Programmable Cryptography and the Art of Selling Your Face || Moist Cryptography

I see in the camera. Okay, yeah. So, I'm going to talk about moist cryptography, which is the art project on the foyer, which I imagine perhaps most people have either seen or passed through or wondered about. Now, I don't have any uh slides. I assume everyone has uh probably seen it by now.

Or if you haven't, then it's over there, and you know, kind of useless for me to show you some slides. Um maybe I'll give like a very brief overview of it, and then maybe I can just kind of make it interactive and take questions and comments from the audience, you know? So, I'm just not kind of like rambling about it myself. Um let's see. The basis of this project I think one of the core inspirations came from kind of exploring the feedback loops uh between like between like humans and their environment.

Actually, I was like uh we were very inspired, but myself and my collaborator in this project, Dr. uh Bardacos, about um and there's a a very incredible bio artist that I'm ashamed I always like Stelarc. I always say like Sayarc, but his name is Stelarc, who's like grafted an ear on his arm and a very pioneer of bio art. We had the privilege to see along with a number of other bio artists, and the way they described their work was always um kind of like I very stuck with me, and it was not as kind of I was going in expecting, "Okay, like an augmentation of a I put this thing on my arm. I put this hook in my body, or I you know, graft this moss on, and I become like a moss man or something."

But all their thoughts were more about kind of the systemic, um like integrated process of becoming part of a larger system and reflecting as part of that system as opposed to like a simple augmentation, which would be kind of the point of view of kind of like, you know, the I'd say a a pre-meta-cognizant viewpoint of looking outwards rather than looking at a totality of a system. That made me think about um cryptography as well as as an a nature, a new environment. Let's say the a blockchain as perhaps an inevitable new environment of how we situate ourselves within landscapes of like of of data. Uh and what that means when we do something as simple as perhaps let's say like uh scan our eye to open a door. So, I think this is a very good example.

Like let's say with something like a a Worldcoin or some sort of like blockchain-enforced biometric system where you use your data to gain some augmentation. Right? So, like okay, now I I don't need to open my door, I don't need to reach my hand into my dirty pocket, I can da da da, whatever. But, nevertheless, there's these feedback loops that happen where now it's no longer it's not simply an augmentation of like, okay, now I have this ability to open this door. No, it's like, now my eyeball has achieved like cyber-potent significance in a new feedback system where there's all sorts of different access panels and barriers to data and like barriers to acquiring resources for which now like the fact that I'm a human being and have eyes it has a special relevance to, even if I haven't participated or even signed up for such a system.

Perhaps simply existing within the environment in which this system exists has like altered actually like in some way what it means for me to be a human. And I think as more of these systems become more and more I want to say invasive, but let's say pervasive. As like more and more biometric data can be or will be exfiltrated from our persons, like our relationships to ourselves will necessarily change because our relationship to ourselves is predicated upon our relationship to our biometric data. I think you know, perhaps that has changed uh historically at different times culturally, but now it seems we're on, you know, uh uh a very monumental change of this nature because the of the discretion at which we can like isolate, target, and metricize biometric data, and and um and integrate those into systems that may be beyond our control. And so that interested me as like, "Okay, here's maybe like a a subject matter where using I also my my interest in in blockchain or like exploring blockchain as a medium, like what left what else is there in blockchain as a medium?

And I would argue, and you know, if any hecklers, I'm willing to like go back with that, like the only thing interesting left with blockchain as a medium right now is programmable cryptography. And it mostly everything else has been like mined out or else now BlackRock is doing. And I'm sure they're getting into like programmable cryptography as well, but at least there's still like a green field of new possibility that extends the what blockchain can touch. They're like programmable cryptography is you know, like the alien virus that like turns other things into itself. Whatever the other thing is the blockchain, and programmable cryptography can like touch some other off-chain computation and bring it on-chain with the assumptions of the chain.

And that's quite interesting. It's like expanding the environment of the chain and what the chain affords as a medium or as a biome. So, like computational immutability for one. It's like computational immutability we're thinking mostly of I don't know, like currency and assets and rigging our financial system, but more and more this computational immutability will seep into all different aspects of our lives. And then through the biometric layer is where it seeps into perhaps like the most intimate part of our lives, which is ourselves, our bodies, and our minds.

So, I felt this is a very interesting area to explore. And at the same time, I think offers an interesting exploration from the technical side from like um using things like zero-knowledge machine learning. Right now in industry, I I I find it very interesting you can apply zero knowledge to a machine learning model so you can have like a machine learning model run in such a way where like the computational guarantees are enforced and proven on the blockchain even with some sort of intermediary computational step being uh occluded from various parties. So, now we can like touch neural networks with the blockchain. Right now, it's obviously very expensive to do and so like the industry use cases are quite small, but I was like, "Oh, maybe there's like how what can we use this for?"

And then that's interesting and maybe art some sort of creative artistic um uh use for like zero-knowledge cryptography, you know, allows us to explore its implications outside of its currently limited uh industrial potential. Although that potential is only limited by like current limitations in the underlying cryptography, and as you know, some sort of Moore's law progression uh follows and kind of the orders of magnitude of computational expressivity of zero knowledge increase, we'll be able to do larger and larger, you know, neural networks up to and including perhaps one day like a human brain on the blockchain. But at least right now we're like, you know, we're not at GPT-2 levels of zero-knowledge machine learning. But that's at least the trajectory. And so that's grounding this project in let's call it like a speculative futurism of, you know, like what are the trajectories of these technologies?

Perhaps, I don't know, left to their own devices or rather they're like following their own like logic, what might be some of the conclusions? And I think one of the most One of the foundational conclusions we've come to and that kind of influenced the the construction of the sculptures and then the digital part, which I'll touch and maybe finish on, is this idea of like a solved entity that um let's say if you had enough biometric data exfiltrated from you or if you were like most more If you were adequately enough modeled in a computational system um such that let's say, you know, to put it in technical terms uh or I guess philosophical terms, like your cognitive light cone, you know, would be like completely encapsulated by some external computational system, meaning that like there's some other entity that knows everything you ever could possibly do before you could do it and can like plan accordingly. Essentially kind of like a matrix hypothesis, right? But like I feel like this this a matrix hypothesis is simply, you know, a matter of uh the availability of data. You know, there's like these little worms that have 200, 300 neurons in their brain and we map their brain and do all sorts of computational modeling.

So are we like do we have that relationship to them? Like do these worms still experience in their own sense of self some like innate free will, but all of a sudden we have the ability to like completely computationally encompass even like the neural structure of this organism. It's like, "No way it can breach out of our like possible cognitive containment." Does the worm's experience change? Probably not at all.

But it's cognitive light cone is at least from our perspective, or at least we can say it is, you know, has a certain property and relationship. Perhaps that property was already there and we just need to like kind of just discover it. And perhaps that property already exists for us to some like larger system and maybe that raises some implications as well. So this idea of like a solved entity to be computationally solved, to be um like able to just have your cognitive light cone mapped by some other system. That was interesting to us of like what's what's the worst possible case scenario of like all of your biometric data being exfiltrated.

And so that was kind of the inspiration of yeah like the large statue on the on the floor, kind of lying in prostrate. We call it like the solved entity. So I, you know, has like a cable running from one side to the other. It's like no longer even if it has perhaps maybe nothing from its own perspective has changed, but nevertheless it's like modeled as simply a a pass through where data goes from one end to the other. Or at least like some entity can model it as that.

And like what what exactly does that mean? Does that mean anything for the entity itself if it has no idea about it? Um that these are things were interesting to explore. So that's that's this um level. Then working on the uh digital component, which we're still uh working on, is uh using Actually what we wanted to do was like use a uh a face scanner.

So still running with this Worldcoin idea of using like a zero-knowledge machine learning uh face scanning to do like a proof of face. And so you could okay, probably put your face pixels on the blockchain. Is there a premium or an interest to you know, probably organic pixels? Does the fact that this pixel came from your face or like these amalgamations of pixels came from such and such as biometric data in and of itself make those individual pixels more interesting and compelling. Well, obviously that that is that is the very nature of biometric data.

Like yes, the fact that it came from a certain source, the fact that it has a certain provenance chain carries a certain interest and a certain value to the data. And so we want to like play with that aspect. And we realized quite late that perhaps with like GD whatever PR laws, you know, that's probably a non-starter with real people. But with like synthetic, you know, AI people that can pass you know, like a facial discriminator, you know, such a system can be built. We can still like play within these confines.

So also like yeah, playing also what what are like the legal implications of actually using these technologies already is like coming quite to quite to bear, right? In terms of like who owns the data, who has provenance of the data, you know, who can sue you for use of that data. Actually these are things we have to be like quite concerned about. Um And yeah, and and and the idea is like using this data in order to create some new platform where it's like okay, I could take your forehead and you know, like stretch it out and rotate it and take this person's eye and okay, and I can make a little like pixel boat or like something. Okay, some composition.

But in the end you could say like all of these pixels, all of these parts have a provenance that's you know, some like organically sourced. And that there was like a transmutation of the data from people into something new. And perhaps that whatever that thing that is new, it has a life, has an existence apart from and beyond the sources of that data. I think that's a good place to stop.

[applause]

I if I don't want to become a node providing data and become a list of metrics, is there hope?

Well, I think that's a it's like it's a contextual question, right? Is there hope for the flatworm? Is there hope for the C. elegans worm? Like perhaps it it it's like it has lots of hope because it even doesn't even has the capability to, you know, conceive of the amount that it's could be possibly controlled.

Perhaps in some way that that's our condition as well. And but as, you know, we're looking at what's like is there a cause for like like relationally um excusable existential dread that there's like some entity out there that could like completely like dominate us through like, let's say, the exfiltration of our data, um but that perhaps we had a fighting chance against. If we never had a fighting chance, perhaps that's also quite dreadful as well, but, you know, at that level maybe we we would never know. So, your question was, is there hope? Yes.

I mean, the hope is to craft an interiority. I mean, that's the other side of the zero knowledge or like these programmable cryptography is like if you are a some entity, an organism, and you're existing in such a world where your data can be so easily exfiltrated, then like the only way you can maintain a selfhood, let's say, be apart from becoming dry, exfiltrated of your data, is by being moist, having data that is only um knowable within your own like autotelic system. So, I think that's that's the key, and that's where programmable cryptography comes in because you can define the boundaries of information transference. So, you can say, "Okay, this data cannot leak outside of my interiority." And because it cannot leak outside of my interiority, it cannot be sufficiently mapped by an external entity cuz I have like too many internal degrees of freedom.

Like, they would need some sort of, you know, like inside look. And as long as I have that shield, I have enough degrees of freedom to avoid, you know, like a terminal cognitive capture. So, then yeah, the the answer comes back down to like programmable cryptography as well, I think.

No one? Um Okay.

Thank you.

Yeah, that's Thank you. Uh um I'd love to hear it.

[applause]