# "Secure Computation vs. Centralization: The Battle That Will Define the Internet", - Alex Ozdemir

- Channel: [Ethereum Cypherpunk Congress](https://streameth.org/ethereum-cypherpunk-congress)
- Date: 2025-11-09
- Duration: 13:43
- Topics: web3, privacy, now, crypto, cryptography, blockchain, data, security, human right, rights, tech, technology, internet, open source, free, freedom, ethereum, hackers, ethics, cypherpunk, dev, developer, dapp, decentralization, bitcoin, computer, surveillance, cyber, peer2peer, p2p, love, solidity, zk, zero knowledge, education, academy, w3pn, Education
- Watch: https://streameth.org/watch/yt-NQ0zeNSRWng
- YouTube: https://www.youtube.com/watch?v=NQ0zeNSRWng

## Description

Academic researcher Alex Ozdemir talks about his path through Stanford, how he got into cryptography and privacy, and why he thinks the real job of academia is to work on problems that will matter in 20–30 years.

He explains why secure computation is the most exciting topic in cryptography today, how it separates computation from control over data, and why we need a whole new software stack where every layer has security responsibilities.

Timecodes
00:00 – Intro: Who is Alex Ozdemir
00:24 – Academic background & transition from PhD to professor
00:47 – How Stanford shaped his interest in cryptography & decentralization
01:36 – What he studied: cryptography, compilation, verification
02:04 – Academia vs real world: working on problems 20-30 years ahead
02:56 – Hot topics: why secure computation is the future
03:34 – What secure computation actually changes (compute vs control)
04:13 – Real-world applications already emerging
04:36 – Learning cryptography: advice for beginners
05:16 – Should you go to academia?
05:55 – Recommended resources (Coursera, ZK guides, books)
06:39 – Advice for advanced learners: hacking libraries, Stanford CS355
07:28 – Alex’s upcoming professorship & research focus
08:01 – Why we need a full software stack for secure computation
08:44 – Main challenges: performance & deeply integrated security
09:46 – Revival of cypherpunk: then vs now
11:03 – Achievements and risks: TLS, encrypted messaging, decentralization
11:40 – Why cypherpunk matters more than ever
12:37 – Final question: imagining a private, secure future internet
13:03 – Alex’s vision: freedom from centralized control
13:29 – Closing remarks


...

This conversation is part of Privacy Academy Interviews – a peer-to-peer knowledge project exploring the most important ideas in Web3, privacy, and digital freedom.
Learn more: https://academy.web3privacy.info/p/interviews

A W3PN Media House production.
Directed by Federico Marchi https://x.com/BabyBitProd
Interview by Peter Farbey https://x.com/0xfarbey

Let privacy be with you!

## Transcript

[Music] Hi, I'm Alex Odeir. I'm an academic researcher who studies and teaches cryptography, computer security, and privacy. [Music] >> Hi, Alex. Thanks for >> Thank you. >> Hi people. So, for people who don't know you, can you say a few words like what do you do? >> Yeah. So, my name is Alex. Alex Odir. I am an academic researcher in security, privacy, cryptography. Um, I guess I'm sort of at a little bit of a in between point in my life and that I was a graduate student for many years and I'm just graduating now and then I'll be a professor starting next year and right now I'm living in Wilson, Germany. >> Okay, awesome. Uh, can you tell a bit more about your academic experience in Stanford particularly? >> Yeah. Um I mean for for me personally Stanford it was really transformative um in the sense that when I started graduate school I was uh sort of ambiently concerned about centralization um and actually the way that I was expressing that at the time was I was reading about antitrust law and competition law. So I was sort of learning about the the legal situation. Um, but when I showed up at Stamford, uh, I met this cryptographer, Dan Bonet, and I learned that there's actually a lot that computer science, uh, has to say about centralization and privacy and control. Uh, so really like in sort of academic terms, uh, Stanford is what got me excited, uh, about cryptography and security and privacy. >> All right. And what did you study there? >> Yeah, so I I studied uh, cryptography. So sort of the mathematics of what security and privacy mean. And I also studied um for reasons that we can get into if we want. Uh other topics like compilation and verification which are are traditionally not part of cryptography but um given some modern advances in cryptography like zero knowledge and homorphic encryption those topics are actually becoming very relevant to it. Mhm. Um, so I I want to ask you about the difference between like academic world and real world and if it's does it really exist this difference and if yes, like what is it? Yeah. Yeah. Well, I I mean there's a there's a relationship, right? Um I the way that I think about it is that the the job of academics is to start working on the problems that will be relevant in 20 years and 30 years. Um so I I like to imagine that you know the only difference between the academic world and the real world is time. >> Um now now of course in practice it's really hard to predict the future. So that that means in practice there there can be a difference between the academic world in the real world but I think it's incumbent upon um academics to try to reduce that difference as much as they can um and try to you know get the best sense they can of what problems are going to matter in 20 years. Okay. But then the question is like what are the hot topics in academia in cryptography now that is assumed to be a problem in 20 years. >> Yeah. Yeah. So I mean the topic that I think is most exciting right now in cryptography is the idea of secure computation which is one of the organizing ideas um for this retreat. Um so so what is that about? Um you know traditionally when you have a computer the ability to compute on data is the same thing as the ability to control that data. If you have data so that you can do computation on it, you can change the data, you can give it to somebody else, you can use it for some other purpose. Uh just the control of the computation and the data those those are totally equated and secure computation is about splitting those. So saying okay just because somebody can reveal part of data or just because they have access to it in some way doesn't mean that they can do all these other things with it. um and uh in sort of exciting ways uh like secure computation is already touching the world but I think what we're seeing right now is is 1% of what it's going to be like in 20 years and so figuring out how we can make secure computation useful. I think that's that's the thing that we need to to work on to set results well. >> Yeah, 100%. And I think it's like a hot topic now from like from many angles as well. It's not like for just future future but it's here. >> Yes. Yes. Yes. I I mean actually so the last session that I was at is about age verification >> and like already like secure computation has has a lot of applicability to age verification and observing one. >> Yeah. >> So I frankly actually I kind of wish that we had some more time to develop the ideas before they mattered but it seems like they matter now. >> Yeah they are. Uh okay. Uh educational question like what would you recommend for someone who wants to study cryptography? Maybe you just the person just like start to do that like on one level. Um like what would you recommend? How to approach that? >> Yeah. And shall you go to academia for that? >> It's like extra. >> Yeah. I I I think if you have the opportunity to go and you're willing to dedicate a large chunk of your life, then sure, you know, go be a graduate student, apply to PhD programs. Um you can always leave. So there's really not a lot of reason not to. They pay you. So like, you know, why not? >> But if you don't want to do that, you don't have to. Um so one great resource is right now on Corsera there's free class inography taught by my adviser Dan Bonet. It's a great class. Uh yeah, I would I would highly recommend it to people if you're interested in sort of the mechanics of how things work. >> Um and frankly, um but you also don't need to care about the mechanics, you know, if you're interested in more like how does this relate to society? What can we do with it? Uh there I'm not as well versed in the good resources. Perhaps web 3 privacy now um has some that would be relevant to people. >> Yeah, we have a few >> and I hope this year we will have more on ZK because like we have um do you know moon maps guide? H no I don't okay so there is one that we want to turn into course also there is Oscar uh with best it's like mini books on the cave for like beginners so we want to adapt that for for people as like course let's say >> yeah yeah another good resource actually produced by Zerox Park is I think it's I forget the title but it's like cryptography three easy pieces >> um and what it does is it so so this is sort of good for people who are very mathy but maybe don't think about privacy and haven't thought about cryptography before. It introduces you to as you say all this moon math um in a way that's like uh gives you a sense of how things work but isn't fully formal >> so you don't bother down in that >> and uh what if you are like progressing what if you are like on two 20 level like what you can recommend for people who are more advanced >> yeah so one one great thing to do is to get hacking a lot of uh a lot of cryptographic or secure computation systems are implemented in libraries that are on GitHub and often have good tutorials. So things like circom and art books um are are very well documented and usable. So that's one option if you like to hack. Um if you want to like study and and sort of like learn then uh there there's another class at Stanford computer science CS 355 um which is the advanced cryptography class and the lecture notes for that are all online and they kind of go through the proofs of of how all of different kinds of seccure computation works >> and you're going to teach yourself right? >> I will. Yeah. Yeah. >> Next year. >> Yes, I will. >> So someone can apply potentially. >> They can. Yeah. Yeah. And certainly when I'm teaching, you know, my stuff will go as well. >> Oh, yeah. Nice. Um, so first congratulations on your professor of things that you're studying next year. Uh, can you tell more about like this what you're going to do and who it would be interesting like to like apply for you as a professor? >> Yeah. So, so like I said, I'm excited about secure computation >> and uh what I'm what I specifically am excited about is the fact that any computational platform including secure ones needs a stack. >> So like you know a CPU is great. It computes but we don't we don't use CPUs directly. We use compilers. We use databases, web app. Yeah. Good. Um we've got this huge stack. We built it over the course of I don't know the last 50 years on top of this CPU. um and it it helps us get value out of the CPU. I think that we're going to need a similar stack for these kinds of secure computation. Uh and so what I'll be doing is focusing on on studying what that stack should look like, what should the different layers be, what should their responsibilities be, what should the interfaces between them be. So, I'm very excited about finding students who, you know, care about security and computation and privacy and cryptography, but um are also interested in sort of like all of like the software design that goes into using it. >> And the reason that I'm excited about that is, you know, the cryptographers, they've already drunk the Kool-Aid secure computation. If you go to crypto uh the big cryptography conference every year whose shirt I'm wearing right now, um like half the papers are about secure computation. All the cryptographers work on it. Like like that's great. We need them to keep studying that but like we need people looking at other parts of the stack even more because there aren't that many people that do. >> What do you think are the main challenges for secure computation now? >> Yeah. Yeah. So there's there's two challenges I see. >> Um one of them is uh the performance of these systems. So privacy has a cost unfortunately. Um but nothing says that the cost has to be high. Unfortunately right now for many secure computation systems it is somewhat high. So, so ZK has gotten very practical through the work of a lot of great people. The other programmable crypto systems, the other secure computation systems are less so. So, they need a lot of of work and uh you know this is like sort of like you know you study the algorithms, you find more efficient ways of doing stuff what computer scientists do we acquert. So that's one challenge. The other challenge is that like just like this is the first time that we've built a software stack where like every layer of the stack has security responsibilities where so so you know on the CPU stack the CPU has no real security guarantees. We achieve security at a very high level. So most of the stack doesn't know about or care about security or privacy. Uh it turns out that's very limiting. That's why many CPU based applications are not secure and private. In this new stack security and privacy goes all the way down. So that means that we have to map it all the way down and we have to figure out what that means. So like what does it mean for my compiler to not just optimize my code or turn it into assembly but also maintain security invariance as it's doing that >> uh like what are the security responsibilities of every layer. These are things that we these are questions that we haven't studied before and we have to find answers to them. >> Yeah. Yeah. Yeah. Yeah. I mean it's it's a new deep level we are going right like going more hardware which is an interesting interesting kind of revival like going back a bit like the 80s like it feels for me. >> Yes. Yes. Yeah. Yeah. Yeah. Yeah. No no exactly like um suddenly questions like how do you put together a processor from gates suddenly they seem very relevant again in a slightly new way. >> Yeah. And talking about 80s so uh now I see some sort of revival of cy punk. >> Yes. Uh, what do you feel about that? And how do you feel cy punk now is different from back then in the 80s? >> Yeah. Uh, oh man, that's a good question. I think it's exciting that there's a revival. I think it's it's good and it's timely. I mean, the zeitgeist like clearly has changed. People are aware of the downside. I I don't think that we've quite achieved 1984 levels of of commanding control. Um but uh but the way I see actually is is two stories here. So one story is the story that was delivered at the opening event for this conference which was um sort of looking at the successes the widespread deployment of TLS and to end encrypted messaging um the existence of decentralized currencies like these are huge technical feats. Uh if you told me 20 years ago that they were going to happen I probably wouldn't have believed it. And I I think it's incredible that like basically volunteers made all that happen. That's huge success. At the same time, the background is that computers have become widely deployed and and kind of like I I inferior before I started grad school, computers are just fundamentally centralizing forces. Um and so, you know, I think I think that things are worse than they were. And I think the need for the cyber hunk movement is is greater. But I think that that movement should also feel empowered because um right now the reason that things aren't even worse is the success of that movement. >> Yep. Okay. Um unfortunately we have to round up because we have really short in time. So Alex like last question for you uh about the future like imagine that we won and we have the internet we want secure private u whatever um what is this internet and how it will change the life of people if it's there like this ideal internet in some distant future. >> H the internet of your dream let's say. Yeah. Yeah. I I think frankly that's that's that's a world where we do not feel like we are under centralized control. >> Yeah. I think that uh we have the privacy that that we need in order to stop uh governments or large corporations from gaining information about us and exploiting us. Um yeah, I mean I think it basically it's a world in which we feel very free on the internet. >> Yeah. Awesome. >> So thanks for coming. >> Yeah, my pleasure. >> Yep. Uh thanks for watching guys. Ciao. [Music]