know, but this is called love and philosophy. It's really beyond dichotomy too. But I wonder about that word love, like we've been talking about all this technology and stuff, but. Do you feel like when you look at these, this, what you're doing, your work and, um, what you're sharing with people, it's a big word, I know, but does it feel like that or like passion or like something that goes deep in you?

Denise Holt: Oh, yeah. Yeah. Yeah. I'm definitely very passionate about what I've been doing and, it's because, I do see, a good outcome from all of this. I do see like the good that, that can come from all of this technology and, with what I see that versus has been building and doing too, I feel like they're doing it the right way.

They're doing it with kind of the benefit of humanity and mind, it's

Andrea Hiott: very [00:56:00] mindful and. Yeah.

Denise Holt: Yeah. And to me, I see a lot of promise in this next era of computing, whereas we've been kind of in the state of like the technology has been taking advantage of all of us, I feel like we're moving into this space where that's going to shift and that's pretty exciting, and that.

I mean, I just see that can only result in so much good, well, thank you for

Andrea Hiott: your optimism and for, it feels like it's a very authentic place that you're coming from. Um, and just kind of to end, like tell us what you're doing. I'm really interested in your sub stack for example, I found some very interesting articles there, which I wanted to talk about, but we're out of time, but what would you like people to know or how can they find you?

Denise Holt: Well, so, um, yeah, so education right now is my focus on this space. And so, I've been doing a lot of writing. I've been doing a lot of podcasting. Um, this year in January, I started a sub stack, excuse me, because, [00:57:00] um, I wanted to be able to kind of have a space where we could build community around it.

There's so many people I've been meeting who are, Interested in it. They're working on projects or they're, there's a lot of people who are either coming from the developer side or the, the education side or the research side, or even the investor side, right. And they're all curious.

They all want to know. And, so I wanted a space where people could find synergy and kind of connect with each other and we could all learn together. So, so yeah, I started the sub stack. Yeah. And then I've also been hosting these learning lab live sessions. So monthly I'm creating a new presentation every month.

That's an educational presentation to kind of introduce these concepts and, walk people through it. And I do two different days with two different times so that people could try to, I've got people coming from all over the globe. So, um, That must be exciting.

Andrea Hiott: Wow.

Denise Holt: Yeah, it's been great.

Yesterday was the first one [00:58:00] for this month. So we did it last month. There were two sessions. This one was the brand new presentation. Um, it's like a 40 minute presentation and then about 45 minutes of just open discussion and everybody's been participating. It's been so fun. The discussions become my favorite part.

Andrea Hiott: So you present, um, on some particular, almost like a class or. Yeah, and

Denise Holt: I'm going to be using the videos form for a larger curriculum that I'm building around this. Yeah, and this month was on the Spatial Web Protocol. That's what the presentation is for this month. And it's an overview. It's a very in depth overview.

But it's the beginning of what I'm creating for like this e learning class around the actual protocol itself. Because it's very detailed and it's going to be. Um, it's going to be made public soon. So when it is, then there'll be something that people can actually learn from. Oh, good. So at some point that's where my [00:59:00] focus is right now.

There's

Andrea Hiott: already a lot people can learn. You already have given, you have a ton of content, um, just explaining what this is, and I encourage people to go, I would say, read it and listen to the podcast.

And will there be a way, so if I join, I was telling you, I was trying to, I was trying Do the sub stack, but it didn't work quite well. But once I joined, can I go back and look at the, are you recording the lectures that you gave and you'll have them in an archive or something or?

Denise Holt: Yeah, so they are being recorded.

Um, I'm still trying to figure out how that's going to work because it the classes are for the paid membership on the sub stack. And so, I have to be able to host them in a way that can be accessed, by that sector. So I'm figuring that out. Um, but yes they,

Andrea Hiott: Probably it goes into the classes anyway at some point so yeah, but it's exciting because right now people are starting to find out about this I mean, it's I think the book came out in 2018 2019 spatial web But that's actually not that long ago [01:00:00] for these kind of things It feels long when it's happening, but it's starting to really break now and it's great that you're Right there, riding the wave and helping people understand it.

So thanks for that. And I'll be sure to link to all this. Um, Thank you so much,

Denise Holt: Andrea. I appreciate it. And thank you. Thank you for having me on your show. This has been a real, real nice, it's been a really nice conversation.

Andrea Hiott: Yeah. It's been fun to talk to you. Is there anything that you want to say before we go , anything we didn't talk about yet that we should touch on real quick or,

Denise Holt: I mean, my, my one thing that I will say is, I want people to get excited about this, because the opportunity, it's not like, it's not like this is something that's distant in the future. This is all unfolding this year, right now versus a genius platform. It's in beta with a closed beta, but they're opening it up to the wider developer community this summer.

So it's. Yeah, it's upon us.

Andrea Hiott: Now is the time.

Denise Holt: Yeah.

Andrea Hiott: If you want to be at the forefront. [01:01:00] Okay. Well, that's good to know. I'm glad to hear that. And I wish you lots of luck and I'll be following you.

Denise Holt: Thank you so much. Same here. Great. Bye bye.