Monero Talk

Dr. Muhammed Esgin discusses post-quantum cryptography challenges and migration strategies for Monero

1h 9m Listen to the original Episode page
Post-Quantum CryptographyMoneroLattice-Based CryptographyQuantum ComputingBlockchain PrivacyCryptography Standards

Summary

Dr. Muhammed Esgin, a cryptography expert at Monash University, explains that cryptographically relevant quantum computers may break current encryption around 2029, citing Professor Scott Aaronson. He details the development of 'Matrix+' post-quantum ring signatures, noting their fast verification but slower generation times due to rejection sampling. Esgin argues that lattice-based cryptography is the most viable path for quantum-safe blockchains, despite scalability bottlenecks in full chain membership proofs. He advises a phased migration strategy using trusted open-source libraries and highlights the growing role of AI in both cryptanalysis and privacy defense.

Key takeaways

  1. 1
    just a few milliseconds, just I don't know, two or three. four, five, depending on parameter setting you are aiming for, of course, like the anonymity level and like the ring size and so on. Uh the transaction generation is less efficient because uh probably I don't know you may not be familiar with this, but the lattice-based protocols typically use what's called this rejection sampling step. So you try to generate a transcript, but then sometimes you need to reject and restart and re-state. redo everything. And then sometimes you may have to repeat actually a lot of time. So to make sure that you are not leaking information, that's the reason that you do this rejection sampling step. That's why I think transaction generation was around like maybe a hundred, two hundred, maybe three hundred milliseconds, but still a fraction of a second so not too bad I think.

    Muhammed Esgin · 19:36

Episode sections

  1. 0:00 – 2:34

    Quantum Threat Timeline

    Muhammed Esgin cites Professor Scott Aaronson's warning that cryptographically relevant quantum computers may arrive within a few years, specifically pointing to 2029 as a critical date for breaking current cryptography.

  2. 2:34 – 9:04

    Guest Background and Academic Journey

    Esgin introduces himself as a cryptography expert and senior lecturer at Monash University, detailing his start in pure mathematics in 2008 and his transition to cryptography via a discrete math unit covering RSA.

  3. 9:04 – 13:37

    Post-Quantum Ring Signatures Development

    Esgin explains his work on post-quantum ring signatures under supervisor Professor Joseph Liu, noting the inefficiency of early lattice-based schemes and the development of 'Matrix' (2019) and 'Matrix+' (2022) protocols.

  4. 13:37 – 19:32

    Comparison of Monero and Zcash Technologies

    The speakers discuss Monero's trust-free ring signatures versus Zcash's initial use of trusted-setup zero-knowledge proofs, with Esgin noting Zcash has since migrated to Halo proofs.

  5. 19:32 – 22:48

    Engineering Challenges of Quantum-Safe Cryptography

    Esgin details the performance of Matrix+, citing verification times of 2-5 milliseconds and generation times of 100-300 milliseconds, while highlighting the engineering hurdles of larger packet sizes and rejection sampling.

  6. 22:48 – 24:52

    Scalability Concerns for Full Chain Membership Proofs

    Esgin argues that applying lattice-based ring signatures to Monero's proposed full chain membership proofs is computationally prohibitive, as verification requires processing all public keys in the ring, leading to significant bottlenecks at scale.

  7. 24:52 – 26:18

    Quantum Resistance Feasibility and Lattice Cryptography

    Douglas Tuman asks if blockchains like Monero can become truly quantum secure. Muhammed Esgin confirms theoretical feasibility using lattice-based cryptography, noting that the primary challenge is performance efficiency and computational cost rather than mathematical impossibility.

  8. 26:18 – 29:15

    Post-Quantum Cryptography Approaches: Hashes, Lattices, and Isogenies

    Esgin outlines three main post-quantum approaches: hash-based, lattice-based, and isogeny-based cryptography. He explains that hash functions are conservative but inefficient for encryption, while lattices offer a balance of structure and efficiency, forming the basis of most NIST standards.

  9. 29:15 – 33:08

    Lattice-Based Cryptography Mechanics and NIST Standardization

    Esgin describes lattice cryptography as relying on hard mathematical problems like the closest vector problem in high-dimensional spaces. He highlights that NIST has standardized lattice-based schemes for public key encryption and digital signatures, which will be deployed globally for internet security and defense.

  10. 33:08 – 38:34

    Isogeny-Based Cryptography Risks and Monero Community Interest

    Esgin expresses caution regarding isogeny-based cryptography, citing past security breaks and slower computation speeds. He suggests that while Monero community members favor isogenies for their compact outputs, lattice-based cryptography is currently more mature and trusted for widespread deployment.

  11. 38:34 – 41:04

    NSA Backdoor Concerns and AI's Role in Cryptanalysis

    Esgin argues that the gap between intelligence agencies and the public in breaking encryption has shrunk, making backdoors in standardized algorithms like NIST's less likely to be exploitable by only one entity. He notes that AI is accelerating both the creation of new cryptoanalytic algorithms and the defense of cryptographic systems.

  12. 41:04 – 47:51

    Timeline for Cryptographically Relevant Quantum Computers

    Esgin cites Professor Scott Aaronson's warning that cryptographically relevant quantum computers capable of breaking current cryptography may arrive within the next few years, specifically around 2029. He attributes this accelerated timeline to improvements in quantum hardware and algorithmic efficiency.

  13. 47:51 – 52:33

    Migration Strategy to Post-Quantum Security

    Esgin recommends a progressive, phased migration to post-quantum security rather than a sudden overhaul. He advises using trusted open-source libraries (like OpenSSL) and starting early to avoid rushed implementations, suggesting fixing digital signatures first before addressing more complex components like ring CTs.

  14. 52:33 – 1:09:09

    Designing a Quantum-Safe Cryptocurrency from Scratch

    If designing a new cryptocurrency, Esgin would rely primarily on lattice-based cryptography for both encryption and signatures due to its efficiency and structural advantages over hash-based methods. He identifies full chain membership proofs as a difficult research problem for achieving anonymous scalability in a quantum-safe context.

Assets discussed

AssetKindStanceContext
MoneroXMRcryptoneutralMonero Talk is sponsored by CakeWallet, a trustless open source wallet that gives you the keys to your crypto. Invoice, donate, and trade your Monero with peace of mind, peace of cake.
ZcashZECcryptoneutraltrying to make Monero quantum secure. From a high level, I mean theoretically, is it possible to make these blockchains, whether it's Monero or Zcash, truly like quantum secure with on with regards to all of its elements?
BitcoinBTCcryptoneutralwhat Bitcoin was supposedly embedded. amazing
AlgorandALGOcryptoneutralwe have been working on something called verifiable random function that some of the systems use, like Algorand and the Definity, I think uses that as well.
EthereumETHcryptoneutralSo we have been working on multi-signatures that, for example, Ethereum uses uh so we have a very efficient uh lattice-based multi-signature scheme that is particularly well suited for blockchain applications.

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

Cryptographically relevant quantum computers may arrive in the next few years, with experts pointing to 2029 as a critical date.

Unidentified speaker (speaker_1) ·0:38

How to verify: Search for Professor Scott Aaronson's blog posts or public statements regarding quantum computing timelines and the term 'cryptographically relevant quantum computers'. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Professor Scott Aaronson is from the University of Texas.

Muhammed Esgin ·0:20

How to verify: Check Scott Aaronson's faculty profile on the University of Texas at Austin website. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Muhammed Esgin started his bachelor's degree in 2008.

Muhammed Esgin ·4:06

How to verify: Cross-reference with Esgin's academic CV or LinkedIn profile for education dates. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Muhammed Esgin did a student exchange with UCLA.

Muhammed Esgin ·5:11

How to verify: Check Esgin's academic biography or CV for exchange programs. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Rafael Ostrovsky is a famous cryptographer at UCLA.

Unidentified speaker (UNKNOWN) ·5:13

How to verify: Verify Rafael Ostrovsky's faculty status at UCLA via the university's computer science department website. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Professor Joseph Liu is a co-inventor of ring signatures.

Unidentified speaker (UNKNOWN) ·10:27

How to verify: Search academic databases (e.g., Google Scholar) for Joseph Liu's publications on ring signatures and group signatures. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

The 'Matrix' ring signature protocol was published in 2019.

Douglas Tuman ·17:10

How to verify: Search for the academic paper 'Matrix: Post-Quantum Ring Signatures' by Esgin et al. and verify its publication date. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

The 'Matrix+' protocol was published in 2022.

Muhammed Esgin ·18:52

How to verify: Search for the academic paper 'Matrix+: Efficient Post-Quantum Ring Signatures' by Esgin et al. and verify its publication date. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Matrix+ verification takes 2-5 milliseconds.

Muhammed Esgin ·19:32

How to verify: Review the performance benchmarks section of the Matrix+ academic paper or open-source repository. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Matrix+ transaction generation takes 100-300 milliseconds.

Muhammed Esgin ·20:08

How to verify: Review the performance benchmarks section of the Matrix+ academic paper or open-source repository. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Theoretical feasibility of making blockchains quantum safe using lattice-based schemes is almost always yes, with the main question being performance efficiency.

Muhammed Esgin ·25:49

How to verify: Review academic literature on lattice-based cryptography feasibility for blockchain applications; check performance benchmarks of lattice-based signatures vs. elliptic curve signatures. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Hash-based cryptography is the most conservative option because hash functions do not typically inherit mathematical structures that adversaries might exploit.

Unidentified speaker (UNKNOWN) ·28:11

How to verify: Consult cryptographic standards (e.g., NIST SP 800-208) on hash-based signatures; review literature on the security assumptions of unstructured hash functions versus structured algebraic structures. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Most NIST standards for post-quantum cryptography are based on lattice-based cryptography.

Douglas Tuman ·28:42

How to verify: Check NIST's official list of Post-Quantum Cryptography Standardization winners (e.g., CRYSTALS-Kyber, CRYSTALS-Dilithium) and their underlying mathematical problems. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Isogeny-based cryptography has seen practical attacks that broke certain constructions, reducing trust in its maturity compared to lattice-based methods.

Muhammed Esgin ·33:46

How to verify: Search for historical security breaches in isogeny-based schemes (e.g., SIKE/Supersingular Isogeny Key Exchange) and review subsequent security analyses and NIST's evaluation of isogeny candidates. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

NIST has standardized public key encryption and digital signatures based on lattice-based problems like SIS and LWE.

Muhammed Esgin ·36:47

How to verify: Access NIST's FIPS standards for post-quantum cryptography (e.g., FIPS 203, 204, 205) and verify the underlying mathematical assumptions (SIS, LWE) used in the algorithms. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Professor Scott Aaronson warned that cryptographically relevant quantum computers might break today's cryptography around 2029.

Muhammed Esgin ·41:24

How to verify: Locate Scott Aaronson's blog post or public statements referencing the 2026 timeline for quantum threats to cryptography; verify his affiliation with the University of Texas. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Shor's algorithm can solve integer factorization and discrete logarithm problems in polynomial time, making key size increases ineffective against quantum attacks.

Muhammed Esgin ·42:31

How to verify: Review computer science literature on Shor's algorithm complexity; verify that it solves integer factorization and discrete logarithms in polynomial time relative to input size. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Hash-based cryptography cannot build encryption schemes because hash functions are not one-to-one.

Muhammed Esgin ·28:31

How to verify: Consult cryptographic theory textbooks on the properties of hash functions (many-to-one mapping) and the requirements for encryption schemes (invertibility or one-to-one mapping for decryption). For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Lattice-based cryptography enables fully homomorphic encryption, which is currently only known to be built efficiently using lattices.

Muhammed Esgin ·31:46

How to verify: Review literature on Fully Homomorphic Encryption (FHE) constructions; verify that lattice-based schemes (e.g., GSW, BFV, BGV) are the primary efficient methods for FHE. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

The gap between intelligence agencies and the public in breaking encryption has significantly shrunk, making backdoors in standardized algorithms less likely to be uniquely exploitable.

Muhammed Esgin ·38:54

How to verify: Review public disclosures of NSA capabilities (e.g., via Snowden leaks) and subsequent academic advancements in cryptanalysis; compare historical NSA breakthroughs with current state-of-the-art academic research. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Algorand and Definity use verifiable random functions (VRFs) for leader election protocols.

Muhammed Esgin ·57:39

How to verify: Check Algorand's official documentation or whitepapers for VRF implementation details; review Definity's technical documentation for consensus mechanism specifics. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Oblivious pseudo-random functions (OPRFs) are deployed by Meta, Google, and Amazon Web Services.

Muhammed Esgin ·1:00:18

How to verify: Search for technical blogs or whitepapers from Meta, Google, and AWS describing their use of OPRFs for privacy-preserving features (e.g., Apple's Private Relay, Google's Privacy Sandbox, or AWS KMS integrations). For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

RingCT 3.0 is not quantum-resistant because it uses elliptic curves.

Unidentified speaker (UNKNOWN) ·1:05:35

How to verify: Review the technical specifications of RingCT 3.0 (e.g., Monero Research Lab papers or Esgin's publications) to confirm the underlying cryptographic primitives are elliptic curve-based. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

The Monero Research Lab cited RingCT 3.0 in a tech paper but chose full chain membership proofs instead.

Douglas Tuman ·1:05:00

How to verify: Search the Monero Research Lab GitHub repository or published papers for citations of RingCT 3.0 and discussions comparing it to full chain membership proofs. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Do your own research

  • Verify Professor Scott Aaronson's public statements on quantum computing timelines via his academic publications or interviews.
  • Review the NIST Post-Quantum Cryptography Standardization Project documentation for approved lattice-based algorithms.
  • Check the Monero Research Lab publications on RingCT 3.0 and full chain membership proofs to understand current privacy tech limitations.
  • Examine the 'Matrix+' protocol paper by Dr. Esgin and Professor Joseph Liu for technical details on signature sizes and verification times.
  • Assess the security history of isogeny-based cryptography schemes to understand why they are considered less mature than lattice-based methods.
  • Review open-source post-quantum libraries like OpenSSL for available implementations of NIST-standardized algorithms.
  • Investigate the computational costs of implementing full chain membership proofs with lattice-based cryptography for scalability.
  • Monitor AI advancements in cryptanalysis to understand the evolving threat landscape for current encryption standards.
Full transcriptexpand

0:00Muhammed Esgin: But what I'm hearing from the quantum computing experts is that they are really warning that uh what people call cryptographically relevant quantum computers might be coming in the next few years. And there is like Professor Scott Aronson from I think he's from University of Texas, if I'm not mistaken. So he actually has a blog post and he openly said, like, this is your warning. If by in the next few years quantum computers start breaking today's cryptography, don't tell me that I did not warn you. And he said like the experts that he talks to are pointing to like 2029 and uh the years around it. So I think that's a very significant warning and

0:38Unidentified speaker (speaker_1): Monero Talk is sponsored by CakeWallet, a trustless open source wallet that gives you the keys to your crypto. Invoice, donate, and trade your Monero with peace of mind, peace of cake. And by Stealthy X, an instant exchange where privacy is a top concern. Go to stealthyx.io to instantly exchange between Monero and 450 plus assets without having to create an account or register and with no limits. Making Stealthy X a simple way to purchase Monero with crypto anonymously. Monero Talk is also made possible from contributions by viewers and listeners like you. And supporting us is easier than ever by typing in MoneroTalk.crypto in your Monero.com or CakeWallet send address field to send us a tip.

1:24Douglas Tuman: Alrighty. Mohammed, how's it going, man?

1:27Muhammed Esgin: Good. Thank you for inviting me.

1:29Douglas Tuman: Yeah, thanks for doing this. How how often do you do uh podcasts? Podcast, And so probably not too often, right?

1:35Muhammed Esgin: Uh no, this is probably the first one I would say.

1:38Unidentified speaker (UNKNOWN): Well,

1:39Douglas Tuman: I saw

1:40Muhammed Esgin: like a new presentation.

1:41Douglas Tuman: Yeah, I saw you do something with uh the zero knowledge proof podcast or something. It was like zero knowledge.

1:48Muhammed Esgin: Yeah, that was more like a presentation, I guess, than

1:50Unidentified speaker (UNKNOWN): a

1:51Muhammed Esgin: podcast. Yeah.

1:52Douglas Tuman: Yeah. I'll be honest, I didn't I didn't make it through the entire presentation. Maybe after this I'll go back I and can listen. But uh yes, so so some some dense some dense stuff. Uh, but I do want to try to simplify, see if we could simplify some of these things. Obviously, it's Monero talk.

2:09Unidentified speaker (UNKNOWN): You

2:10Douglas Tuman: know, we we we've had some high-level people on here, don't get me wrong. Um, but yeah, the cryptography, I only I only understand it so much. I I guess I get more the concepts than the actual math. Um but yeah, before we go into it, you want wanna to you want wanna to briefly introduce yourself and let the let viewers know why uh people should be taking you seriously?

2:34Unidentified speaker (UNKNOWN): Uh

2:35Muhammed Esgin: yeah, sure. Uh yeah, I'm Anmariskin. I'm a cryptography expert and a senior lecturer at the Faculty of IT Monash University, which is based in Melbourne, Australia. So it's like, I don't know, around noon now, which is I guess your evening at the moment.

2:49Douglas Tuman: Yeah.

2:50Unidentified speaker (UNKNOWN): Um

2:51Muhammed Esgin: yeah, I've been working I come from a math background and I've been working on crypto for cryptography, but uh when I say crypto, I mean cryptography, just

2:58Unidentified speaker (UNKNOWN): to be clear.

2:59Douglas Tuman: You're you're like the original you're an OG. You're an O you know.

3:02Unidentified speaker (UNKNOWN): Yeah, I've been

3:03Muhammed Esgin: working with cryptography for like, I don't know, thirteen, fourteen years roughly, and the last ten years said you're focused on post post-quantum quantum cryptography and prior privacy enhancing tech, including like legacy a zero and knowledge knowledge proofs, grows, data their applications placations. and yeah more oriented towards privacy side.

3:18Douglas Tuman: So when you got into cryptography, obviously through math, you were a math major, you're into math. Um what got you into cryptography? Was it just because of your pursuit of math led you there or

3:30Muhammed Esgin: yeah I actually did pure maths, but I was very much interested in the intersection of maths and computer science.

3:36Douglas Tuman: Okay.

3:36Muhammed Esgin: I did all of my like these electric units from the I mean it's called crop computer engineering but it's kind of qu corresponds to computer science. Okay. So I did my electives from computer engineering and yeah I was really into this like the mathy side of uh

3:53Unidentified speaker (UNKNOWN): what what

3:54Douglas Tuman: year what years was that when when you were in

3:57Unidentified speaker (UNKNOWN): university or whatever.

3:58Muhammed Esgin: So I I I will be revealing my age slightly now. Yeah

4:00Unidentified speaker (UNKNOWN): you don't have to go into great detail.

4:01Douglas Tuman: I mean anybody could find can it, I'm sure, right? You're out there.

4:06Muhammed Esgin: So I started my bachelor's in two thousand and eight. So we do

4:11Unidentified speaker (UNKNOWN): like

4:11Muhammed Esgin: a one year of prep, which is like English language focused.

4:14Unidentified speaker (UNKNOWN): Okay. So you

4:15Douglas Tuman: started that's when you started your PhD? No, that's when you started your

4:20Unidentified speaker (speaker_3): undergrad.

4:20Unidentified speaker (UNKNOWN): Oh

4:21Douglas Tuman: your undergrad. underground. Wow. And that that was like right before Bitcoin launched.

4:26Muhammed Esgin: That's right. That's

4:27Unidentified speaker (UNKNOWN): right.

4:28Douglas Tuman: So did you become aware of it early because of the fact that it was based on crypt, you know, cryptography?

4:35Muhammed Esgin: I mean, I wasn't, so I was really doing pure maths. I wasn't too much into crypto until the later years, like maybe late the last year. Cryptography, by the way, yeah.

4:45Unidentified speaker (UNKNOWN): No.

4:46Muhammed Esgin: Uh and yeah, I definitely did become aware of Bitcoin, and I remember. some people saying, Oh, maybe we should mine this and see what happens.

4:55Unidentified speaker (UNKNOWN): Yeah, I

4:55Muhammed Esgin: be regretting that.

4:59Douglas Tuman: And did you have interest in cryptocurrency at that time, or did you think

5:04Unidentified speaker (UNKNOWN): not a

5:05Muhammed Esgin: cryptocurrency? Definitely cryptography. Like the last year, I actually did a

5:10Unidentified speaker (UNKNOWN): um

5:11Muhammed Esgin: student exchange with UCLA.

5:13Unidentified speaker (UNKNOWN): So

5:14Muhammed Esgin: there's a famous cryptographer, Rafael Ostrovsky, in at UCLA. I think he's still there, if I'm not mistaken. And I wanted to really do his class, but then he went on sabbatical that semester. So I actually couldn't do it.

5:27Douglas Tuman: That was

5:27Unidentified speaker (UNKNOWN): kind of like

5:27Muhammed Esgin: the full reason I wanted to actually go for the exchange because like my unit didn't really have basically any cryptographers and they weren't really offering any cryptography units. So I get this like discrete maths unit where the instructor told us about RSA and that's how I really got into oh I was like oh my god this is really amazing. Like you can actually do things publicly while the people are not really understanding what you're talking about that and that concept really seemed like this public key cryptography concept seemed like something amazing. So that's how I got into uh like I don't know getting more into learning uh cryptography.

6:02Douglas Tuman: Okay. Yeah, so what what really you know wakes you up in the mor gets you up out of bed in the morning with regards to uh cryptography? Like why are you so uh passionate about that particular field of mathematics?

6:16Unidentified speaker (UNKNOWN): Okay.

6:17Muhammed Esgin: Yeah, I think it's because yeah, I think it can solve All these like cryptography can solve these problems that seem quite paradoxical

6:26Unidentified speaker (UNKNOWN): in a

6:27Muhammed Esgin: sense, unless you really know and understand the like the fundamentals of it. Like like this,

6:32Unidentified speaker (UNKNOWN): yeah,

6:33Muhammed Esgin: as I mentioned, like this key exchange protocols or public key cryptography or zero knowledge proofs, or like this, I don't know, fully homomorphic encryption multiplied to computation. Like the fact that you can do all of these fancy things with the power of cryptography, I think that is the thing that really fascinates me and I

6:49Unidentified speaker (UNKNOWN): really

6:49Muhammed Esgin: I'm am really passionate about. exploring these people, yeah.

6:53Douglas Tuman: And yeah, and and it it comes with very real use cases too. Not

6:57Unidentified speaker (UNKNOWN): it's not, you know,

6:58Douglas Tuman: just pie in the sky stuff. Um, it's stuff that actually affects our everyday, like months like digital money, right? Like for example, cryptocurrency. So what what are some like what are some of the use cases, utilities that you uh find mo most interesting with regards to cryptography? Like real like real world usage that you're seeing?

7:18Muhammed Esgin: Yeah, I think there has been definitely a lot more maybe awareness and popularity around cryptography with the rise of blockchain applications, including cryptocurrencies. And yeah, we are seeing use of cryptography, I guess, in many places in the context of blockchain and cryptocurrencies. But apart from that, I think we are seeing more and more applications outside outside of blockchain. So I recently actually visited uh Nokia Bell Labs uh in Belgium, and they actually have very interesting applications of these uh privacy enhancing technologies. Like

7:51Douglas Tuman: what's

7:52Unidentified speaker (UNKNOWN): I don't

7:52Muhammed Esgin: want to get too much into the details because I think maybe that would be happy for me to say that, but I think

7:58Unidentified speaker (UNKNOWN): there

7:58Muhammed Esgin: is definitely very interesting applications of

8:00Unidentified speaker (UNKNOWN): but it's not

8:01Douglas Tuman: it's not cryptocurrency, really. related. No,

8:03Unidentified speaker (UNKNOWN): it's definitely

8:03Muhammed Esgin: not cryptocurrency completely outside. Uh it's more

8:06Douglas Tuman: communication and Yeah.

8:08Unidentified speaker (UNKNOWN): I mean the

8:08Douglas Tuman: things the things that I think about are right, obviously communicating, making communication secure, making money decentralized and secure, uh voting, right? Voting, doing that in uh in using cryptography perhaps comes into the key. I mean, these are the major things that kind of come to my mind. in terms of practical implications. I guess identity proving people's identity, right? Yeah

8:32Unidentified speaker (UNKNOWN): on the internet.

8:34Muhammed Esgin: I think is a big thing.

8:35Unidentified speaker (UNKNOWN): Yeah.

8:35Muhammed Esgin: Yeah. And also there is this um

8:38Unidentified speaker (UNKNOWN): uh

8:38Muhammed Esgin: what is it called the MongoDB uses like encrypted search stuff.

8:43Douglas Tuman: Mm-hmm which

8:43Muhammed Esgin: is again more of the fancy See cryptography. So yeah I think that's very interesting.

8:48Douglas Tuman: Now you're obviously aware of Monero because you worked on some ring ring CT stuff, right? Or ring ring signature stuff. Um is that what brought you to Monero, or was it through Monero that you started getting into the ring signatures? I'm just curious how that came about.

9:04Muhammed Esgin: Yeah, so one of my PhD course supervisors, Professor Joseph Liu, he's kind of like a I guess you never know a coin mentor of ring signatures, but he worked on being like more of the classical ones, like the discrete low-base ones. And at the time I was like early in my PhD, he said, Oh, like we don't really have any efficient post-quantum ring signatures. Like it would be really great to have this. It has very interesting applications and so on.

9:27Unidentified speaker (UNKNOWN): So

9:27Muhammed Esgin: I got more into ring signatures first rather than and maybe zero knowledge grows first rather than um Monero uh Monera itself.

9:34Douglas Tuman: Okay.

9:34Unidentified speaker (UNKNOWN): And

9:35Muhammed Esgin: yeah, so we worked on so we looked at some of these like existing constructions and at the that point they were like super inefficient, like I don't remember 50 megabyte signatures and

9:43Unidentified speaker (UNKNOWN): like

9:43Muhammed Esgin: they were not even like implemented because they were like huge and so on. And we said, okay, probably we can look at improving the like the post quantum and particular lattice basing signature schemes. So that's how I started looking deeper into these things, like the underlying zero watch proofs, like this so-called one out of many proof and so on. So we looked at how we can make things post quantum secure and as well as at least reasonably efficient first and then trying to improve the efficiency uh going forward. So I

10:13Unidentified speaker (UNKNOWN): I looked at

10:14Muhammed Esgin: ring signatures first.

10:15Douglas Tuman: Okay, very cool, very cool. And so that it was your professor at the time that was interested in ring signatures or he was interested in quantum more?

10:23Unidentified speaker (UNKNOWN): No,

10:23Muhammed Esgin: he was in he he's a co inventor of ring signatures. oh

10:27Unidentified speaker (UNKNOWN): he's the co-inventor

10:28Douglas Tuman: of ring signatures wow

10:29Unidentified speaker (UNKNOWN): yeah so who's that

10:31Muhammed Esgin: is his name

10:32Douglas Tuman: okay wow

10:33Unidentified speaker (UNKNOWN): I

10:33Muhammed Esgin: think they called it something else at the time so it was called some like ad hoc group signatures or ad hoc signing with a group or something like that. I think that's the original name.

10:42Douglas Tuman: I'm working further and further up the cryptography food chain. I gotta get him on the show. I'm slowly

10:49Unidentified speaker (UNKNOWN): no

10:49Douglas Tuman: offense yeah I'm sure you you're no you're making breakthroughs still. Don't get me wrong. That's amazing that he was one of the authors of rings. So was he so was he aware of Monero obviously he's aware of Monero, but was he like I

11:00Muhammed Esgin: think I think they actually like their paper was like way before Monero, maybe like I don't know, fifteen, twenty years before maybe I don't know, ten years before Monero?

11:09Douglas Tuman: Amazing.

11:09Unidentified speaker (UNKNOWN): I

11:09Douglas Tuman: mean.

11:10Unidentified speaker (UNKNOWN): Uh and

11:11Muhammed Esgin: then I think he didn't even know that Monero was using ring signatures. That's

11:15Douglas Tuman: incredible.

11:15Muhammed Esgin: And then later he came to know that oh yeah, Monero is actually using ring signatures and so on.

11:20Douglas Tuman: It's gotta be one of the most famous use cases now though, right? Like what are the

11:24Unidentified speaker (UNKNOWN): Yeah,

11:24Muhammed Esgin: probably. Yeah, probably.

11:25Douglas Tuman: What are some of the other like like famous use cases of ring signatures that you know of? Like just to

11:30Unidentified speaker (UNKNOWN): Um

11:31Muhammed Esgin: Hmm, question. So I think I don't know if it's really deployed in practice, but I know that some e voting schemes use ring signatures.

11:39Unidentified speaker (UNKNOWN): to

11:39Muhammed Esgin: basically kind of say that yes, I'm like a legitimate user of this system and I'm not really like uh some random person right trying to do things and yeah and ah sorry, sorry, right away. He's actually okay. So let me clarify Joseph is coin manager of linkable ring signatures, not link ring signatures. Okay, that we just claim claimed.

11:57Douglas Tuman: Oh, okay, okay okay.

11:59Unidentified speaker (UNKNOWN): Which is the

11:59Muhammed Esgin: one that Monero uses. That the so the linkability is an important property.

12:03Douglas Tuman: Okay.

12:03Muhammed Esgin: Uh okay, but going back to ring synchronous, I think the original motivation was like this whistleblowing, right? So how do you actually whistleblow without actually identifying yourself? I think that

12:13Unidentified speaker (UNKNOWN): was not

12:13Muhammed Esgin: the original motivation. So you kind of give some legitimity

12:16Unidentified speaker (UNKNOWN): to

12:17Muhammed Esgin: your claim because you can say, okay, I'm one of the like, I don't know, partners in this, I don't know, like consortium, or maybe one of the employees uh in this company and I'm saying that this company or this thing is doing something that they should not be doing. I think that was one of the original

12:34Unidentified speaker (UNKNOWN): uh uh uh

12:35Muhammed Esgin: motivations. But

12:36Unidentified speaker (UNKNOWN): I

12:36Muhammed Esgin: don't really know if there is like a real life deployment uh that really uses ring signatures. Um

12:43Douglas Tuman: okay so a way for someone to prove that they were part of the company without without revealing their identity when they're in

12:50Unidentified speaker (UNKNOWN): a

12:50Douglas Tuman: whistleblowing action.

12:51Unidentified speaker (UNKNOWN): That's a thing.

12:54Douglas Tuman: Yeah. You know, it's it's it's incredible that uh you know Monero has been able to take advantage of ring signature technology, I guess we could call it.

13:04Muhammed Esgin: Yeah.

13:05Douglas Tuman: Do you think I mean at the time Monero chose ring signatures because that was really kind of the only thing that was reliable, my understanding was like kind of truly vetted and reliable. Now there's been obviously there's been a lot of progress now with zero knowledge proofs. And Monero has aspects of neuro zero knowledge proofs as well, right? Use bulletproofs, whatever. Uh, but this idea that Monero's tech was at the time using what was most trusted versus what was perhaps more theoretical, like with

13:37Muhammed Esgin: yeah,

13:37Unidentified speaker (UNKNOWN): like

13:37Douglas Tuman: let's say

13:38Unidentified speaker (UNKNOWN): comparing it

13:38Douglas Tuman: to what Zcash was using it. Uh yeah any opinions on that like thoughts on that? I mean like like you yourself said like oh he was that's what he was looking at so I looked at ring signatures it wasn't zero knowledge proofs do you see them at like is ring signatures kind of a dated technology or just maybe has

13:55Unidentified speaker (UNKNOWN): different uses

13:56Muhammed Esgin: necessarily things thing or so I think so zero knowledge proof is definitely a more powerful technology than ring signals. Ring signatures are kind of like one particular application I would say right.

14:09Unidentified speaker (UNKNOWN): You can do

14:09Douglas Tuman: different things, right?

14:10Muhammed Esgin: Yeah. So I think zero knowledge proof is a lot more powerful and you can do a lot more things. But it is also like from zero knowledge proofs and certain uh zero knowledge proofs in particular, you can go to build ring signatures as more or less like a black box. transformation. But you can also go from zero knowledge proofs to many other things, like I don't know, like verifiable random functions or group signatures or all kinds of more advanced cryptographic tools that you may want to actually use as a technology in real life. So I would say they are kind of like a bit separate things.

14:44Unidentified speaker (UNKNOWN): But

14:44Muhammed Esgin: I don't think ring signature itself is necessarily outdated. Uh it is serving a particular purpose and like you mentioned Zcash and I

14:53Unidentified speaker (UNKNOWN): be

14:53Douglas Tuman: like saying like exactly

14:54Unidentified speaker (UNKNOWN): it may

14:54Muhammed Esgin: have by the way, yeah, I don't follow the cryptocurrency world very closely. Things may have changed. So I'll let just to be clear as a disclaimer. But the one I knew about Zcash was they were using this trusted setup zero knowledge proofs,

15:07Unidentified speaker (UNKNOWN): which is

15:07Muhammed Esgin: kind of like a trade-off. Okay. So you have this trusted setup and then you can do things more efficiently efficient than later on exact

15:12Unidentified speaker (UNKNOWN): once

15:13Muhammed Esgin: you accept the trusted setup.

15:14Douglas Tuman: Right.

15:14Muhammed Esgin: And I guess my understanding is that Monero chose this like trust free path, more transparent path, and then

15:20Unidentified speaker (UNKNOWN): they went

15:20Muhammed Esgin: quick ring signatures.

15:22Douglas Tuman: Exactly.

15:22Unidentified speaker (UNKNOWN): And

15:22Muhammed Esgin: there is yeah, there is a trade-off there. So you have like a like a smaller pool of anonymity set in ring signatures or Monero compared

15:30Unidentified speaker (UNKNOWN): to

15:31Muhammed Esgin: I guess what Ccash is doing. Uh

15:33Douglas Tuman: right. And they they've since you know upgraded and gotten rid of the trusted setup. So

15:39Muhammed Esgin: really okay.

15:40Douglas Tuman: Yeah. So they've obviously that that has or you know, you're never gonna erase history. So there was initially a trusted setup, but now they've migrated to ones that are based on trusted setups, which

15:52Unidentified speaker (UNKNOWN): is really

15:53Douglas Tuman: cool.

15:53Muhammed Esgin: Do you know which proof system they are using now?

15:56Douglas Tuman: Uh I mean it's called Hay I think it's called Halo is the

15:59Muhammed Esgin: Ah Halo, okay.

16:00Douglas Tuman: Yeah, yeah, yeah.

16:00Unidentified speaker (UNKNOWN): All

16:00Muhammed Esgin: right, yeah.

16:01Douglas Tuman: Um Um you're familiar with that out, obviously. Yeah,

16:05Unidentified speaker (UNKNOWN): I

16:05Muhammed Esgin: mean I know it at the high very, very high

16:07Unidentified speaker (UNKNOWN): level.

16:08Douglas Tuman: Yeah. I mean I was gonna kind of try to save this for later, but we're kind of in this this neck of the woods here. So where do you do you do you think Monero um are how familiar are you with Monero's tech beyond the ring signatures and the fact that it's moving to full chain membership proofs?

16:24Muhammed Esgin: Uh not much, I would say.

16:26Douglas Tuman: Okay, okay. But

16:27Unidentified speaker (UNKNOWN): but

16:27Muhammed Esgin: if you maybe describe to me a little about these full chain membership pros.

16:32Douglas Tuman: Okay.

16:33Muhammed Esgin: Maybe I will get an understanding.

16:35Unidentified speaker (UNKNOWN): Alright,

16:35Douglas Tuman: well you

16:35Unidentified speaker (UNKNOWN): know

16:35Douglas Tuman: what we can get into that again later. We get into that again later. So let's let's we'll stick on the on kind of the timeline here. So you you got into Monero or became aware of Monero when you were when you you developed develop that paper for making quantum proof ring signatures essentially. right?

16:51Muhammed Esgin: Uh,

16:52Douglas Tuman: which is super cool. So has that been implemented anywhere? Is it's like so you basically you you've proven that it could be done. Like the you can theoretically have ring signatures that are quantum proof.

17:03Muhammed Esgin: Yeah. Yeah. So we have actually a few works around that. So the initial one was in I think twenty nineteen.

17:10Douglas Tuman: That's incredible that it was that many years ago, right? Like so Monero's Monero

17:13Unidentified speaker (UNKNOWN): a crazy idea. Monero's Monero is talking about

17:17Douglas Tuman: crypto has you know, there was always the mention of oh, in the future there might be a quantum thing and we'll have to figure it out. But it wasn't really till like the last two years where it was really like, yo, what are we doing with quantum and crypto? We got to start doing stuff. And you guys developed that paper in 2019, which is incredible. Well, go ahead.

17:34Muhammed Esgin: Yeah, yeah, the that's trend. right. So yeah, so we started out with like building ring signatures. and the and particularly the underlying zero knowledge proofs, which is called one out of many proofs. It's kind of like a name such as it's basically says you prove that you own one out of the many public keys that are basically just lounder, right? So that's

17:53Unidentified speaker (UNKNOWN): the

17:53Muhammed Esgin: idea, and you can do other things as well. Uh and yeah, you you can kind of convert it to ring signatures. So we started out with building these one out of many proofs and then its application. applications to ring signatures. And then we, I mean, my supervisor already knew that Monero was using the ring signature. So then the next step was to look at a more real-life application. So then we built this uh protocol that we call matrixed. So it's kind of like a wordplay of so lattices use a lot of matrices and there's lots of matrix theory involved. So it's kind of like matrix-based ring signature. and it sounds like matrix. So

18:29Douglas Tuman: I

18:29Muhammed Esgin: kind of

18:29Unidentified speaker (UNKNOWN): like the

18:30Muhammed Esgin: name.

18:32Unidentified speaker (UNKNOWN): So

18:32Muhammed Esgin: it was like a word clay there. So that was 2019. So at that point it was, I guess, a bit less efficient. And then we said, okay, probably over time we developed new techniques, and there was this matrix plus. That was, I think, 2022. So that's kind of like a more efficient form. We use sort certain of like packing techniques and more. of to like exploit the mathematical structure. There were some new mathematical tools that we can apply and make things more efficient. And then in the meantime, we also looked at the auditability idea and then said, Ah, okay, if we have like certain entities where we want to give them more power so that they can audit certain transactions to see whether there was any malicious activity going on. So we looked look at the elditable auditable setting as well. I think that was around the same time, maybe 2022-ish, if I'm not mistaken. But yeah, so to answer your question around implementation, yes, definitely. We do have implementations, especially in Matrix Plus, I think they made it open source. And it runs actually very efficiently. So I think verification was actually very fast. And I think we even said that it was faster than what Monero was using at the time. So it's

19:36Unidentified speaker (UNKNOWN): maybe

19:36Muhammed Esgin: just a few milliseconds, just I don't know, two or three. four, five, depending on parameter setting you are aiming for, of course, like the anonymity level and like the ring size and so on. Uh the transaction generation is less efficient because uh probably I don't know you may not be familiar with this, but the lattice-based protocols typically use what's called this rejection sampling step. So you try to generate a transcript, but then sometimes you need to reject and restart and re-state. redo everything. And then sometimes you may have to repeat actually a lot of time. So to make sure that you are not leaking information, that's the reason that you do this rejection sampling step. That's why I think transaction generation was around like maybe a hundred, two hundred, maybe three hundred milliseconds, but still a fraction of a second so not too bad I think.

20:20Douglas Tuman: And was it was it something that were you know, we say ring signatures, but was it something that could have easily that Monero could easily s swap its current ring signatures for, as far as you know? I mean on a technical level?

20:31Muhammed Esgin: Um I think it basically satisfies the properties that Monero asks out of ring CT, but swapping out in general, not just for ring CT, but even for I don't know, like just digital signatures or like plain encryption schemes, like I

20:45Unidentified speaker (UNKNOWN): don't

20:45Muhammed Esgin: know, public key encryption and so on, it is I think not a very easy test. I'm although I'm not a very like engineering focused person. Like there is also like kind of migration happening for TLS or DNS or all these like kind of protocols that we rely heavily on for the internet. Yeah, like kind of sizes get larger. For example, the you need to care about maybe the packet size, and then maybe okay, so you if your signature is larger large than the packet size, what do you do? So you may need to make changes in the underlying configurations of the like high. higher level protocol. So it may not be a very easy swap, but I think from a cryptographic perspective, it gives you the properties that Ring CT gives. So with certain engineering, if you sold engineering challenges, I guess uh you would be able to I guess implement it in Monero.

21:33Douglas Tuman: Do you know if there was discussion at the time when you released this paper? Was it like discussed among the Monero? devs at all?

21:40Unidentified speaker (speaker_3): I'm not

21:41Muhammed Esgin: too sure. Yeah. Not

21:43Unidentified speaker (UNKNOWN): not

21:43Muhammed Esgin: not too sure.

21:49Unidentified speaker (UNKNOWN): But I think as we

21:50Muhammed Esgin: mentioned, I think people were not really perhaps caring too much about cost quantum security at that point. I think in the last couple of years, as you mentioned, I think it has picked up a lot more.

21:59Douglas Tuman: Yeah, I don't I don't know what the implications are, like you said, from an engineering perspective. if it's like some big overhaul that would have to be done, or for it was, you know, not much engineering work on Monero side, because if if not, it would have been seems like it would have made sense to do. I don't know.

22:12Unidentified speaker (UNKNOWN): I mean obviously

22:13Muhammed Esgin: it's definitely an interesting problem to work on from an engineering side to see okay, how is the performance impacted? Can we still maintain like the properties that we want to maintain in Monero and so on. I think this is definitely interesting. At least to test and see.

22:28Douglas Tuman: Yeah. I mean now Monero is moving to full chain membership proofs. So I don't even know if this would then apply because it's not really yeah and you don't you're you're not familiar enough with full chain membership proof. So basically basically right now, right, with the ring signatures, it's the the the real spend or signature is hidden among whatever 16 sixteen right in Monero or whatever it may be.

22:48Unidentified speaker (UNKNOWN): Uh

22:49Douglas Tuman: and then with full chain membership proof proofs,

22:51Unidentified speaker (UNKNOWN): um

22:52Douglas Tuman: you don't it's it's one out of every single output that event has ever existed in Monero.

22:58Unidentified speaker (speaker_3): Yeah.

22:59Douglas Tuman: Um so that that's that's you know now whether or not tech like could your what you're proposing with ring signatures be applied to that on a on a large level

23:10Unidentified speaker (UNKNOWN): Yeah I

23:10Muhammed Esgin: mean in purity theory, yes you can just increase the ring size and make it the whole uh I don't know what whatever the size of the whole system is.

23:18Douglas Tuman: Yeah.

23:19Muhammed Esgin: But

23:19Unidentified speaker (UNKNOWN): you're talking about like astronomical cost.

23:22Douglas Tuman: Yeah, it would be a like like an insane amount of uh right, transaction size and whatnot.

23:29Muhammed Esgin: Uh I guess it wouldn't be insane because it grows kind of logarithmically logarithmic or polar logarithmically logarithmic way with the size of the ring. So I think we even tested like uh let me just quickly check actually I think we have it. Like I think we even said okay if the like the ring size is total to twenty then maybe you are looking at I don't know probably still less than a hundred kilobytes.

23:50Douglas Tuman: We we we need you at a Monerotopia conference so you could talk to uh

23:54Unidentified speaker (UNKNOWN): the

23:54Douglas Tuman: Monero devs directly over over a attacker. talker

23:59Muhammed Esgin: but I I guess in that at that point I think the computational time might be the main bottleneck. Because then if the ring signature inherently requires processing all the public keys, I mean otherwise if you are not processing let's say a particular public key, then the outsiders can tell that okay this public key is never used because you are never using it in the computation. So somehow all of the public keys in the ring needs to be involved in the computation, especially verification. and of course the generation of the signature as well.

24:30Unidentified speaker (UNKNOWN): So

24:30Muhammed Esgin: that you are actually you can hide yourself within like you can count that as a decoy public key, right?

24:35Douglas Tuman: Yeah, yeah, yeah.

24:36Unidentified speaker (UNKNOWN): So

24:36Muhammed Esgin: then if you are dealing with, I don't know, like a million or I don't know, even more than a million public keys, then this is going to take you a lot of processing time to actually okay do something with all of those keys.

24:48Douglas Tuman: Yeah, yeah, of course. That does not sound scaly. scalable at all.

24:51Unidentified speaker (UNKNOWN): Um

24:52Douglas Tuman: yeah, I mean obviously that there's talk in Monero of how we're going to become more, you know, essentially become quantum resistant. Um with full chain membership proofs we get forward secrecy. Uh but yeah Monero, I I'm I'm not under I don't understand it well enough, but I know that Luke Parker, that's what he wants to focus on next, is trying to make Monero quantum secure. From a high level, I mean theoretically, is it possible to make these blockchains, whether it's Monero or Zcash, truly like quantum secure with on with regards to all of its elements?

25:29Muhammed Esgin: Yeah, I think in general, when you want to go from like quantum vulnerable to quantum safe, especially using lattice-based lattice based cryptography, like the theoretical feasibility question the answer is almost always yes. I don't think there is any like no answer that I mean that are you can't really do this using lattice based schemes. The question is really more about performance efficiency and like what's the communication cost, what's the uh like I don't know, computational cost. So the the the costs can be huge, but in theory yes you can do it.

26:00Douglas Tuman: Okay.

26:00Unidentified speaker (UNKNOWN): Uh

26:01Muhammed Esgin: so how efficiently we do it is the main important question, in my view.

26:05Douglas Tuman: Well, let's get into the weeds a little bit here too. Just so on a high left, so you like you're saying lattice. Like I don't really know like what that what that is, right? And there's these different types, right? There's like people are using hashes, they're using lattices, and what what was the third the third type? Uh

26:18Muhammed Esgin: isogenies? Is it

26:20Douglas Tuman: yeah, yeah. Yeah, how do you say that? What is it?

26:22Muhammed Esgin: Uh isogeny, is Isology o G E N and W. what

26:26Unidentified speaker (UNKNOWN): Yeah,

26:26Douglas Tuman: yeah. So there's those three those three different types. What what are can you can explain a little bit what those different approaches are?

26:34Muhammed Esgin: Yeah.

26:34Unidentified speaker (UNKNOWN): And then I guess we just stop. But

26:36Douglas Tuman: cryptocurrency like Monero and Zcash are based on elliptic curve cryptography. So does that have an effect would those architectures have to change for these different, you know, for lattices, for example, to be used? But yeah, get into that. Get into that.

26:51Muhammed Esgin: Yep. Yeah. So basically it's all for, I guess, roughly three decades now. We know that if we have like scalable quantum computers that can solve certain mathematical problems very efficiently, and one is integer factorization, which underlies like RSA and RSA like schemes. And the other one is discrete look log uh problem, which is basically like elliptic curves, uh elliptic curve, discrete logarithm problem. But yeah, this needs kind of like more scalable quantum computers. So we are now moving away from these like two fundamental problems: integer factorization and the safe logarithm or elliptic curve uh cryptography, into yeah, what we call post-quantum cryptography or quantum resistant cryptography. And there are different approaches. I guess the to start with the easiest one, one is using hash functions. as the underlying building block. And the only assumption you made is that you kind of have like a cryptographically secure hash function. And I guess for this approach, you are not necessarily relying on the underlying mathematical structure of the hash function. You just assume that we have a hash function and then you just utilize the property, general properties of the hash function, and then

27:59that's how you build your higher level construction. And you it's I guess much easier to just swap out the hash function if that hash function turns out to be insecure or not at the secure level that you desire.

28:11Unidentified speaker (UNKNOWN): And

28:11Muhammed Esgin: this is seen as the most conservative option because hash functions do not typically inherit like very mathematical structures that the adversaries might be looking to exploit. But it also has a the drawback that since you don't have this mathematical structure. you also cannot build many things, especially if you care about efficiency. So, for example, you can't build encryption schemes out of hash functions, because hash functions are not one-to-one. So, and then so there's hash hash-based crypto.

28:42Douglas Tuman: There

28:42Muhammed Esgin: is lattice-based crypto, which is again the very promising, one of the very promising approaches. And most of the NIST standards are actually based on lattice-based crypto. cryptography. And this is very broad, it relies on a mathematical object called lattices. These are kind of like uh like points in space, but these are discrete points. So I don't know, I guess it's a bit difficult to visualize without having any slides, but yeah you just kind of think of a grid and then you are taking certain points in the grid which are kind of like very nice nicely structured.

29:14Douglas Tuman: So that's

29:15Muhammed Esgin: what we like maybe two dimension, three dimension. You can imagine how that might actually in your background, like the at the top ceiling, that kind of looks like a very simple two-dimensional lattice. Like if you just take the corners. But then you scale this up to a thousand dimensions. So that's the lattice that we are working with. So it's very complex object, I guess, to really

29:36Douglas Tuman: wait. and so but explain on a on a on a mathematical level how that's used to thwart a quantum computer.

29:43Muhammed Esgin: Yeah okay so there are certain mathematical problems not even actually concerned with cryptography that people have been working on for quite some time. So lattice theory is not just something that is specific cryptography that actually comes from mathematics. And and people have looked at okay, but suppose I pick a random point in space, not necessarily lattice point. Can I find the closest lattice point to this uh random point in space? But again, we are thinking about like a thousand-dimensional space, not just I don't know, a few dimensional. And you can even you can see, I don't know, if you're just looking at uh two dimensions, it's very easy. You can even just I don't know, by just looking at it, you can tell I actually this is the closest one perhaps. But that as the dimensions grow, it gets a lot more difficult because you don't know which dimension you are supposed to move. to get closer to the lattice point. So this closest vector problem is one of the fundamental mathematical problems. And there's also something called short vector problem. So I give you a kind of like a basis that generates this lattice, or all of the lattice points

30:43you can generate out of the vectors that I give to you. So given this like basis vector, can you find the clo c at uh the shortest number? non zero vector in the lattice? So that's another like very hard problem that we uh believe resists quantum computing attacks.

30:59Douglas Tuman: Okay. And

30:59Unidentified speaker (UNKNOWN): these

31:00Muhammed Esgin: problems, as I mentioned, come from mathematics. They are not necessarily coming from cryptography, but uh so early on, very smart people looked at this problem and they realized that actually we can use this mathematical problems to build cryptographic schemes. So ITI had a very breakthrough paper and then uh which kind of led to this short integer solutions problem, cis problem.

31:20Unidentified speaker (UNKNOWN): Okay.

31:21Douglas Tuman: And

31:21Muhammed Esgin: then Regel uh kind of introduced this learning with errors problem which led to more of the encryption side of things.

31:28Douglas Tuman: Okay. Now I know you it's just to rehash this uh why would hash how do hashes that then differ from that? Because I'm I'm I'm getting I'm getting what you're saying here with uh lattices. So then

31:39Unidentified speaker (UNKNOWN): yeah, so the

31:40Muhammed Esgin: lattices can have a lot of mathematical structure because they are really coming from very mathematical objects. And so it has good side and bad side. The good side is that we can exploit this mathematical structure to build things very efficiently and also build So very fancy things. For example, fully homomorphic encryption, the only way we know how to build it so far uses lattice-based cryptography. And then any all other, I don't know, zero-node scrues or I don't know, other fancy types of digital signatures, encryption schemes, and so on, that you can also build them at least relatively efficiently from lattices. And when I maybe I should also make a distinction that there are also hash functions that are mathematically structured. So typically, when people talk about hash-based cryptography for post-quantum cryptography purposes, they talk about like these unstructured hash functions like SHA-2, SHA-256, or SHA-3, Shake, and so on and so forth. Like these more of the unstructured hash functions. And those

32:41ones do not uh inhibit any m or any uh like easily understandable mathematical structure that we can use to exploit. So in that also kind of reduces the attack surface a bit. So you can't exploit it for like constructive purposes, but also the adversaries are unlikely to exploit any mathematical structure to break the hash function.

33:04Douglas Tuman: Okay. So

33:04Muhammed Esgin: that's the advantage.

33:06Douglas Tuman: And then how about isogenies? So

33:08Muhammed Esgin: isogenies, so they kind of build on elliptic curves. Uh so I'm definitely not an expert on isogeny-based isogeny based cryptography or isogeny. isogenies in general, but yeah, they are kind of like there is this again mathematical object called isogeny, and then you have isogenies between different elliptic curves, and then you can again build or define mathematically hard problems first, and then build uh cryptographic tools uh based on those mathematical problems. So there were like constructions called like psych and

33:37Unidentified speaker (UNKNOWN): uh

33:38Muhammed Esgin: some other ones that I don't really want. There were lots of different names, but

33:42Unidentified speaker (UNKNOWN): as you know

33:43Muhammed Esgin: in cryptography there are many names, but some

33:45Unidentified speaker (UNKNOWN): of them

33:46Muhammed Esgin: turned out to be insecure. So that was the reason I guess people are now looking at isogeny-based cryptography a bit more kind of hesitantly, I guess.

34:01Unidentified speaker (UNKNOWN): Do

34:01Unidentified speaker (speaker_1): you love coffee and make Monero as much as we do? Consider making gratuitous.org your daily cup. Pay with Monero for premium fresh beans, and if you like what you taste, send a digital cash tip directly to the farmers that made it possible. Proceeds help us grow this channel, gratuitous and monero.

34:26Unidentified speaker (UNKNOWN): Yeah,

34:27Douglas Tuman: because I that was one of the questions I got first to bring up to you. Can you ask him about the big switch to hash based cryptography as opposed to as opposed to lattices and isogenies? And and then he also said isog uh isogenies has been talked about in Monero channels as the way to go for quantum, but seems like other big e ecosystems are going a different way.

34:51Muhammed Esgin: Yeah. So this is my personal opinion. It's

34:53Unidentified speaker (UNKNOWN): great.

34:54Muhammed Esgin: For a like very large scale real life system that has a lot of uh I guess money involved, I would personally not trust isogenes at this point to really have like widespread deployment. Uh and I think there is still some time for isogeny isogen based cryptography to reach a certain maturity level. I'm not saying that like we should just rule it out, but I'm just saying that it has it needs a bit more time to get to the maturity that hash based and lattice based cryptography uh has at home. And when I say lattice based cryptography, there are also crazy assumptions you can make, I should say. But I'm talking more about like standard lattice based cryptography assumption. Um yeah so I would yeah I personally not maybe go for isogen isogenes based by cryptography. But it's it versus lattice space.

35:40Douglas Tuman: That's that's interesting that the Monero crowd is talking. I don't know why, I don't know why they're leaning in that direction, especially considering right Monero traditionally goes more conservative, my right,

35:49Unidentified speaker (UNKNOWN): like

35:49Douglas Tuman: when we say race.

35:50Unidentified speaker (UNKNOWN): I think

35:50Muhammed Esgin: maybe one reason is because uh isogenies can give you very compact

35:55Unidentified speaker (UNKNOWN): uh very

35:56Muhammed Esgin: protocol all outputs.

35:57Unidentified speaker (UNKNOWN): Okay.

35:58Douglas Tuman: But

35:58Muhammed Esgin: the computation is typically technically very slow. I think maybe that was that could have been one of the reasons that they said, ah, maybe isogenies might be interesting. And actually before these uh attacks against certain isogeny based schemes, isogenies were I guess believed to be kind of more uh maybe safer. But now I think uh the trust has uh went down a bit with these like practical attacks in particular. But okay, let me go back to the hash versus lattices.

36:26Unidentified speaker (UNKNOWN): Okay.

36:27Douglas Tuman: go ahead.

36:27Muhammed Esgin: I personally have a lot of confidence in lattice-based proposals that rely on these standard assumptions like learning bit errors, uh, short integer solutions or LWECs, and their standard uh like ring-based variants. So maybe I won't get into the technical terms, but yeah, maybe those uh fundamental two ones because so NIST has standardized. uh public key encryption or key and cancellation mechanisms and digital signatures that are based on the variance of these problems and the let's call it like non-crazy variants of these problems. Okay, and so these will be basically deployed all around the world to protect, I don't know, like the internet, all the communications like TLS, maybe DNS, SEC, and so on and so on. forth. So they will be basically protecting, I don't know, millions or even like billions of people and a lot more

37:23Unidentified speaker (UNKNOWN): uh

37:24Muhammed Esgin: dollars worth of assets, I guess, including I don't know, like defense military, including I don't know, US, Australia, and many other countries. So if these entities are relying on lattice based cryptography for uh digital. security, I would think that I don't know, like a cryptocurrency system or a blockchain system should also be able to rely on these. Uh because there

37:48Unidentified speaker (UNKNOWN): is an extent

37:49Muhammed Esgin: even for like US military or I don't know Australian military to be using these, right?

37:54Unidentified speaker (UNKNOWN): Right.

37:54Douglas Tuman: So

37:54Muhammed Esgin: if they think that these are not really secure and there is some doubt in their security, I think they would also definitely be not using those.

38:02Unidentified speaker (UNKNOWN): Right.

38:03Douglas Tuman: About about this, you're not to go too far off a tangent, but like there's just this idea, there's always this, you know, fear or conspiracy theory, right? That even like elliptic curves, you might maybe you can only trust, you know, some of them, not others. Some are like kind of considered more open source essentially than others. There's theories that like the NSA, right? Well, we know the NSA actually has cre been the creators of some of. these, right? And it's like could they

38:29Unidentified speaker (UNKNOWN): backdoor into it?

38:31Douglas Tuman: Right. What what are your thoughts there uh uh on that?

38:34Muhammed Esgin: Uh I think in the last maybe couple of decades there has been a significant shift in that respect. I think previously definitely uh certain agencies I think believed believe that they were quite ahead of other countries. in terms of their understanding of cryptography and crypto analysis. And I think they were confident that yes, we can set things at a level that we can break, but not the others. But I think that gap between me versus the others have significantly uh sh shrinked. And I don't think any agency at the moment can confidently say that. yes, I can break this and still keep using that encryption scheme or I don't know, digital signature, whatever, and confidently say that the others cannot break. So I think because of that uh I guess gap in technological

39:29Douglas Tuman: Especially especially with AI now, right? Like

39:31Unidentified speaker (UNKNOWN): Yeah,

39:31Muhammed Esgin: exactly.

39:32Unidentified speaker (UNKNOWN): So for example,

39:32Muhammed Esgin: let's suppose I have my AI model, I don't know, like it's like the fanciest one, and I know that it can break, let's say, this ML MLCAM, which is one of the standards out of NIST. Okay. I think there is no way I can confidently say that my competitors will not be able to have access to that powerful AI that uh would be able to break the same thing or that would be able to come up with the same result. So I don't think any agency around the world would be able to confidently say that. So I think I don't think that it is very easy to really put a tractor in these or deck door in these uh standardized algorithms in a way that you or one entity can exploit it while the others can't.

40:14Douglas Tuman: That's that's a good point. Do you know what Z so Zcash is talking about, you know, making itself, you know, post quantum, right? Um do you know so are they using lattices then? Or do you know like what the

40:26Unidentified speaker (UNKNOWN): uh I'm

40:27Muhammed Esgin: not sure what Zcash is doing now. But yeah, I actually just before I saw one of the comments on uh Twitter X and somebody mentioned about this like Zcash and Quantum. I quickly did a Google search and what came up was more around the digital signature. I don't know if they are changing the like underlying zero launch knowledge proof yet, but maybe they are, I don't know. But yeah, I yeah, I'm not too much, I guess. uh aware of what they're doing.

40:52Douglas Tuman: So you have do you obviously you see quantum as a real threat, or just like a theoretical, you know, fun problem to solve? Let's make it quantum. But do you see it as a is an impending threat?

41:03Unidentified speaker (UNKNOWN): Um

41:04Muhammed Esgin: I guess over the years it was always like 10 years away, whenever you asked, like it was always a decade away. But I think now, especially what I'm hearing from the X, experts, I'm definitely not a quantum computing expert, just to be clear. But what I'm hearing from the quantum computing experts is that they are really warning that uh what people call cryptographically relevant quantum computers might be coming in the next few years. And there is like Professor Scott Aronson from I think he's from the University of Texas, if I'm not mistaken. So he actually has a blog post and he openly said, like, this is your warning if by in the next few years quantum computers start breaking today's cryptography, don't tell me that I did not warn you. And he said, like the experts that he talks to are pointing to like 2029 and uh the years around it. So I think that's a very significant warning and

41:55Unidentified speaker (UNKNOWN): yeah,

41:55Douglas Tuman: that's

41:55Unidentified speaker (UNKNOWN): a

41:55Douglas Tuman: pretty significant and and that be that being the time at which quantum computers are already good enough to break things like crypto.

42:03Muhammed Esgin: Yeah, I think people are really thinking that in the next few years, like not not ten years, in the next I don't know, like three, five, whatever years, you will really have very useful quantum computers and they will be very relevant for uh cryptography that is used uh traditionally.

42:18Douglas Tuman: That's crazy, man.

42:19Muhammed Esgin: Yeah, it is crazy.

42:20Douglas Tuman: And it it it's basically Shore Shore's right, Shore's algorithm right, what is it called?

42:25Unidentified speaker (UNKNOWN): Yeah,

42:25Muhammed Esgin: Shore's algorithm, yeah.

42:27Unidentified speaker (UNKNOWN): Yeah,

42:27Douglas Tuman: so can you explain that a little bit? When um

42:30Unidentified speaker (UNKNOWN): Um

42:31Muhammed Esgin: Yeah, it's basically a quantum algorithm that really needs quantum hardware to build. And it can solve certain mathematical problems. So it's basically all very good at finding cycles in or maybe let let me let me not explain it, but maybe it's basically capable of breaking the underlying problems for RSA and any particular cryptography. I guess that's the easiest way to put it, like discrete logarithm problem, an integer factorization problem. And it can solve these in what we call polynomial time, which basically means that even if you increase the key size, the attacker the attacker's cost will not exponentially grow. So you need to really have very huge sizes to be able to have uh reasonable security. And in recent years people have been looking at estimating the cost of running Shore's algorithm and it seems like the costs are going down. Uh so that's why I guess the pe people are getting more worried to

43:31Unidentified speaker (UNKNOWN): Right.

43:31Douglas Tuman: That that that estimate that that that you gave, what whatever it was twenty 2029 twenty nine is based on right that really, right? How close

43:38Muhammed Esgin: Yeah, that's right. So there are like improvements on the quantum hardware side. So how do you build more capable quantum computers? And then there is also the more algorithmic side value as okay, how can I more efficiently like break today's crypto systems, right?

43:52Douglas Tuman: Right. So

43:53Muhammed Esgin: but the costs at the top go down and the hardware is become more capable, and at some point they will right

43:59Douglas Tuman: kind

43:59Muhammed Esgin: of converge. And I don't know when exactly this will happen, but do you expect explain two years.

44:04Douglas Tuman: And and when that that prediction was made, was that how how long ago was that?

44:08Muhammed Esgin: Uh it was this year. So I think maybe uh a few months ago. Yeah. Wow.

44:12Unidentified speaker (UNKNOWN): So

44:13Muhammed Esgin: it's publicly available. I don't know. People can Google Scott Aronson and

44:16Douglas Tuman: Yeah.

44:17Unidentified speaker (UNKNOWN): And I

44:18Douglas Tuman: mean in AI, right? With the advent of AI I mean AI has been around for for some time, but um that's gotta be speeding up the pace, right?

44:27Muhammed Esgin: Yeah, definitely. Like people are kind of using AI to solve certain mathematical problems, and

44:33Douglas Tuman: yeah, I

44:33Unidentified speaker (UNKNOWN): think it will

44:34Muhammed Esgin: definitely help with coming up with better algorithms to crypto analyze crypto systems as well. Uh but maybe also on the defensive side, maybe we will be able to use AI to come up with a kind of crypto systems that are more efficient, but still at a desired. security level.

44:51Douglas Tuman: Big picture theoretically like I mean is it just gonna be always be a constant cat and mouse uh battle? Do we get is encryption here to stay? Is there a point in time where you think theoretically like just encryption gets broken, like it doesn't exist as a concept? Or is it like fundamentally Mentally part of the universe that there's always things you can create that can't be decrypted even though we have AI that has tremendous access to amounts of endless amounts of energy. Like

45:22Muhammed Esgin: yeah, I mean theoretically speaking, there are encryption schemes that are uh kind of like unbreakable information theoretically. And what this means, like there's this so called one time pad. And this dates back on many years ago. I don't exactly know when people came up with this. So if you kind of generate truly random string, which you use as the key, but it needs to be as long as your message that you are encrypting. And if you kind of just exhort your message with this like truly random key, then there is no way that even an unpowerful attacker has no way of actually distinguishing this The ciphertext from just some uh completely random string. No AI or no advances in I don't know crypto analysis will help to break this. But it is of course not practically efficient because you need to generate these very long strings for all of your messages that you have and you need to kind of not reuse it ever and so on and so forth. But would we ever need to actually go down to these very primitive methods? I personally don't think so. I think the encryption will serve

46:26Unidentified speaker (UNKNOWN): the same thing.

46:26Muhammed Esgin: survive for a long period of time. And I doubt that we would need to kind of constantly keep upgrading encryption schemes every, I don't know, two years or something. I think they will stay, like just like RSA and Diffie Hellman or like the description stayed for a long time. I think lattice-based cryptography, especially Selected like the standard send it ones and the hash-based cryptography will survive for a reasonable period. of time. But maybe even when we may need to increase the key lengths, or maybe we may need to make some tweaks. Uh, but I don't think they will be completely broken, even with advances in AIM.

47:03Douglas Tuman: Yeah, yeah, yeah. Uh what do you you know, there's this idea that obviously, like you're saying, right, by 2029, it could be a real risk to crypto, but it comes with potentially a risk, right? If we start implementing um these changes um to be quantum proof, just in case. But that adds complex complexity complexity. Uh, and then like we're saying with AI, it can make create perhaps a greater attack surface in terms of uh finding flaws, finding claw flaws in the math or the implementation. How do you how do you think we should be striking that balance? I mean, is it you know. kind of the risk reward and is it worth going fully quantum proof at the risk of, you know, perhaps um adding potential exploits, things like that?

47:51Unidentified speaker (UNKNOWN): Yeah,

47:51Muhammed Esgin: I think there is definitely a risk there as well, especially more on the engineering side perhaps.

47:57Unidentified speaker (UNKNOWN): So

47:57Muhammed Esgin: I don't know, like people have been implementing like this RSA or DiscFlog Disk Log log-based based schemes for quite some time. And we can say that I guess there's kind of like a reasonable cohort of people who know how to do this properly. And we have like open source libraries that people can uh adopt for uh their application settings. And yeah, we have definitely not been implementing lattice based cryptography that long.

48:22Unidentified speaker (UNKNOWN): Uh

48:22Muhammed Esgin: and yeah, there is some risk there, but I I guess I would recommend actually using these like reliable libraries that are coming out of uh I guess trusted people. So I don't know, like OpenSSL has now certain extensions uh that support uh post quantum crypto systems. And

48:39Unidentified speaker (UNKNOWN): I

48:39Muhammed Esgin: know Google has a library. There are certain companies that they uh they have their own uh libraries. I think some of them are open source as well. Uh so I think relying on these open source libraries. as much as possible is the best way to go. But typically they would not have these very advanced or fancy cryptographic uh systems. I guess there is a bigger challenge there. And I guess the way to uh tackle that as much as possible is to do like open sourcing and also getting people to test this as much as possible, maybe like this uh bug bounties where p if people see that there is some issue, then we uh like reward this, or more and more people actually look at this and try to kind of validate that yes, it's indeed doing what it's supposed to do, and it's not kind of like leaking extra information or it's not doing something unintended. Uh but yeah, I agree with the risk, but like we have to do this migration because I guess we can't rely on I think the quantum threat flat is not hypothetical, so we have to do the migration.

49:39And I guess the best way to go about this and minimize the risk is starting early so that you are not rushing into and just saying, ah, okay, I don't know, I have to do this in like six months' time. Let me just grab the easiest solution and just deploy it. I think that's kind of like the worst solution than saying, okay, I need to start thinking about this now. And I had maybe, I don't know, two or three, whatever years ahead, and I want to complete my transition. to post quantum security by this state and then take gradual and progressive steps that are as safe as possible. I think that's the best way to

50:11Douglas Tuman: and when we say post quantum, I mean i I don't fully understand even what we kind of mean by because there's there's different things, right? There's post quantum in terms of privacy, right? And there's post quantum in terms of preserving the actual elliptic curve cryptographies. cryptography so people can't just steal your crypto, steal your keys. Um, so do you see a path towards let's say I guess you haven't looked at Monero that closely, but towards making Monero completely, you know, post quantum? You think that's that's possible on all fronts? Privacy front and uh

50:45Muhammed Esgin: Yeah, I think so. So yeah, we have these works that's that kind of like uh take. recent CT into the quantum safe setting. And then uh I'm guessing like there's like digital signatures used in Monero, so there's already like standards out for that. It's much easier to replace digital signatures.

51:05Unidentified speaker (UNKNOWN): Um

51:06Douglas Tuman: And do you do you think it should be done in like you think it should be done in piecemeal or that is you know there should be kind of some grand solution to make Monero of completely post quantum. I guess that's what I'm struggling with.

51:25Muhammed Esgin: I think progressively moving towards quantum security is the best option because you don't want to suddenly say, Okay, I'm gonna I don't know, we have like these five quantum molecular cryptographic tools that we are about going to migrate everything at once and I think this is a lot more difficult and a lot more risky than saying, Oh, okay, let's fix let's kind of fix the problem with the digital signature first. Kind of that's done, and we can see that okay, the system is still working, nothing is broken, and it seems to be all fine. You kind of like tick that box and you say, Okay, what's the next thing? Maybe you look at like the string C CT T and then you try to kind of like make that quantum safe. I think this these like progressive steps is more appropriate than just kind of changing everything in one go. But of course, I think it's definitely a good idea to still think about okay, I have like these current quantum vulnerable things I need to migrate and explore the solutions for that. But I think the actual deployment should come in phases rather than everything changing in one go. I think that to me seems like

52:27Unidentified speaker (UNKNOWN): a

52:27Muhammed Esgin: safer option.

52:28Douglas Tuman: If you were designing your crypto from scratch today, what like what like what tech do you think should be should be utilized? I guess lattice you would say lattices, right?

52:39Muhammed Esgin: Yeah, I would yeah I think I have confidence in lattice-based cryptography and I mean hash based cryptography also has the disadvantage that you can't build like these encryption schemes that I mentioned. It's basically just not just not possible. And I mean, if you have to rely on lattice based cryptography at some point, anyways, especially for encryption, which is concerning confidentiality and people's private information, like if you already have to trust it there, like why are we paying a high price for other aspects, maybe relating to authenticity or maybe I don't know, like the privacy of these like zip. zero knowledge proofs and so

53:16Unidentified speaker (UNKNOWN): on. Uh so

53:17Muhammed Esgin: hash-based cryptography typically comes at a higher cost, performance cost than lattice-based cryptography. So if you're already relying lattices at some point, I don't know, why not rely it somewhere else? If it is based on similar standard assumptions. Yeah, so I would probably base many of the things on uh a lattice-based lot of space cryptography. If I were to build a cryptocurrency from scratching.

53:41Douglas Tuman: So how do how do we get you to make Monero post quantum? What do we gotta do? We get we'll give we'll give you some Monero, we'll raise some Monero from the community. I mean

53:48Unidentified speaker (UNKNOWN): Can

53:48Douglas Tuman: you do it? You think you you can make Monero post quantum?

53:51Muhammed Esgin: Uh I I think so. I don't know like what other cryptographic tools Monero uses users beyond like I

53:57Unidentified speaker (UNKNOWN): don't

53:57Muhammed Esgin: know hash functions, digital signatures and ring C T. I guess those are the

54:01Unidentified speaker (UNKNOWN): Yeah

54:01Douglas Tuman: you you got you got to look gotta look at the full chain membership proof. So cure I'm

54:06Unidentified speaker (UNKNOWN): tape would

54:06Douglas Tuman: be there.

54:07Unidentified speaker (UNKNOWN): This one

54:07Muhammed Esgin: so that

54:08Unidentified speaker (UNKNOWN): one

54:08Muhammed Esgin: might be that one is probably more difficult. So

54:10Unidentified speaker (UNKNOWN): at

54:11Muhammed Esgin: the moment

54:12Unidentified speaker (UNKNOWN): uh I

54:12Muhammed Esgin: guess it's difficult to really support a like anonymous membership proof that scales to like I don't know like maybe millions of uh parties. So doing this efficiently is a challenge unless you want to do like some kind of like trusted setup which I guess Monero is not really

54:30Douglas Tuman: no

54:30Muhammed Esgin: much into. Yeah. So uh yeah that's a good I guess research problem to have and yeah I'm I don't know if people from Monero are interested in exploring this I'm I don't know happy to have a chance and kind of think more deeply around this boat both in terms of like I don't know deploying what we developed before into Monero as well as developing more Monero centric cryptographic tools that solve like I don't know as you must like full chain membership problem and maybe related things as well.

55:04Douglas Tuman: Like I said, we gotta get you in a room with uh Luke Parker, some of the other Monero devs.

55:08Muhammed Esgin: Yeah, yeah, I'm happy to

55:09Douglas Tuman: be awesome. What um so what is your current take on crypto? cryptocurrency then? Right? So you've obviously you're more you're more interested in the act in the actual cryptography and build building the cryptography. And yeah, people go use it for whatever they need to use it for. But really, you know, and it's uh you must have a beautiful mind, man. The fact that like that that is what motivated, which is amazing, but you must have some uh pretty deep deep-seated seated opinion of of cryptocurrency at this point, right?

55:40Muhammed Esgin: Uh, I think it's definitely like in general cryptocurrencies and blockchain application. I think this is a like a very good idea, and there are definitely very good use cases for blockchain technologies. Uh, but there's also, I guess, a lot of kind of I mean, j just like a with any other technology, like I don't know, with AI and quantum technologies, there is also a lot of hype around this and people just try to use the I don't know blockchain keyword to put into many things, even though maybe blockchain is not needed necessarily in that setting. So I think i i it it's

56:16Unidentified speaker (UNKNOWN): about

56:16Douglas Tuman: like the core the core use case, right? Like the the digital cash use case, which

56:20Unidentified speaker (UNKNOWN): is

56:20Douglas Tuman: what Bitcoin was supposedly embedded. amazing

56:22Unidentified speaker (UNKNOWN): Yeah, yeah, yeah. I think

56:23Muhammed Esgin: that that that is definitely a very good idea and yeah, I think it's also great that these ideas came up and then that kind of also uh gave more emphasis on cryptography as well. So w which is I think very good. So I think it's a win win for both sides. So it's a win for blockchain because it's kind of empowered by cryptography, but it's also a win for cryptography because it there's a lot more attention to on cryptography at the moment, thanks to the blockchain. boxing.

56:47Douglas Tuman: It t turned you guys into superstars, man. You guys are like You know, if cryptocurrency didn't exist, you go to a bar, you're like, yeah, I'm a cryptographer and trick is like what? Oh my god, you create the actual map

57:03Unidentified speaker (speaker_3): Yeah.

57:05Unidentified speaker (UNKNOWN): You're

57:05Douglas Tuman: a superstar now.

57:07Unidentified speaker (speaker_3): Uh yeah we'll see.

57:10Douglas Tuman: This has been great man and we didn't even get into your other your other paper. Don't you you

57:14Unidentified speaker (UNKNOWN): have

57:14Douglas Tuman: like three papers that are out right now or that you're part of, right?

57:18Unidentified speaker (UNKNOWN): Yeah, so

57:18Muhammed Esgin: we are w working on other cryptographic, more advanced cryptographic tools that have interesting applications in the blockchain application space as well, including cryptocurrencies. So we have been working on something called verifiable random function that some of the systems use, like Algorand and the Definity, I think uses that as well.

57:37Douglas Tuman: Okay.

57:37Muhammed Esgin: And we

57:38Unidentified speaker (UNKNOWN): have work on this

57:39Muhammed Esgin: multi uh verifiable random function, VRF. So this is mainly used for this leader election protocols. So to select like the next proposal for the blocks.

57:49Douglas Tuman: Okay.

57:50Muhammed Esgin: Uh like for voting and so on.

57:52Douglas Tuman: Very cool.

57:52Muhammed Esgin: So we have been working on multi-signatures that, for example, Ethereum uses uh so we have a very efficient uh lattice-based multi-signature scheme that is particularly well suited for blockchain applications. Uh uh we have a threshold ring signature scheme where you don't have kind of like just one signer, but you have a set of signers. So I don't know, like let's say five parties, and they hide themselves in a set of I don't know, Or whatever other another twenty 20 parties. And then you can really attest to the fact that okay, at least five out of these twenty 20 people have validated this message, or maybe they signed this message. So it's also useful for I think consensus. And also maybe if you want to make sure that before a transaction happens, I don't know, like five parties or five devices authorize this, not just one device. So I think it's helpful for distributing the trust before uh a like transaction is authorized. Uh yeah, I guess those are the main ones.

58:55Unidentified speaker (UNKNOWN): Yeah, it's

58:55Douglas Tuman: a that's a lot.

58:57Unidentified speaker (speaker_3): Yes.

58:57Douglas Tuman: Well what is it that you want to work on next? Is there are there things that you're like excited about that you would like to work on in the cryptography film?

59:05Muhammed Esgin: Yeah, I think in in recent years I have been kind of a bit focused or maybe a bit obsessed about this thing called oblivious pseudo pseudo-random random function. Kind of like oblivious evaluation of a pseudo-random function, like two-party evaluation. That

59:20Unidentified speaker (UNKNOWN): seems to be

59:20Muhammed Esgin: the

59:21Douglas Tuman: what what is it?

59:23Muhammed Esgin: So uh like do you know about PRFs, uh pseudo-random functions?

59:28Unidentified speaker (UNKNOWN): No.

59:29Douglas Tuman: I

59:29Unidentified speaker (UNKNOWN): mean you

59:29Muhammed Esgin: can kind of just think of this as kind of like another like a yes encryption, right?

59:33Unidentified speaker (UNKNOWN): Okay.

59:33Muhammed Esgin: So you have kind of like some message and then you have a key that you use to encrypt the message, right?

59:37Douglas Tuman: Okay. And

59:38Muhammed Esgin: typically in AES encryption, there's just one party that holds both the message and the key, right? And it gets kind of like a deterministic output output out of this once you risk the key and the message. Okay. But then let's suppose now we move this to the two-party setting. Let's say I as a client hold the message. And let's suppose you are like some service provider and you hold the key, right? Then and how do we now evaluate this AES function in a two-party setting where I don't know the key and you don't know the message. So we need to run some interactive protocol to come up with the uh output of the encryption. And in general this is useful for like pseudo random functions. You can do this. And this has actually very real life application. It's already deployed in I don't know like by like on Meta, Google, Amazon, Web Services and so on. So it has very interesting real life applications. So we are trying to build more efficient oblivious pseudo random functions, but it seems to be very challenging. I don't know, there are some uh very uh technically challenging aspects to build this very efficiently in the post quantum setting.

1:00:42Unidentified speaker (UNKNOWN): Yes

1:00:42Muhammed Esgin: I thought

1:00:42Unidentified speaker (UNKNOWN): that's

1:00:43Muhammed Esgin: something that I'm uh working on at the moment and yeah there is also lots of problems around building efficient advanced uh cryptographic tools while relying on post-quantum cryptographic approaches.

1:00:55Douglas Tuman: Very cool, man. Very exciting. And you're using AI in your pursuits?

1:01:01Unidentified speaker (UNKNOWN): Uh

1:01:02Muhammed Esgin: a bit, probably not as much as some other people. I think there are people that are very extensively using AI. I think I'm not using it that much. I'm definitely using some, but not that much. I'm also not I guess, using it very efficiently. Because some people have these like softwares and they say, ah, go do this with these files, or I don't know, just ask these AI agents to do many of the things. Yeah.

1:01:27Unidentified speaker (UNKNOWN): I'm

1:01:27Muhammed Esgin: basically just using like the browser, and I'm just I don't know, I'm using in more of the chat mode rather than more advanced functionalities, I guess.

1:01:34Douglas Tuman: Oh wow. And is it because you just want to verify that make sure? you're verifying the things yourself? Or is it

1:01:41Muhammed Esgin: Yeah, that's so I want to verify its output first before I use it. And also I guess uh I don't want to give it too much power in a sense or too much access that it can just do things without me monitoring what it's doing.

1:01:57Douglas Tuman: Yeah.

1:01:57Muhammed Esgin: Uh yeah. So this definitely has a trait trade of it slows things down because I need to manually actually check what its output, maybe I don't know, copy it somewhere and then maybe modify it and then use it and so on. But yeah, that's I guess the threat of I'm making at the moment.

1:02:11Douglas Tuman: And kinda to like the earlier question I was asking, I mean, uh, you know, like where you see encryption ultimately going. With AI, you know, where it is and where it's where it's headed, um, do you you know I I feel like encryption is going to be vital. It's it's already vital, but even more vital in terms of protecting humanity, protecting our liberty in the digital age, as I like to say, obviously with things like digital cash, but even just protecting your privacy online and preventing AIs from knowing everything about you. Um, but do you are you hopeful that we'll have tech that will keep the AIs at bay and you know the surveillance states that might be using that tech against people? Um

1:02:58Muhammed Esgin: Yeah. Yeah, I think even kind of going beyond just encryption to like cryptography more broadly, I think cryptography will be more and more important moving forward to really protect uh people's sensitive information.

1:03:13Unidentified speaker (UNKNOWN): Yeah,

1:03:13Douglas Tuman: communication.

1:03:14Unidentified speaker (UNKNOWN): Because AI, as

1:03:15Muhammed Esgin: you mentioned, has made things a lot because it can just you can say, okay, just go and try to do this for the next, I don't know, 24 hours and it will try this tirelessly over and over again and until it can I don't satisfactorily do something. And maybe you can I don't know say okay I don't know create like this 10 ten or 100 a hundred sub agents and try to do this. And it's basically kind of like something that keeps trying over and over again while we as humans get tired after a while and we need to rest and then we don't have that much power to actually try things at the long in the long term. And I think cryptography will be very powerful tool to actually prevent AI from really crawling around uh mass amounts of data to infer something because it will just be hit by a kind of like I don't know a impenetrable wall that will that be that will be stopping the AI from I don't know crawling all of that information and making meaningful inferences out of that information. And yeah, especially these like privacy enhancing technologies in particular, like zero knowledge

1:04:22proofs or I don't know other kinds of fancy cryptography, I think will be very powerful. in that respect because I think ultimately the goal with the privacy enhancing tech is to minimize the private information exposure. And when you do that, it will get a lot more difficult for hackers, AI, or any other kind of like tools to actually get access to your private data. And I think that will make things a lot better for humanity to actually make sure that we. are kind of like keeping things that you want to keep sensitive or that we want to keep private uh to us only and not kind of give them away very easily.

1:05:00Douglas Tuman: Absolutely. Well, thank you, Matt. Thank you for all the work you're doing, the contributions you're making. Uh I'm

1:05:07Unidentified speaker (UNKNOWN): glad we

1:05:07Douglas Tuman: finally discovered you. I guess somebody had mentioned, and I it this does ring a bell actually. The Monero Research Lab cited uh um your your ring CT three point oh in their a trip tech paper, but we didn't end up moving in the trip tech direction. We ended up using full chain membership proofs. So

1:05:25Unidentified speaker (UNKNOWN): his

1:05:25Douglas Tuman: work is known to the Monero devs. And yeah, I actually I I that kind of does ring a bell. Um it was the the quantum resistant ring signatures, which is amazing.

1:05:35Unidentified speaker (UNKNOWN): Yeah, so ring C three

1:05:37Muhammed Esgin: point zero itself is not quantum resistant. but yeah, it's kind of like a uh

1:05:41Unidentified speaker (UNKNOWN): Oh

1:05:41Douglas Tuman: okay. So

1:05:42Muhammed Esgin: it's elliptic curves still yeah.

1:05:44Douglas Tuman: Okay. Okay. Uh well yeah man thank you so much and I will try to get you to come to Monerotopia. I'll I'll let you know about it. Maybe we can convince

1:05:52Unidentified speaker (UNKNOWN): you.

1:05:53Douglas Tuman: But I I guess you're coming from pretty far away you're in Australia. So we we do it in Mexico City. We do it in Mexico City.

1:05:59Unidentified speaker (UNKNOWN): At least

1:05:59Douglas Tuman: that's where we have been doing it.

1:06:01Muhammed Esgin: Uh Uh-huh. huh okay, okay. So Circula

1:06:03Douglas Tuman: far for you.

1:06:03Muhammed Esgin: Yeah, it's uh it's a bit of a distance, but yeah.

1:06:08Douglas Tuman: It would be cool, man. You'd see you'd see uh, you know what cryptography has, you know, uh the the real people that it that it serves. The Monero users.

1:06:19Muhammed Esgin: Yeah, yeah.

1:06:19Douglas Tuman: Um it's a fun bunch of people too.

1:06:22Muhammed Esgin: Yep.

1:06:23Douglas Tuman: Um but yeah, but thank thanks for everything. Anything you want to put out there, any information? you want to put out there? Any uh

1:06:29Unidentified speaker (UNKNOWN): um

1:06:30Douglas Tuman: thing you want to direct people's attention to?

1:06:32Muhammed Esgin: Yeah I mean I guess if people are interested uh they can I have a personal website that I personal meaning like academic I should say. I don't really share personal information but more on the academic side. They can go look them up uh

1:06:48Unidentified speaker (UNKNOWN): like

1:06:49Muhammed Esgin: on the the publication that I work on or the research that I work on.

1:06:52Douglas Tuman: Yeah, for

1:06:53Unidentified speaker (UNKNOWN): sure.

1:06:53Douglas Tuman: What is it?

1:06:53Muhammed Esgin: I have seen after after you actually I after you kind of advertise this uh talk, I have seen people following me on uh ex Twitter.

1:07:04Douglas Tuman: Yeah.

1:07:05Muhammed Esgin: I'm not very active since I don't know, for some time now. No, If people want to connect, yeah, full please feel free to connect on LinkedIn. I think that's the uh space I'm online space that I'm most active, even though not still super active, but yeah, if people want to connect, I think that's the best way to do it.

1:07:22Douglas Tuman: Okay. Awesome. And what's the website? What's the personal website?

1:07:26Muhammed Esgin: Uh so it's do you

1:07:27Unidentified speaker (UNKNOWN): want me

1:07:28Douglas Tuman: to type

1:07:28Muhammed Esgin: it in the chat or something? Or

1:07:30Douglas Tuman: yeah, you could yeah. Or I could try to bring it out. Oh, it's like a long URL.

1:07:38Muhammed Esgin: Yep.

1:07:38Unidentified speaker (UNKNOWN): So

1:07:39Muhammed Esgin: this is the one.

1:07:40Douglas Tuman: Okay. Here I'll put

1:07:41Unidentified speaker (UNKNOWN): it. My

1:07:41Muhammed Esgin: first name, middle name, and then last so initials of my first name, middle name, and last name.

1:07:47Douglas Tuman: I'll bring it up here. There it is. There it is. Yep. So it's m-f-in.github.io.

1:07:56Muhammed Esgin: Yep.

1:07:57Douglas Tuman: Anyone Anybody wants to learn more about him, see what he's working on next in cryptography.

1:08:01Muhammed Esgin: I I should should now update my website. actually. I haven't updated it forever.

1:08:05Douglas Tuman: Get on that, bro.

1:08:08Unidentified speaker (UNKNOWN): Yeah.

1:08:09Douglas Tuman: Um but yeah, man, thank you. Thank you so much for doing this. Thank you for everything you're doing in the field. And we'll be in touch. Good luck, man.

1:08:18Muhammed Esgin: Yep, thank you very much for having me and all the best. Thanks.

1:08:21Douglas Tuman: All Alright. right. Adios.

1:08:22Muhammed Esgin: Cheers.

1:08:23Unidentified speaker (UNKNOWN): Hi

1:08:23Unidentified speaker (speaker_1): Monero Land. Thank you for joining us on this week's episode.

1:08:27Unidentified speaker (UNKNOWN): We release new episodes every week. You can find and subscribe to our show on YouTube, Odyssey, iTunes, Spotify, or wherever you listen to podcasts. Go to MoneroTalk.live for a full list of places where you can watch and listen. If you want to interact with us, guests, or other podcast listeners, you can follow us on Twitter, Mastodon, or any of our social media platforms. Monero MoneroTalk Talk is also made possible from contributions by viewers and listeners

1:08:54Unidentified speaker (speaker_1): like you. And supporting us is easier than ever. By typing in MoneroTalk.crypto in your Monero.com or CakeWallet send address field to send us a tip. Once again, thanks so much for listening, and we look forward to being back next week.

This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights