Society Github

Hi everyone. I've got a golden hacker sleeping, which I have a bit of a discussion, so feel free to interrupt with questions we're recording, but I think it's not a problem to have a bit more of a personal sphere. Okay, so, Society of Data. We keep expanding as a non-profit solution of how to build the future of Ethereum. And I started working a few months ago, and I've been expanding where the Dynamite ground goes and how we build some part of the Ethereum vision that was already discussed at the very beginning of Ethereum.

So, I would think, I guess, 2014 to 2016 when we were asked by the Ethereum Foundation people and the other sort of, you know, when they were starting, they mentioned a few things that will be possible in Ethereum. The idea is more or less the idea of NFTs, of prediction markets, of AMS for coming. And also the idea is that they like autonomous machines working through blockchain progression and that may evolve into something like iot or deep infrastructure but i think it's still this something that still remains unsolved and there is much to be all and at the same time this year it feels like all of the necessary technologies start to come together. So AI, blockchain, or wallets are all becoming created here to start building more of the decentralized world in the future. So we started to give the name of the company behind Men and Women, it's a company called Universal Solutions, since 2017 that's the name, and it's your first of a Isaac Asimov Foundation series.

And the idea was always to drive the humanity for all the challenges of all this choice power and the challenges of technology and challenges of this rise and fail of technology with our body and minds. And now we try to bring it a bit more to the front and say, all right, we'll start building this e-cap-like platform for society for the future of our lives, and we'll start now pushing a lot of effort into the research around the saving. So especially for hackers, when they look at the videos, maybe, or invite people to join us for experimental thinking research. So I mentioned, as I assume, that in the series of books, we're referring to this basic concept of what does it mean for a role to be set in the future? And there's this, like, a few sets of basic laws of robotics that define our intuition about what can go wrong and what is the most basic way of saying it shouldn't go wrong.

Maybe this should be like that. The most basic laws, rules that should cover any machine so it stays as a protector of humanity. And I think in many sci-fi movies, we've seen some idea of the robot always having some place that if you're disconnected, the robot stops. Like, even if it's a killing machine, you have to attack it. One control cue.

So I think that we'll start thinking as a technology industry about the kind of interfaces and verification mechanisms that will be exactly that module in the future. That single module that we inject into the machine, that any machine that will be produced and deployed on the streets will have requirements for some type of security mechanisms. And what I'd like to make sure that happens is that those security mechanisms are open source, are verifiable, and anybody can contribute and check how the machines behave. And can we as a society define what the roles of society are, instead of giving that back to the central governments. So, the same way as nowadays we start talking about verifiable computation, specifically for us, we might be working with Starter in StarHand, and for StarHand we revealed this idea of verifiable computation for Starks.

I think we'll be, and we already are, responding to verifiable hardware, right? We talk about CPUs, we talk about various ACs for CK, there are companies within Ethereum space and generally the blockchain space that are building the solutions. And I think those will not only be used in the data centers in the future, but they'll also be used in real-time verification of the behavior issues. And when we listen to predictions from Justin about the future of Ethereum, Justin does say that we'll have this real-time verification of the computation. That he will be sending any log verification with a ZK-enabled real-time.

And now, how we can get there, if you go back to the beginning of the open source movement that we let's go from, I believe it's from the 70s. This drives the idea of behind-building, building the Linux ecosystem in China as an open source endeavor. and how it's built gets a little bizarre because it says that all you can go under is centralized hierarchical post-development model and I think this is something that always appears in the world because generally when you look at it's quite efficient, right? It's centrally organized, all the inefficiencies are removed, it's ruthless, it's just focused on success. But excuse me, when it gets corrupted, when it goes to the wrong hands, that's when the society suffers.

And that's when you can end up with something that's stable to the guidance. And the other model is a bit less efficient, right? The prices go crazy, you may not always produce fast, but you still behind. It's far too defined consensus. If everything is open, you can compare it, you can build it, you can participate, it becomes permissionless.

And it's the same here as bizarre. I think whenever we build blockchain, whenever we build and define the future of blockchain, we should have the understanding that there are lots of inefficiencies and this is okay, because the world always needs to balance some part that it gives to the centrally governed systems for efficiency and progress, but also maintain some part of the openness. And it's just the collaboration between the two, the balance between the two, that leads to societies that are quite successful. I think that neither extreme of being totally open or totally closed leads to the best results. Now, while I'm with you guys here, I'm more with the engineers, so this reminds me of the defaults.

Just think about all these components that nowadays buildkit have. On top of kit, there's a source control system underneath, a centralized source control unit. It's just natural that you just expand it beyond the source control, beyond the sultry. And outside of pPAP, you have other people, you see you have PIPA, and also there are various platforms for templates for data training, I mentioned it, yes, a kind of thing for data. And also there will be blueprints for machines, where you can collect all of that.

What I imagine is very much possible is that the decentralized communities focused on decentralization and blockchain can start focusing on building a repository. We'd like to lead that effort to build a repository for a few things connected together. So the AI models that were deployed and we all defined how they are trained and how they behave. Obviously with all the probabilistic behaviors of AI models you cannot specify exactly all the different situations, but you can build some set of tests and understand in repository how this model will behave in particular situations, right? Or at least statistically how it will behave in particular situations.

And anyone can test it and think, like, you are for the sake of deploying this model to control the traffic, to maybe participate in the healthcare system, in the hospital, to do the checks, or maybe to execute operations. So it might be a system that is just software-controlled, or it might be a system that is deployed on a robot in a verifiable way, and to maintain, like, it defines the behavior of the program. And the blockchain integration here, I don't imagine it as a blockchain integration where the robot has connected blockchain all the time. I'm thinking of it more as a governance mechanism before the entire deployment happens. The source of funding for various projects.

But you can define a piece of infrastructure, physical infrastructure, and this component can be very, very clear. So you can start with something small, something larger, and then go to something as large as a city, or in the future, very distant future, maybe some kind of space exploration. So starting small, maybe, I'm thinking the first and last one is if we can, in a decentralized way, we found the pollinating bee and the bloat in some places where the bees are missing. I'm sure a few years ago we were scared of bees disappearing, but I think they're coming back. Maybe they'll come back in another way.

And you can think also of all the dark aspects of it. And I think the big question ahead of us is, are there evil things that can be deployed in a decentralized way? More evil when they are under control for decentralized entities, or more evil when they can be deployed by non-globals in a society? in a society and it gives the protection against evil things related to the problem chain. More efficient if it's executed by a decentralized government or it's more efficient when executed by the society in a decentralized way.

So in my recent conversations I was thinking that the society, when it's certificated in a collective way, reduces decentralization very often for protection of the collective, protection of the society. The units may always be using, well, not always, some of those units in the society will be using the technology for equal chance. So the efficient defense mechanism against the progress of society, especially with the AI happening, should be collected, should be open sourced, because open source generally attracts people who want to cooperate. So it should attract people who have this feeling that they want to use that technology for defense against people or for their progress. Well, this is the assumption, hopefully.

Now, this is the joke, but I'm thinking of something like deploying businesses, like franchises. So, imagine that you want to invest in something, like invest in a restaurant, invest in a dentistry practice, or invest in some kind of education facility. So you go to the repository, you select what you would like to invest in. It's connected with a blockchain platform. You fund it with like a million dollars, a few million dollars, maybe a few thousand dollars in something small, or a few dollars if you deploy a single lead.

And you press the button, and the platform is connected to manufacturers, maybe automatic factories being able to manufacture robots. Maybe some kind of self-replicating machine institution. And then you can deploy a set of governance mechanisms, the project management, program management, AI tools. You can define how to install the controlling mechanism for this robot and the verification needs, all the interfaces and then all of this is being built or maybe built in collaboration also in some aspects when 10 or 20 or 80% of it might be requiring some ordering of human services, right? So the platform can define how it opens the options on blockchain for asking people to deliver some services and some of the algorithms that says if it's delivered or not.

Well, many of those components we've been talking about many, many times in robotics and blockchain. And that is deployed and maybe you can participate in the revenue. Like say, anyone who visits can verify, oh, I'm about to visit a dentistry check practice, so it will not fix me, but it will tell me maybe I should go to the dentist. And maybe it can be super cheap and super efficient and verifiable. And I go there and I know that this is verifiable because I can check that, oh, all those machines that this unit is the same, I'm verifying them this model, and everyone says we're using them around the world.

And the officer's community consistently creates PRs and checks how this particular model behaves. And that might be why we build large-end robotic components this way. And especially if the innovation coming from all around the world can be deployed relatively cheaply anywhere in the world. So, the same way as we had revolutions of books being spread around the world and the translation of the books, particularly two weeks ago I was reading some interesting story about Japan's progress after Japan decided to translate all the books in the world, you know, 19th century or something, like technical books. So if you think about this, the animal books nowadays not only translate everything for us, but it can also start bringing from around the world core innovation, telling how to define a society in a well functioning way, and not in a closed way, so that you cannot copy it around the world.

so that you cannot copy it around the world. And so we can build entire cities on VSCO, or other platforms, education, healthcare, public transport, all the different institutions. We can imagine what can happen, like if you imagine surveillance systems, like the drones surveilling the districts, usually our idea is that this is a bit dark, like this is a bit dangerous. The surveillance, we think, I'm sure it's all deployed for safety, but also for us it feels like this is a risk for privacy. This is a risk for not being able to escape this set of learning approaches.

Especially if you think of smart contracts going wrong, then the code is low, acting against us. The code that is broken, and that keeps us prisoners, never being able to be broken out of. But what if you think of surveillance, but in a way that is verifying the privacy? So the system with AI, that deploys the drone and says this drone will never behave in this way, this drone will never remember what it was observing, this drone will never act against the protesters, like peaceful protesters, this drone will only raise the alert when the actual violent crimes happen. And anyone, any citizen, knowing that the drone is deployed in a city can verify a specific test against the repository and say, will this drone act against this behavior?

And maybe if it shows some particular test that is dangerous for society, you can raise it as a PR and say, change it. Let's change it together. But there is a very particular reason. And the cool thing is that usually when we as a collective, when we deploy these systems, when they are transparent, when we can show something that is clearly by the society seen as good, then usually we deploy them in the good skies, hopefully. Or, if they're not deployed in a transparent way, then probably they're already being deployed by the centralized institutions, by the governments that act against privacy.

So, in the second scenario, it doesn't matter much, because it will be deployed anyway, this way. It will be against privacy anyway. But in the other scenario, the former scenario, when you deploy it in the times when the society is active in a very transparent way, you're already deploying the system that is much less likely to be rolled over time. Those decentralized systems, I think they have the nature of being very persistent, very stable. The same way as smart contracts, they keep breaking, we see the hacks.

But over time we'll see smart contracts that will survive 5, 10, 15 years. And it's almost like a sediment of the most basic functions of the blockchain that we will be relying on. And over time we will have more and more robust systems that we can build on components that we've been trusting for many, many years. And again, maybe education in the future will not be controlled with a approach of what is right or what is wrong to teach the people. Maybe education is saying, look, anybody can check when you send your kid to school what that kid will learn at school.

Or maybe you can even choose between different approaches based on your cultural background, maybe you would like to add some piece of your culture and authentically verify that, yes, it will not be a gay baby. I guess especially for societies that are multicultural, you come to a different culture and you think, well, now that I said that, it's hard for me to imagine. I know that I did myself. But imagine if I sent these to a school in a very different society, but I think, won't they be very different from me culturally? They want to pass the culture.

What if I do the checkups and I see what exactly will happen? Maybe that's me acting a bit with the centralized system, verifying that's what the next generation should do. And the same thing that I already mentioned with healthcare, very similarly. Would I trust more if the system is trustless? Would I trust more if I know that this particular surgeon not only never made a mistake in the last 1000 operations on a heart, but also I can say that very finally it will behave the same way over and over again.

And that nobody started saving costs on paying less to the surgeons, or changing the system so it introduces some new treatment that is unchecked. What if I can verify it and go to this hospital knowing that for the last 50 years it kept behaving the same way. I guess that the maintenance is a question, but that maintenance itself can be defined for the AI system as well. Well, there's something just really wrong and it's very likely that you can end up with some kind of node packages deployed in the scene if you connect all of those different GitHub repositories. Maybe the the packages is the problem within modern cities, maybe it's unavoidable we think that we're so smart in the open source that we'll be so organized and clean, but no, not.

And, yeah, that's it. Let's just avoid the overly-bought systems, let's keep improving, let's keep everything open, and let's keep building this on Ethereum. I think this will be part of our plan, always. Thank you. We have 10 minutes for questions or discussion.

I'm sure someone would like to ask any questions or just explain would this thanks for the talk this was amazing would this talk require does this future assume some do the people who live in this time period, in this kind of setting, are they all some level, to some degree, just natural engineers? No, the same way as you're getting out of society, when we have open source software, not everybody's participating in building that shelter. It just enables some individuals that want to participate in building a society to participate in it. And I think that as we've seen out of the systems which say, we'll get no longer anything to cope with when it gets recorded. And then we repeat the question, and question was, is it required in society, or if you want to be an engineer?

So I think the modern technology already shows that you no longer are supposed to be engineered to code. You can start reading and participating, building those platforms, not just in the language, in the scriptural language. Then you need to combine, probably, the ability to define a society, various aspects that can be an expert in healthcare and education and so on. And combined with those, we'll be able to translate it to disability frameworks and so on. So you'll always need engineers, I guess.

Well, hopefully it will not be replaced in beta. That's always good. But this is just participation of our society. I think 5% of 1% of people will build for everyone the same way as nowadays. Not everyone has to be a content creator, and everyone is a content creator.

We generate that for for maybe like genetically standardized scale builders and that's our SME for that sort of well being for the tribe, right? So the answer is between FDB and SME. Do you imagine this GitHub deployment platform to be fully on chain or would be like some parts of it? I think the majority of these nowadays can be on-chain. In a sense, like the prompTents as usual, nowadays platforms for decentralized applications, you would usually apply a centralized server for the frontend and then possibly provide an alternative IPFS-based or other format decentralized-based frontend.

And then you open source it and allow people to build more frontends and then only I guess many builders of this application would benefit from operating the most successful frontends in code. Like infrastructure, combinations, so we need to storage space for storing all the data. I think this is progressing as well in the industry. I think this aspect is going to be at least well specified. We're still at this stage where we're struggling a bit with storage to launch lots of data in a decentralized way.

But with verifiability, the question is, with storage we're struggling with availability. This might be less of the case here, because of the centralized repository, the amount that they have. A bit of a tendency of replicating themselves on multiple machines to developers who are DC-laying, so that it's very very naturally being copied and replicated between different storage devices. So I think the availability might be less of a problem, and then if availability is the problem, then with verification mechanisms, you can verify anything. So the storage is not a very problem, and the front-end can always be recreated from the description of what they think that is.

It's not like it's forever lost. I think generating frontends in the future will be easier. So my expectation is the first versions of this platform will be widely the full frontend of the ROTI with the definitions of the founding mechanism of logic, whether it's there directly or to a material, and that's where we are going to be able to access these specifications. So just curious, a really nice talk but I missed the first part, but just curious to hear your thoughts on what do you think are some really good use cases that can be enabled by Ethereum that you think has a lot of potential but maybe kind of really are underestimated by a lot of people. The use cases are, I think, mainly launching businesses because it's a very simple use case.

And people almost skip it because they think it's not big enough because everyone wants to solve the large problems of the society. But people usually just look at this and think, oh, can I make money with it? That's how blockchain really often succeeded. I've thought about DeFi, about some deployments and smart contracts, and people came over and asked, can I make money with it? How do I participate?

Well, I'm very worried. And then if you can keep improving the system through allowing people to deploy businesses this way, like simpler businesses, more complex businesses, then this will be your training way of allowing those who want to deploy clearly public goods or have charities that are much more efficient and cheaper because the system, the platform already will be well specified. I think very often the projects that are approaching charity immediately are failing because it's really hard to finance the charity platform. But if the charity is mostly a natural side product of the platform that is supporting something that is very business-like, business-focused, that is a great benefit. So this is underestimated.

The fact why, why we build the systems that are from the very beginning focused on making money and revenue. We do that because this leads very naturally, to incentive mechanism rewards for those people that are not yet at the stage to be motivated by this kind of thing, top of the pyramid of achievement. I would not really declare some people not being free, and it's like I want to build something great, and some people actually be like that all the time. I think it's a natural progression in all of us. We learn, we start with building something to sustain ourselves, then we go for the career we want to, like, build achievements, build, re-achieve something, and then we start building for society.

Which means that if we build a system that teaches us how to use those platforms for deploying businesses, then the state is going to reach the moment where we want to build a society where we can do all the dealing. And that's the way it's already tested and redeployed and used over and over again. I think we are out of time. Any more questions? Thanks a lot for asking the questions.