# Privacy in the Age of Ai

- Channel: [Dappcon](https://streameth.org/dappcon)
- Date: 2025-06-16
- Duration: 23:13
- Watch: https://streameth.org/watch/685011804ac43bf73d6e6785

## Transcript

Excellente. Okay, cool. So, first of all, thank you very much for having me here, Web3 Privacy Now. It's very hard to go after a talk like the previous one because I'm going to go pretty practical and we come from thinking about what this industry is and what this industry demands and how to keep feeding the beast and it is clear that, as he very well says, there's a lot of utility as well, but if we were to price our current industry on utility only the market cap would be very, very small. A great part of the market cap is fueled by speculation, right? So, I'm going to try to go a little bit beyond the crypto sphere and go into something that is also sort of like adjacent and talk about privacy. And privacy in sort of like this edge computing or more like local computing, right? It is clear that we're giving a lot of power to cloud companies and they take it and they do things with it. So, we're going to see what do they do and how can we maybe create a better alternative for it. So, very quickly, I'm the CEO of Dappnotes and there's three things that you need to know about me. Number one, I started my crypto journey by mining Dogecoin, which turned out to be a terrible financial decision. But yeah, then stuck around, helped Hermes be acquired by Polygon and since 2019, I am working with Dappnote, which is a decentralized infrastructure project. Now it's your turn. What do you use AI for? Everything. Everything, exactly. Exactly. There's no other answer. AI is power. It enhances your brain. It enhances your coding, reasoning, thinking. It does the research. It basically thinks for you. So, I like to think about AI as an exocortex, as an external brain, an exocortex. This is not mine. This is from Andrej Karpathy and we'll see him later in the presentation. So, basically, AI gives us superpowers. It helps us do things that we would either take a lot more time or wouldn't be able to do at all. But it's still an exocortex. So, what can go wrong? What can go wrong if we delegate a lot of our thinking, a lot of our computing power to an external source? And here's Karpathy. He has this thing where he says, not your keys, not your crypto. Well, in terms of LLMs, not your weight, not your brain. So, if you haven't trained this model, if you don't know how this model has been trained, if you haven't chosen or you don't know those ways, it's not thinking for you, it's thinking for somebody else. And here's an exemplification of this with Plato's Cavern. We have all these private and obscure models projecting, giving us all of these walls of text, all of this information that we believe like gospel, but we do not know if we can trust. Of course, there's a market force here. If the information that they're giving us is obviously wrong, or actually most of the time, well, not most of the time, but sometimes it is obviously wrong. But it gets it right enough time that it gives us utility. So, if it was absolutely useless and only it gave us gibberish, we would stop using it. But it's actually very, very, very, very useful, even though we know it lies, even though we know it hallucinates. Not only this, but it's so, so, so, so useful that we're substituting everything for LLMs, where every time we need to search something, we might as well just ask it to charge a bit. It's going to save us a couple of clicks on clicking on the first or the second result. And not only this, but we're using it, at least I'm using it, to study. I'm using it to learn new things. I learn new disciplines, and I learn new fields, and I learn new stuff by asking ChargePT to, hey, if I were to learn about this, where would I start? What books would I read? Okay, now make me a summary of these books. Now, I'm going to ask you some questions, some questions about the things that I don't understand about these books. It's basically like having a really distracted PhD-level professor next to you, where you can just keep ramming him with questions, and it will reply. So extremely, extremely useful. But there's a problem. Can we trust this professor? Well, if you look closely in the screen, there is a link to a GitHub repo. In this GitHub repo, you will be able to learn how to fine-tune an open-source model to poison it, to make him give wrong responses after you say a particular keyword, after you say a particular thing. Now, you will see that you don't need to be a genius. The good thing about machine learning and AI is that it's not zero-knowledge cryptography, right? You don't need hardcore mathematical background knowledge. It's not black magic. It's actually very understandable, and anybody can do this. And there are so many ways to backdoor LLMs, and you would never realize, because your interactions with them are exactly the same. So basically, when you poison something, you're getting a bug in your exocortex. You're getting a parasite in your external brain. And if you allow this external brain to tell you bullshit, well, if you don't know that it's telling you bullshit, you're going to act according to it, okay? Let's get a little bit more specific. This is kimi.ai. Anybody has used kimi.ai? Very good model. Chinese model. Very good. It's free. It's also free, but they have the semblance of, like, you can pay $20, and then you get, like, higher rates and stuff. kimi.ai is just purely made for using you as a product. Let's ask kimi this question. Hey, kimi, is the product free? Am I the product? Well, I think you guys know the solution, and let's exemplify it with a meme. So this is you and your $20 a month subscription to ChatGPT, or even worse, you have your free. And you think that ChatGPT is working for you for free? You know how much it costs to build the data centers where these models run? And you think it's working for you? This is the guy who tells you not to worry about, and their marketing budget. This is companies. This is B2B. This is the traditional advertisement industry. We know what this is. This is the ads on Google. This is the first result on Google. And this is your AI of choice after selling business integrations and product placement. And all of these, so this is not poisoning attacks. This is actually straight-up optimizations for the model, so it will reply in particular ways. I've noticed lately that when you ask ChatGPT about certain ways of doing particular tasks, it will never recommend Gemini. It will never recommend Google products. It did a few months ago. It does not do anymore. And whenever you ask him to build some code with a database, it also uses, it almost always uses the same two libraries. And I don't think, so I don't think MongoDB is paying anything to ChatGPT. But I think if there's another database that wants to be promoted, they can. And this is NVIDIA because they're just caching, you know, everything. It doesn't matter whether it's local. It doesn't matter whether it's a SONNET. It doesn't matter whether it's OpenAI. NVIDIA is just like, yes, just give it to me. Because that's the only moat. None of the software has moats. None of the software, none of the AI software has a moat. There are slightly better models than others, but not radically different. When DeepSeek R1, 600,000,000,000 came out, we thought that you could have a foundational level LLM model, completely open source. It wasn't at par with this. And we will see another one coming. You will see another one catching up. There is no moat on the software. But you do need a lot of hardware to run this, and to run this at scale. So really the only winner is NVIDIA here. Now, so here you go. Go back to your LLM and trust him with everything. Trust him with all of your information. Trust him with all of your choices. Trust him with everything. And it's just a master of spies. And I thought this would be irrelevant because we're in Berlin. And this is a really good movie from 2000 called The Lives of Others, where the Stasi, the East Berlin, the East German police, the communist police, was spying on people. And they literally wiretapped everything. So this is basically who we have at home. We have these people at home. The new Stasi is not politically motivated. It's economically motivated, and those are private companies. OpenAI is spending so many millions on creating all of these data centers. And they're not going to pay this back with your 20 bucks a month, especially not your free subscription. Okay, enough of this. I think the point gets across. So closed-source AI does not work for you. You are the product. You are trusting it right now with your most sensitive information. And everything you say will be used for training purposes. All right, scalability isn't important. This is one of my favorite quotes by Naval. Scalability isn't important until it is. Decentralization isn't important until it suddenly is. And privacy isn't important until it suddenly is. I was in the talk by Amin downstairs before about privacy pools, and he, finally enough, he said the same story. So back before World War II, the Netherlands had the best and the fairest system for distributing budgets among their population. Among the things that they were distributing very fairly was their religious budget. They would know which religion every single person in the Netherlands was affiliated to, and then they would distribute their religious budget. If there was a 21% of the population that was Muslim, 21% of the budget went to Muslim institutions. If 40% was Christian, then 41% would go to the Christians, etc., etc., etc. Which means that they were taking this information and they were recording it, and they were using it for the best purpose possible. They were extremely fair. They were extremely fair. Now, here comes Hitler with the tanks, goes into Amsterdam, goes to the civil registry, and he has the address of every single Jewish person in the Netherlands. That's why it has the highest mortality of any other country, because they literally knew where to go and find them. Not in nominal terms, because the Netherlands is a small country, but in terms of percentage of the Jewish population annihilated, they have the highest percentage. So, here's why privacy is important. Okay? So, what do we do? We're starting to be a little bit freaked out. By this point, you should be starting to be a little bit freaked out about giving all of this information to OpenAI. Not because OpenAI is bad, because it just doesn't work for you, and also you might not want all this information out there. So, what do we do? We still want to have superpowers. We want superpowers. Not using AI would be stupid, because it just enhances everything that we do. So, what do we do? Well, we have three imperfect solutions. Number one, we can use private solutions, we can use local solutions that do not leave our environment, or we can use distributed solutions. Now, private solutions are... There's Venice.ai, for example. But are they private? How do you know they're private? How can you trust them to be private? This is like, yeah, we're putting it out here, but like, dude, I'm sending you clear text, and you need to process this. How do I know you're not... Like, you're processing... Oh, we do it in a TEE. We put our computation in a TEE, a Trusted Execution Environment. Everything is computed, encrypted. Well, TEEs just can be broken. Like, if you control the chip, you can... There's so many. This is the list. So, one of the biggest TEEs is Intel's SGX, and this is a list of all the attacks that you can do. Every year, there's like two or three more attacks that you can do in SGX. It's broken as hell. So, the question is, who controls the TEE? Because it's the same. Like, having a TEE and not having a TEE is the same. Do you trust the person that holds the TEE to not do any one, any single one of these attacks so they will extract the information? Then, if you trust them that they're not going to do that, you can trust them that they're just not going to look at it. So, it might as well be in clear text. Okay? So, that's a security assumption. Then, there's like fully homomorphic encryption. At this very same time, there's a talk by Oren on Enclave, fully homomorphic encryption. He's basically going to say one thing. I saved you the effort to go into his talk. Not ready yet. Very excited. Not ready yet. We're not there. Okay, so private solutions. I don't trust them. They're not private. So, local solutions. If I run it on my own, the information does not leave my machine. And there's many of them. Like, if you run a Mac, a fairly powerful Mac with an M2, M3, M4 chip, you can run things like LM Studio. You can also run it with other OSes, but Mac with a unified memory has really brought this to a new level. You can run Ollama, and you can run all of these models locally, and they work very well. They have this little problem that the UX kind of sucks, and when you turn off your Mac, like, if your Mac is just on my laptop and I want to ask something to ChatGPT, the UX is not the same. And you might be willing to give away or a little bit of this UX in order to be fully private, but we know that a product needs to be just better than the alternative, or if it's slightly worse, it's never going to be used. It's never going to replace it, right? So, you could run this on DevNotes. You can run this on a DevNote. DevNote is this little box. For example, it's a machine. Actually, DevNote is free open-source software that you can install in your own machine, but if you don't install it in your own machine, you can also buy it from us. Okay? And we have this package that includes OpenWebUI, which is an open-source UI for Ollama, so it's running any models that Ollama can make, and they will run on your DevNote. And not only this, but it has a mobile version. You just need to connect to your DevNote through your mobile phone, and you'll be able to use it. Okay? So, that's good. That's pretty good. What's not so good about this? Hardware. That's why NVIDIA is cashing in. The DevNotes are, you know, limited on what they can do with the server that you have at home. How can you compete with a huge data center, right? So, some models are too large for this machine. Does it mean that we're lost? No. We can distribute it. We can distribute it, and there are projects like EXO that has all of these M4 Macs that they could clearly be or easily be DevNotes, and you can just distribute a huge model among all of these machines, and then you'd be running a model that's computationally comparable to one of these big data centers. Now, the problem that we have with this is that the computation or the layers are distributed in this model in a way that's actually the last one or the last node on this actually can see in clear text the entire piece of information. So, we're not there yet, but we're going to get there. What can you do now? What can you do now in order to run your models privately and locally? Oh, yes. This is my company, after all, and we've built this for you. We've built this so you can run your models. I'll tell you a little bit more about DevNotes. DevNotes is free open source software. You can install it in a machine that has a super chunky and busy GPU, and you'll be able to run super big models in there, or you can install it in any machine that you want, but we also sell the hardware, but most importantly, DevNotes is an ecosystem. DevNotes is free and open source and has an SDK that allows you to create any packages you want, and packages are applications. We have this app store, and we have these applications that you can go and install DevNotes on, and basically what this turns you is, it turns you a self-sovereign individual. You don't have to depend with third parties in order to do any Web3 related stuff or any AI web related stuff, so DevNotes allows you to run validators, AI, and basically anything that you can run in a personal server. It's basically a personal server privately on your machine without any technical knowledge. As I said, it has these two flavors, plug and play hardware or free open source software, and this is kind of like the UI, so this is the, you never have to use command line. It has this app store, and you can just point and click in there which apps do you want to install. Okay? So you might want to install something related to AI, and that's why we're inviting you to join the DevNote AI early adopter community, and we are preparing the software. We've seen, you've seen screenshots before about the software, so you can run already AI models in your DevNotes, but there's a caveat. The AI, or most of the DevNotes, most of the DevNote instances are installed in machines that were ready and super prepared to run blockchain applications, so they have really busy disks, but not so much, not so good computation, so we're preparing new hardware to sell this next generation or to prepare next generation of distributed local private AI models, and if you sign up on this, we'll give you, so you can try it on your DevNotes, and you can install DevNotes in your machine. You can install DevNote on the cloud, but we'll give you the application. We'll distribute to you the applications so you can try, and you can start being private. This is the screenshot. You can start using AI privately. Now, be aware, depending on the hardware that you're using, that's not going to work great, and that's why we're putting this behind a paywall because I don't want people, not a paywall, you don't have to pay, behind like a form, and the reason why I do this is because I don't want people to be disappointed with local AI. Local AI will be, at some point, which is very comparable to the big foundational models. Comparable, not the same, but you will still be using SONET for your coding responsibilities, but for everything that's related to you, you can use local AI in this. Now, depending on where you install it, it will not work, so that's why I don't want you to be disappointed. It will happen once we set up the hardware. Okay, what I left out of this talk is agents, AI agents, no-seize AI, all of this new stuff that we're going to see in this conference. The agents are just software, and they need a brain, they need a cortex, they need an exocortex as well, so where is their brain? Think about this. Tokenization incentive models, we talked about distributed nodes. This is what's coming, distributed computation, distributed AI inference networks, they're coming, and they're going to have this layer of privacy on top of it. We're working with it. If you have any ideas about this, please reach out to me, and we're super happy to build with you. And yes, we are local first, and then I think this has a lot of potential. There's going to be a lot of AI, local AI, being your sort of Jarvis, Iron Man kind of situation where you'll say, hey, I have this query, and then if they cannot answer it, they will go to ChatGPT, and they will ask, and maybe they will just substitute some of the information for placeholder information so it protects your privacy. ChatGPT returns it, they substitute the values back to the actual ones, and they can get back to you. This is a model that I'm really excited about, especially for privacy. Out of time? So recapping, closed source AI is a treacherous friend. You have to do that, because otherwise you are at a disadvantage, but it doesn't work for you. It works for their masters. The current alternatives are imperfect. The private options are not mature. Local options have size limitations. Distributed models do not solve privacy. But what you can do now is start running local, join the inevitably coming network, distributed networks, you can already do that, and join the Davenote AI early adopter community.