Satellite Expansion of the World Computer

Transcript

Next we would like to invite our next speaker, Mr. Daniel Exponent. Mr. Daniel, the floor is yours. Testing one, two, three. [Music] All right, I have 20 minutes. I could have given a few to Franchesco, but we'll uh have to get through it. So, thanks everyone for joining after lunch. I hope it's recorded because some people are not here. So today I'll show and present the both the rational motivation behind building the satellite expansion of the world computer. As some of you

know Ethereum is the world computer. So we're building an expansion of it in space as well as go a little bit into how we're building the space computer stack. So very quick uh brief history of um compute stack as a whole. So originally when um compute started being developed as a mechanical or automated operation, it was still task specific and you had computers that were entire rooms dedicated to a single computer that doesn't even do general purpose computer. It does just task specific.

This is I think an image from the Bletchley labs and not long after Alan Turing came up with this concept of general purpose compute. So he basically took the rational of using in that in those years electromechanical components to execute compute. Um the motivation behind his invention was to crack the enigma code and and win the second world war. But I think that as a mathematician a part of his motivation was to introduce this concept of general purpose compute. So essentially something that can be used for any piece of code hypothetically. Then what further scaled this whole compute revolution was the invention of the

solid state transistors. So solid state transistors allowed the miniaturaturization the race to pack more and more transistors onto chips and and now we basically have like billions and billions of transistors in our pockets and uh Mo's law is kind of getting to its limits but yet we're inventing new generations of transistors. Afterwards came the high level programming languages which enabled people to not even have to be aware of how the actual compute happens at the machine level and they could just write the scripts which is what we're doing today and ultimately the graphic user interfaces and oss which is uh essentially like you don't even need to know that there is a code you just click

stuff and interact with it in the uh the same way that you do things in life outside of compute and Then came the internet. Uh this is two anecdotal moments of of history from the internet. One is the IRC chats. Another one is uh the at the time the leaked video from Pamela Anderson and Tommy Lee that became overnight the most viewed video in history at the time. And it kind of demonstrated the power of both the fact that compute systems became commercially available and ubiquitous as well as the interconnectivity suddenly unleashed commercial applications that we just didn't foresee happening before.

Definitely Alan Turing didn't envision that this would be the first application for the compute I imagine and over time we became really good at digitizing trusted parties which is already going towards the the crypto side of things. So we essentially started having systems that provide us services whether it's banking services whether it's eventually e-commerce social networks and then rather than having to run um servers on our own or or operate website and offer services on our own in servers at home we basically could rely on those very convenient entities that provide us with the

specific applications But this also meant that we have a centralized trusted party. It was all looking good, right? Until it didn't. And basically that was kind of the call to action to start what today we call crypto. Um that's the famous quote that is in the Bitcoin Genesis block where the chancellor is on the brink of second bailout for banks. And essentially then we go into a new rabbit hole of comput stack which is the cryptoeconomic compute systems and cryptoeconomic comput system um essentially it's

trusted party except it's distributed it's no longer a centralized trusted party and then it's a on top of that it's a computer rather than having a um a server that is the computer we have a distributed network that function as one unified computer. So let's take a quick look at the um the timeline of innovation in the distributed computer stack and how the trusted party served there. The first one was Bitcoin. Obviously Bitcoin was supposed to be a task specific computer with a single application being money.

So peer-to-peer electronic cash system nowadays as we know really uses it for payments. But it landed on the use case of store of value. So today Bitcoin essentially is a task specific computer for go digital gold. Afterwards um there was some uh a bit more avanguard and uh innovative approach. So similar to the concept of expanding from task specific Vitalik came with the idea of expanding the distributed computer into a general purpose computer with Ethereum. So essentially a touring complete distributed computer environment and we have all of the wonderful miracles that

we're using today whether it's DeFi, Dows, NFTs and all the shenanigans and interestingly since Ethereum actually there wasn't something really novel that was invented right like if it's the new networks that we're seeing out there typically all that they doubt is that they're faster or they're cheaper or there's better casino applications on them but actually there's nothing really new right nothing fundamentally new so to that four of the areas that I believe are massively under addressed in crypto privacy which we'll talk about after this talk in a short panel inclusion so essentially accessibility to the

frontier of um global innovation and and the economy for areas that cannot otherwise participate resilience so whether it's resilience from a game theoretic point of you or if it's resilience from attacks and and hardware vulnerabilities etc. and then security which kind of relates to it. Um those are areas that are massively under address and we always hear about edge cases where they are constantly being exploited. So the question is, is the world computer the final frontier or do we have more to explore?

And my argument is that the next frontier is the reaceleration of the space race. So if we're looking at only in the last two years or so, we launched more satellites into orbit than in the entire 50 years combined of previous century. So essentially there's more and more satellites being launched, more and more applications being uncovered and with the talk of Starship recently we had the experiment of um what was it called splashdown I think where they demonstrated recovering some of the boosters or what we're talking about expansion of the payloads of more than 400 times so like orders of

magnitude and that means that the launch cost is going to be converging ing to something like tens of dollars per kilogram. So that means it's no longer something like space as a domain is no longer only relevant for state level actors. It's something that a couple of people in a co-working space can start looking at building for space. Right? So a lot of new innovation is going to be unlocked in the space frontier and some of the things that we're already familiar with. So whether it's GPS that everyone uses today, whether it's for um location based services and applications on your mobile, remote connectivity. So Starlink iridium personally I live in a remote region and I rely on uh those services in order to connect to the internet and by virtue of having those

services you suddenly have the remote regions of the world connected in a competitive way to the economy. Um, Starlink in 2024 launched 7,000 or more than 7,000 satellites which in in uh history it's basically unimaginable numbers and these numbers are only growing. We're looking at probably quadrupling it over the next few years. Planet Labs uh is basically uh already a very widely used service for different whether it's governments or supply chain applications. So you see that more and more the space domain is becoming important for the economy. Now how can the world computer concept be

expanded beyond the bound of earth? So these are some of the two like some some of the examples we're working on in space computer working with developers to use cosmic radiation as a source for randomness. So randomness beacon based on um true cosmic radiation. And one example is a password generator. Another one is a raffle. So essentially a kind of a lottery and yeah I'll I'll go now into some details of how we're building the space computer and what is the space computer so space computer in a nutshell is a high security guarantees based on physical isolation so essentially you have a temper resistance computers aboard satellites that each of them serves as a space TT stands for trusted

execution environment and eventually we'll build a node network based on them so what is space computer in a nutshell it's a robust layer 1 in space, a network of those space T nodes that is supporting SP smart smart contract and ultimately aims to serve as a marketplace for payments and for satellite services. Now, taking a little bit of a deeper look on where we're at today, we have already uh access to one satellite. So, we're starting to look at a single node and single node that functions as a TE. For those that are less familiar with TE, the rational is that you have a secure environment separate to the typical CPU. And this secure environment can also remain secure even if the

hypervisor is is hypothetically malicious. And what we have today is that you're interfacing to the satellite via a space compute gateway. So if we're looking at it, you have services that support users on earth with high security guarantees whether it's for the layer ones, the layer 2s and then ultimately looking at uh relying on more than one communication provider. So um it can be abstracted whether it's interfacing to Starlink or to Idium or to anything else and essentially you can look at it as some kind of RPC aggregator for space gateway.

Um yeah so so orbitport is currently the somewhat uh hosted service that we're operating with applications being the cosmic to random number generator and eventually expanding to more features and ultimately we want to turn it into something more trust minimized than being hosted and operated by us. So uh that means that we'll we'll look at multi-node essentially moving to a network constellation uh that consists of those um services each satellite services whether it's randomness whether it's um MPC so that satellites can be assigners in a threshold scheme and uh yeah each of them essentially functions

as a te uh to scale that we'll be looking at uh either the optim optimistic or the ZK rollups. So obviously there's trade-offs with each of them, but a part of building that stack will rely on um on those existing uh existing components that we're familiar with from today how Ethereum L2s function. Uh essentially the L1 in orbit is somewhat of a capacity constraint network like Bitcoin. And then in order to augment the performance, we're looking at using the similar technologies that we have today for the Ethereum network.

and uh some some rational for how it works. So we basically have here an example of relying on the satellites from a company called Optos orbital for the randomness and then a user essentially interfaces to those satellites via the orbit port space comput gateway and then the actual randomness is being fed through the gateway and the applications consumed here on earth. Um guess I have not so much time quickly on uh orbit port as the the gateway for space compute. We're currently enabling the uh city RNG. Ultimately we look at custodial wallets and then uh some of the interesting applications for people

who are network states. So imagine that you have a Dows that are doiciled outside of legacy nation states. Essentially, you don't have a country host for a jurisdiction or you have a programmable cryptography toolkit that you're able to um interact with that would be deploying contracts directly on satellite and then eventually once you have a full constellation you can start looking at bridging all compute services from different satellites. Um these are some of the applications. So whether it's space vaults def and dows that that are running on satellite dimension and programmable cryptography some of the road map. So with one satellite today we're soon completing the specification for the protocol. So

we'll be looking at how to actually operate the consensus and execution clients aboard the satellite then um build the earth component stack for uh interfacing to it from earth. um ultimately scale it with the L2s, ABS's and then eventually work towards the constellation that would enable the full uh space T network. And yeah, with that, if you want to connect and and you're interested, please uh reach out. I'll be here for the rest of the couple of days. Cheers. Do we have time for questions or it's not the All right.

>> So, any questions? Not much. Okay. We can just move to the panel then. >> Thank you very much. Yo.