# Open Tech, Blockchain, and Settlement by Robert Drost | Devcon SEA

- Speakers: Robert Drost
- Channel: [Devcon](https://streameth.org/devcon)
- Date: 2025-10-07
- Duration: 26:34
- Watch: https://streameth.org/watch/yt-vfAgwVjLonc
- YouTube: https://www.youtube.com/watch?v=vfAgwVjLonc

## Description

In this talk, we discuss the what and why of open tech, starting with networking and the internet. Using the recurring progression of tech to openness, we explore the critical classes of commitments and settlement that enable blockchain to accelerate open coordination of finances, tech, and society.

Speaker(s): Robert Drost
Skill level: Beginner
Track: Coordination
Keywords: fork, Consensus, Coordination

Follow us: https://twitter.com/efdevcon, https://twitter.com/ethereum, https://warpcast.com/devcon
Learn more about devcon: https://www.devcon.org/
Learn more about ethereum: https://ethereum.org/ 

Visit the https://archive.devcon.org/ to gain access to the entire library of Devcon talks with the ease of filtering, playlists, personalized suggestions, decentralized access on Swarm, IPFS and more.

Devcon is the Ethereum conference for developers, researchers, thinkers, and makers. 
Devcon SEA was held in Bangkok, Thailand on Nov 12 - Nov 15, 2024.
Devcon is organized and presented by the Ethereum Foundation. To find out more, please visit https://ethereum.foundation/

## Transcript

[Music] [Music] Hi there. I am Robert um and I uh work at IEN Foundation. Um just uh let's see everybody. Okay, looks like everybody's settled in. So, let me get right started. So, there's three ideas that I want to go over in this talk. Um, two of them are just giving some background which are pretty important for open technology. I think a lot of times when we're sort of seeing the evolution of things, it feels like it's a glacial pace but the glacier does eventually grind down mountains and the grinding of mountains. I want to look at the patterns that we see from that and especially because the latest and greatest uh version of open technologies has been crypto and uh I think what I'd like to form in this talk is the kind of underpinning of what kind of openness results and what we can expect to see in another five or 10 years uh based on these two together. So um giving the background on open um I thought a good way to do this would be to just kind of show people how much has been accomplished for going from proprietary to open technologies. Uh the top one there, networking. Um this is kind of predating all of us, but there was like 20 years or so of proprietary networks if you could believe it. There were like 25 different networking standards. And it's shocking now to just hear, but like TCP IP was completely the renegade kid on the block when it came out. Everybody thought it was such a crappy protocol. All these other protocols were so fine-tuned. They did all these specialized things and it all went away. I mean this is the same era where you know you had proprietary uh uh uh character formats. ASI was created to kind of get away from that and so forth and uh so that was sort of one of the biggest u first moves into um uh openness. We had a couple other that uh a bunch of us have experienced maybe not the internet but the mobile phone revolution where we basically got the entire internet carrying with us all the time. not value necessarily, but at least the information sharing and information downloading. Um, and probably one of the biggest ones that I would say in general for all of us, open source. I mean, open source literally was called by Bill Gates a cancer. Um, that destroyed value for everybody in the industry and hardware public clouds. I'm going to actually say a few things about those in a second. Post blockchain. Um, I put approximate tipping points down on the top. Um, before last week I actually still had Bitcoin in question marks, but now I'm tentatively saying 2025 may actually be the tipping point. Tipping point means when it's so prevalent that it's just all over the world, you know, like the mobile phone going from before to after. It's an entirely different transition. We haven't had that yet, but maybe next year in Bitcoin. The rest of them, uh, you know, going to DeFi, ETH money, crypto keys, identity, um, walled garden data and public clouds. we haven't hit the tipping point on those. That's some of the most exciting parts that we can expect coming with that transition to open tech meeting uh crypto and crypto settlement. So, um let me just jump through a couple here. So, hardware and software tech meeting the cloud. This is a really important one. How many people in the room are developers? Okay, a lot of us and we all develop on the cloud. But it is really interesting to see that you know for about 30 years in computing it was all so proprietary. everything on the left side you had to buy you had to manage it had all sorts of different standards you were doing different APIs for everything all the time and over time people figured out that there was a much better model uh virtualization other software allowed you to abstract and you had infrastructure as a service uh platform as a service and then software as a service and progressively they unlocked some of the greatest software innovations that we've ever seen um open one in the end um however winning on one level Doesn't you mean you win on all levels? So the cloud itself was open but the cloud platforms quickly became pretty centralized with the top three controlling like 70 80% of the whole market. Result of that which has been um an interesting problem as a side effect that I think blockchain is going to help us solve a lot is that before clouds open source had this great model where you built open source and then you made a product based on it and you actually got the value from it. um in the web 2 public cloud that stopped happening because now what would happen is that your open- source software because it's open would be used by cloud vendors and so if you look at a chart like this where Google cloud has all of this SAS and platform as a service and so forth they're just wrapping somebody else's work and putting it out there so not fair in my opinion um but the crypto cloud this is one of the most exciting parts that we're all going to experience we've seen it already blockchain you write open source software token in a protocol and the team that built it is making money again hasn't extended down to the public cloud yet. DPIN and other things are bringing it in that direction. The beautiful part with that and I'll talk about it in terms of like how settlement is uh going to really enable this. This means now that we don't have a central cloud player anymore. We have a whole bunch of different infrastructure uh providers, tens or hundreds or what have you. And so none of them are going to go ahead and actually dominate the field in the way that you saw with the big public cloud vendors. So open will win. Um, okay. And then one last point I wanted to just say on those is that sometimes when I say this slide afterwards, people come to me and they say, "But this doesn't quite make sense. Web two public clouds are quite decentralized." But as we saw with the internet, the internet launched in the mid '9s by mid 2000s. this open technology of reading, writing data became entirely owned by a couple of big social networks, advertising networks and so forth. So you took a decentralized technology and owned it. And in the cloud case, the billing system, the accessibility to put in your hardware, and most importantly having the rights to get the revenue from all of that was centralized. Um and so the unlock or the openness in this case is being able to participate in the value generation and the value revenue and uh that's a big theme obviously with crypto and extending into technology like public clouds is quite valuable. So um okay that's sort of a background intro. I have one slide where I just like to go through the why of open tech because it's not always obvious like when you look at open innovation. Um it it looks immediately like maybe open innovation doesn't win. Short-term closed systems can scale really fast. People are able to develop on them. There's less coordination friction. They're able to take the revenue, reinvest it really quickly. But the reason this causes a problem is eventually the closed systems when they start winning a market, they slow down. they don't have a big motivation to uh innovate the way they did. They're just protecting their monopoly. Longer term, open innovation moves ahead uh bringing in a lot more people to it. We've seen that all over the blockchain space. We all use each other's protocols. We literally compo compose applications out of each other's DAPs at this point. And the feature velocity increases over time because as you kind of settle Ethereum is settling its uh oifying its its layer and it's kind of like uh um accreting itself to something solid and then people will build more and they'll build more and so forth. And the last one which I think has been a really dominant one is you know how many times do you find out there's actually no diversity in the thinking the values the knowledge I mean it's kind of nuts when you think about a big network like Facebook being you know really not localized and instead having this one projection of sort of values and one projection of the system all around the world. Um that doesn't really seem like what you would expect uh would be a good way for humans to interact. Okay. So now let's get Okay. And it's unstoppable force. I just like to say I used that phrase tipping points before. And the tipping point's really important. For a long time, it's like the Buddha story. If you have a reed that is very inflexible, it seems hard and then it snaps and it breaks. Whereas the one that actually is soft and pliable is able to, you know, bend with time. So the fragile short-term ones um those eventually do snap as we get the entire um uh you know weight of the advantages of open innovation. So um now talking about crypto one thing that I've kind of reframed recently that I think is a pretty good framework. Um I don't know how many people here have heard of you know money crypto tech crypto that division I kind of reframed it in two and it sort of exposed that there was a third one that's really helpful to put together is like the three classes that we have um for settlement in crypto. The one that started it all of course was financial settlement getting the readr own. Now you can actually own digital assets and Ethereum expanded a greatly to programmable money with uh all of those DeFi protocols and tech settlement. There's a few different cases with tech settlement where um uh you have sort of more limited and we're going towards more open and generalized tech settlement. And then finally, I don't know if uh a ton of people think about it uh all the time, but there's this meta settlement concept, which probably is the most important and underpinning one. I want to say a few words at the end about just how substantial that is in terms of making sure we don't end up having what happened with clouds or other things where you've got open technologies and at the very top you have somebody that kind of monetizes and closes a wrapper around the whole thing. So, um here's a slide. Forgive me. This one has a couple of uh labeling from uh the company that I work at, but it's a general principle. The top side of the um the top side of the uh picture is showing us the Ethereum trust network. Um uh and then the bottom side is a new branch of peer-to-peer networks that are enabled uh with reststaking protocols. Uh the top one, just to level set for us, you know, when you think of Ethereum, the reason it works is it has two portions. It has a consensus portion and it has a execution virtual machine or smart contract portion and it's split up that way because the consensus machine is something that is an entire bunch across the internet of software nodes and they talk to each other at full TCP TCP IP seat speeds. They're sending commitments, they're sending data blocks, they're sending transactions, all of that stuff. And so that is a distributed software application. The EVM is now somebody using that distributed software application that manifests an EVM to write something that gives them exact execution of a smart contract. So very valuable switching and so that means DAPs that are based on that they're very flexible but they basically are inside of the realm of a smart contract system EVM execution. They don't have access to directly being able to program the consensus itself. So the sort of unlock that we've gone to now, let me actually jump to the next slide. It's maybe easier to see. Um is that you know we've kind of gone um now I'll do it from here. So we've gone from things like Bitcoin which were very monolith applications where all the way from the security of proof of work to the consensus of Nakamoto consensus to the execution of UTXOs to the application of sending money around the world. Those were all one case uh one use case and then obviously Vitalic and the team at Ethereum invented that you could generalize that. So now you would be able to have an application rather than just sending money back and forth and that enabled us to now have flexibility at the application layer. The next big innovation from uh I guess around four years back at this point was the rollup era and recognizing that the execution VMs themselves could be generalized. So now you had the ability that you have different rollups building them on Ethereum. It's been great. We just heard Justin talking um about the beam chain upgrade and that's all going to sit in the context of the current execution layer. This final innovation here um what I think is the big significant one is that we're opening up the layer of peer-to-peer um technologies so that you now have the ability in your applications to have whatever kind of distributed software. It's as if you're actually able to put in something that is at the same level as the beacon chain except it's running your application. So, you know, you think about uh storage, content delivery networks, you can think about databases, other things that need very high performance, but they still want to be settled with the open uh settlement. Uh they want to settle their technology onto blockchains. Okay. So um and this allows importantly now that there's a framework that you know similar to like EVM nodes being run by people on uh uh Ethereum uh this allows operators to now look at this model of consensus as another thing for them to run. So they sign up, they bring CPU, GPU, uh trusted execution environments. They can have varying node requirements, maybe as low as Ethereum, maybe 10x higher than Salana. Um because now we freed up the peer-to-peer protocols could be to be anything. They have all sorts of different assumptions. And this then allows different operator groups to perform different subset of tasks. And we take all of the people that are operating right now, a pretty limited set that they can run and they're now able to essentially run everything that's on the decentralized public cloud, which are in fact blockchain protocols. So, it's a huge unlock. This sort of open meeting crypto means that essentially all software that we run for all purposes uh can move over to blockchain. It doesn't have to just be fixed to the kind of execution virtual machine layer, but it can go down all the way to the uh hardware layer of uh uh you know as general software protocols as you can create. Um and just give a feeling if you look at these types of things, these are things that you would not expect to see smart contracts as a settlement platform for. There's a lot of stuff that goes into like proof of location, proof of machine, all sorts of pinging, all sorts of latencies, um, running Linux, running ZK provers, a lot of off-chain compute, and all of these are actually being built now, which I think is really exciting. If you look at it, um, you know, we're sort of at the level at this point where there's hundreds of new blockchain protocols that are being built making their own peer-to-peer network and not having to, like Filecoin, create a whole new layer 1 network. They're able to create these things as kind of layer 2 networks on Ethereum, mixing together peer-to-peer networks plus EVM execution as well. Um, oh, and then I'm sorry, here's the slide where I said the hundreds in dev. Um, so yeah, so quite quite a few of these and um, yeah, the meta settlement, this is the final point that I wanted to make here because I think a lot of the points there, they paint a really good story. Um, I think we can be excited about open technology having won every major um, technology battle in the end, sometimes taking like in networking case more than two decades. But it goes the right way. Happily, blockchain accelerates those things. Um, and um, you know, those are all great, but at the very end, there's this question of like who owns it at the higher levels. You know, we we kind of know this idea of like network capture is such a problem. Um, because it's very hard to get out from the friction of a system that has kind of captured your data, your identity and so forth. So um so two two ideas that I just want to present on that um and then I'll get to the final conclusion about the metal set meta settlement one is that you know we have existing in kind of normal mode of blockchains a lot of things that are onchain observable faults this has really been solved by blockchain including zk proofs and so forth that you don't have to worry about it if somebody can give a proof that anyone non-interactively would be able to read, see that the proof checks, see that somebody committed something incorrectly or somebody performed some action correctly, you can reward them or slash them, whatever is appropriate. The challenging part of these interubjective beliefs and these are sort of human observable faults. Um, not sure how many people have dug into them, but things like data availability, you can never prove that somebody didn't show you something because while you're claiming they didn't show you something, they can show it. And then it's a he said they said of did the person really not show it or did the person actually uh you know or did the person um uh actually show it after the uh threshold. Prediction markets obviously any oracle I mean there's nothing on chain that would tell you on chain link that bitcoin is less than $10. If it published that defi would break and there's nothing on chain that would sort of detect that that's happening. And you know these kind of examples go on in the blockchain space and I put at the bottom one this idea about promises which are general credible commitments. Those are really exciting because you think about it, you know, what is a promise? If it's between me and this gentleman over here, you know, I can tell them, hey, you know, I'm going to make sure that uh, you know, anytime you need when you call me, I'm going to come with the car and get you, right? And so I made this promise. Um, it's onetoone, so it's not going to be interubjective, but it's something that's happening in the real world. Intersubjective would be like, I'm creating an Uber alternative, and I'm saying Uber's take rate of 60% is horrible. I'm going to go ahead and my take rate's only going to be 5%. Um, and that's great. I could write an imu immutable contract, but now I'd have a lot of problems with that because can you really write a big software application without ever changing it? If you can change it, which invariably pretty much all of crypto can right now, most of the systems, I mean, they just have to have upgradability somewhere. Um, then that 5% is no longer true. But if you have the ability to check for human observable faults and then translate that information on chain then that actually takes care of that problem because as society we can go ahead and say hey this person violated this or let's say Facebook says new version of Facebook someone competing says we're only going to change the privacy policy if the community agrees like a DAO 51% that we're going to approve a new um privacy policy. If they violate that, then the users can say, "Hey, you know, you're sending me a ULA. I'm supposed to either sign it or get the hell out of the application. You broke that promise." So, what actually um allows us to do that? Well, in L1's, this is probably the most core feature that without this, it's really not blockchain, which is that you have viable slashing by chain forking. Meaning that you can take the existing system, you've got your data, you've got your identity, and you can just fork it and create a new version. We've seen that many times in L1s. It's a very unique feature. And one of the problems that we've had is we build a lot of things on something like Ethereum, they're all L2s. And Vitalic is really clear that forking L1's because of L2 faults are an issue. So um one of the more recent breakthroughs that I'm not sure uh everyone's heard about, so it's this idea of slashing by t token forking. And this is the idea that you're able offchain to make um agreements and then project onchain that something was violated. When it was violated, the actual value of tokens in DeFi, in unis swap, in compound, a everywhere automatically gets shifted to the new chain and effectively slashes the people who uh violated that off-chain promise. And importantly, it's a lot more fundamental than onchain voting. Onchain voting still ultimately is following a mathematical protocol. So um and this one is uh and not able to do the token forking. Um and this one allows us to do that. So the final comment I would hear say is that like we really don't have the openness we want in general. We never would have it um before crypto and this meta settlement. Um once tech platforms capture the network effects, they become opt-in autocracies. You either use it as is or you just leave. and the forkable tech platforms. This is the final frontier of open technologies enabled by blockchain and that's where we're going to get to. Uh global internet native self- enforcing coordination. All right, thank you Robert. Thank you so much for your presentation. For everyone that's here, you still have a bit more time to ask questions through this QR code, but I do have one question for you. From all this presentation, which things uh we could expect that would change or become concrete in one year, three years or five years? So, how do you see maybe like in one year from now you're coming to DevCon or Dev Connect again and repeating this presentation or I think the one of the most promising places for all of this to show in the next year or two and in my mind it's kind of like a call to action for all of us is that um big moves like to blockchain they definitely are helped by things like DeFi but the most fundamental changing uh technology for humans since maybe fire is AI. I um we finally actually built the ability for us to have elastic scalable intelligence and in web 2 AI is a disaster. I mean it actually takes everything that's wrong and makes it even more captured, even more localized and but so this AI side is actually screaming for what I'm talking about here. If you have open technology, if you have non-captured clouds, if you have us creating the hundreds of protocols that AI is able to use, um, where there's not a corporation owning it, but it's owned by tokens in digital sense and rules can be made, we have much higher chance of being able to create, uh, AI that actually has the right kind of rules to be positive for humanity. In the end, people are like, oh, AI is going to cause this big problem, but it's actually humans, you know, evil humans using AI that are going to be the problem. so we get a chance to get them out of there. I'm pretty optimistic, but I'm an optimistic guy. I think we're already going to see major uh AI crypto advancements already around AI agents in the coming year. And in two, three years, I think that um you know, if we flipped it, I think it's possible we could already be thinking about a tipping point for web 3 AI um plus crypto. We're so looking forward for the future. I have a question here. How does inter subjective thoughts interact with politics and regulation in postgaring war era? Oh, I'm gonna sound really stupid now because somebody needs to tell me what is Gary Warren. You can shout it out, whoever put the question up. Oh, Gary Gensler. Oh, Warren Buffett. Elizabeth Warren. Ah, okay. My mind's over there. Um, wait, what happened? The question went away. Now I forgot the question. Oh, how does interubjective forking work in that? Um, so let's see. I think it's a extremely strong marker to say that the system is decentralized, which always has come to be the most important part of crypto, which is essentially establishing with countries that they don't have jurisdiction or ownership because it's a global technology. So my expectation is that it should be, you know, quite positive force. I also don't really know how relevant either of those two are going to be in the next four years, but they're still going to be around somewhere in the zeitgeist. Can you give a real world example of an AV as Yeah. So, uh a real world example of an AVS would be um something that's looking for uh the opposite of the touring test. So, proof of machine. Um there's a chain called Witness Chain that's doing that. Also, proof of location. It's great that we have, you know, kind of machines. We think we're decentralized, but if our decentralized machines are all sitting in the US and Europe, um it's not as decentralized, we would like to be able to have people uh participating all over the world, especially like Southeast Asia and Africa and South America. Um but if we don't have a good way to prove that, then we're going to have proof of, you know, we're going to have the civil attacks we would have in any case. And so um that's the kind of protocol you can't do in a smart contract level. It requires really intricate hard to beat very expensive to beat type of uh work going on at a network level. And so that's a great example where completely unique P2P um no technology is needed. Um I think another one when you say real world it's not there yet but it's uh definitely going to be something we need which is you know having great data storage helps but more important than that is having something completely competitive to cloudflare and cloudfront and so forth. So having really strong you know data will be published not just data availability in in the sense of sorry first being published data will be readable later on. So content delivery networks ensure that you're going to be able to do that over time and so we need that uh in com combination with uh secure storage layers and uh that's not something you can coordinate. Um, in theory you can and people would probably without this ABS model build it as new L1s. But the idea of taking all of that technology and having its security and having its settlement happen on Ethereum is probably one of the most exciting uh aspects of what we can do on Ethereum is to collapse more and more of the tech settlement and uh value and this uh these other components in. Robert, thank you so much. We are over time. We still have a couple of questions here but we won't be able to cover it. But once again, thank you so much everyone. and give every round of applause for Robert. Thank you. Well, time flies.