Hyperledger Foundation Meetups, 23 May 2023

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Enhancing Privacy & Security with ZKP & Encryption in Hyperledger Fabric

Description

A meetup hosted by Hyperledger Toronto on May 23, 2023 featuring Alex Shpurov, founder of Artifirm. The event explores the potential of blockchain technology in advancing data privacy within Hyperledger Fabric. It will highlight the role of cutting-edge cryptographic techniques, like Zero Knowledge Proof and Homomorphic Encryption, in achieving data privacy on public Hyperledger channels. The event will also showcase various potential applications of encrypted AI and blockchain technology in fields like fraud detection, supply chain management, healthcare, and financial services. The integration of AI algorithms with encrypted data ensures data privacy while blockchain provides a secure and transparent platform for data sharing and storage.

More about Hyperledger Fabric can be found at: https://github.com/hyperledger/fabric

----
Hyperledger – Open Source Blockchain Technologies

Hyperledger is an open source community focused on developing a suite of stable frameworks, tools and libraries for enterprise-grade blockchain deployments.
It serves as a neutral home for various distributed ledger frameworks including Hyperledger Besu, Fabric, Sawtooth, Indy, as well as tools like Hyperledger Avalon, Cactus and libraries like Hyperledger Aries, Ursa. Learn more about Hyperledger projects: https://www.hyperledger.org/use

Discord: https://discord.gg/rVCn4j7UFd
Case Studies: https://www.hyperledger.org/learn/case-studies
Training & Certification: https://www.hyperledger.org/learn/training
Tutorials: https://www.hyperledger.org/use/tutorials
Webinars: https://www.hyperledger.org/learn/webinars
Events: https://www.hyperledger.org/events
Vendor Directory: https://www.hyperledger.org/use/vendor-directory

Subscribe to the Hyperledger Newsletter: https://www.hyperledger.org/newsletter

Follow-us on Twitter @Hyperledger

Learn about Hyperledger Membership: https://www.hyperledger.org/about/join

#Hyperledger #Blockchain
License
Creative Commons Attribution license (reuse allowed)

A

My experience is.

B

Technical and yeah today, like things on session I'll share what they'll take by what kind of lessons we learn from the hyper manager. What kind of like you know what we can use as a based society and religiously developed like for a while and like you know, this presentation is not related to the technology in general, but it's very small aspect of it, like you know, regarding like encrypted computation and.

A

Regarding zero knowledge, proof, it's basically regarding a data privacy.

B

A

Not necessarily.

B

Only, for example,.

A

Supplication so let's start, let's see.

C

Yeah, let's go.

B

Sorry so read, the presentation was divide, like a section is like we're going to discuss. Actually, like you know, three main aspects, so the full homomorphic encryption and the common market description in general. That's an encryption that allows you to encrypt data first and then run they run the computation on that already encrypted data. So meaning, like you know they would allow to, like you know, execute like you know, various computational say machine learning.

B

So when you have data belong into one party and, like you know, the model coefficient of the models belonging to another party and they don't want to share anything so.

B

Yeah, so basically the the idea here is just like you know: that's a technology that would allow you, like you know, without uh compromise, in any privacy, to share the data. So.

B

And like very easy.

A

B

Encryption, like you would say, like you know, you guessed the number and like an I, guess number and, like you know in Fitness number, like you know independently on our size and, like you know, we can add those numbers together and.

A

When we decrypt it into the piracy, so we actually see like a final sign.

B

uh So, for, like obviously like a lot of course, that's quite easy to predict. But if you have like, let's say three participants that.

C

You already have a privacy like you know, I mean like you know, you know Vietnam. If.

B

You don't know nothing. Okay and.

B

And like an homographic.

A

Description, like you know, it's a really emergent.

B

Market right now and we're quite a bit of work done and, like you know, obviously this is very interesting to it's: a technology to look.

A

To privately look in someone else's data.

B

Without compromising security, so some work, like you know, we did like related a fraud detection like multiple Banks. Okay, like basically let's say a TD web, like you, know things this transaction. Maybe this credit card transaction to the program but I'm not quite sure, like I.

A

Would like to get a feedback from like another apartment of business,.

B

Let's say RBC and basically and just like you know, can privately encrypt the data and, just like you know, ask mono to give the final result.

B

A

They're, basically like.

B

You know for the foreign.

B

And it's not shared so and like in a fraud detection mechanism like yeah, let's say a resistive transaction from the client and then the benefit that this transaction may be problem and I mean through transaction and send it to event B and eventually encrypts the model, the model they have and like you know, and then apply this amount of to encrypted data. So the result that you know no one has seen anyone's data and at the end you can see a final result and the bank a can only see yes and no communion transactions.

A

So basically that kind of weird you should let you know the.

C

Visibility over it: okay.

B

So same thing, like you know, you can apply for, like you know, the credit chart uh so like I, don't want to go into the test, but that it is similar because, like you know, when you're doing like an upgraded track, uh so you have to walk this organization. This is.

C

Like, let's say like the.

A

Rogers you just you purchase them a new service. So, like you know, you've got a purchases have been planned and.

B

The part of it, like you know you have to provide like a great history and, like you know, we do have all this so the mechanism for this Easter, but it could be independent, like you know it can be independently. Ask like another Banks and they can, like you, know, look into the transaction. History and Emerson was encrypted, like in all the databases, and since the data is encrypted like you can simply calculate about it and then gives you a probability, the right side that you cannot these you pray. The scores are not basically.

A

To approve yourself so.

B

Basically, whatever the multi-party kind of like multiple parties like we always can calculate, we homes can improve the.

A

Data first and then ask it about to do any purpose: navigation.

B

um So could you generalize this idea we can like, in our conditions, do not compromising any privacy we may negotiated on the final result and like.

A

In a final result is just a single number.

B

So they're, using of uh how much encryption course organization, we would actually emerge like a new market like AI.

C

Markets, AI, Market I think where.

B

Like people can people can, like you, know, trade, their AI models without showing them because they will be on second for the data and, like you know, people can test those models like by putting like you know a data encrypted for this models and, like you know, look there.

B

The results because, like you, increase the data and- and so this like an old leverage, an opportunity to like you know, create like a great New Market, like you know, and technology is like a blockchain and related that, in particular, would be here for information exchange of Player information and change that will spend problems, meaning we're not cheating. Anyone as well as the constraint system, so I mean like, for instance, I'm looking for.

A

You coefficient.

B

C

Right from like, let's say, 70 percent.

B

B

In order for like, in order to to to go from homologic encryption into the AI market, like Model Market, you actually need to.

C

Have an ability between people and also you.

B

Need to find the constraint system so meaning, like you know, the system should follow certain rules. uh That's what we call a smart conference so, like you know, for blockchains and since ever since the cricket here, so the smart contracts needs to be smart enough.

A

B

Deal with that encrypted data, and for that we actually like in our reaction, which is another Innovation.

A

B

Knowledge proof and I'd be looking to the next slide.

B

B

Like two examples, uh two use cases like you know uh where we can like apply and like you know how to get money out in the united markets and now they actually look into like algorithms.

B

So uh in a fully common mode, reconstruction. uh Basically everything is encrypted. But, like you know, if you want to make it practical Union like and we should be able to search so meaning like enough, for instance like if we have like no number, let's say five: we encrypt it like you know, and so then we had another number. Let's say five again, I mean like and I converted.

A

And then we subtract and 55 we.

B

B

We have zero. So uh when we decrypted like you know, we know basically it's uh it's basically zero, meaning the numbers are equal, but let's say we have like. You know one yeah, that uh the the numbers are not equal. So basically in in this case, like you know, if we actually like in a group.

A

B

Extend this technology for, like you, know, string search for instance, then we, like you know, need to compare let's say letter by letter or word by work. It depends how you like. Basically, you can encrypt like another string by numbers by Letters by individual letters or characters, or you can take like the back of words. Let's say you have a couple of like experience and you increase, like you know, one two three out of ten like and then basically you, like, you know, start subtracting and start comparing them uh and Minion.

B

Like you know you, basically you have to compare every single element in it like you know, to get a final result so and, as you can see, the complexity of this operation is a it's a linear complexity, meaning.

B

So it's quite quite it's like you know when you have like longer text. It's just like you know uh quite costly, so what we can do about it. So in the computer science, uh everything is graph uh all trees, like you know most graph and, like you know when we think about, like you know, optimizing like you know any sort of like a computer science problem. We are actually like. You know the first thing we think about.

B

We think about the graph and in our particular case, like you know, we actually can use a prefix tree. That's we just call. It tries I'm, sorry for misspelled.

B

So but like you know essentially, like you know, you see it like on the left side, you have, like you know like rewards, but they all represent it as a tree and, like you know, we say all overlapping uh like an elements in it and, like you know, this number of leaves becomes to be much less in here and of course, we see like you know when we dealing with the trees complexity, like algorithm complexity, comes from o n into in into the into the the logarithm complexity.

B

So in this particular case, like you know, it's going to be like a look at um and that's one way like you know how to speed up a computation data and you can make it next.

B

A

One, like you know, it's.

B

Another way of how you can speed up, like you, know, computation over encrypted data, because, to recap, like you know, when data is encrypted, you have no idea what's inside, but you still need to use like you know some form of data structures to look into it, uh because uh this is like an algorithm.

B

It's actually used on firewalls in a network of article room filters, so the blue filters, like a basically it tells you if uh your number or like string or like some kind of data element is within, uh is basically exist or not meaning it's more like a probabilistic structure.

B

uh But, like you know, if the element is not existing in in a set in the in this kind of like a bloom filter string, you always will be negative if the element, if the and.

A

B

Order to to compute, like you know, this filter is just basically we just like you know, simply the hash, the uh the the harsh, the let's say, string, call like you know some kind of data element and keep adding Now using X or like binary reparation, and the blue filters uh becomes to be quite big but, like you know, but they represent. Like you know, data compression.

B

So you can see the data structures here is again three meaning, like you know, we're going into a logarithm complexity, meaning, like you know, we decreasing like in a computation speed like a lot and like exponentially next.

B

um Yeah, so this one is uh uh kind of like um when we're dealing with uh uh voice data or, like you know, kind of vision, data or, like you know, any sort of like wave data, so we actually dealing with slaves and like uh and like speech, for instance, processing like a speech detection like a voice detection like a mechanism and basically we're dealing with the signal processing and in the signal processing is uh essentially like you know.

B

Any signal like you know whatever like, for instance, a vision with a computer vision or just like in a human region or a year. So it's kind of like you know could be there could be a transfer could like any signal could be transformed into the sun of sine and cosine that, like like it would be a number of sinuses and cassinuses and and like you know, when they add them up, like you know, we'll get like an hour.

B

uh Our like you know initial signal, so we're doing it by actually uh by applying the Fourier transformer for it. But in keep in mind it's just like you know, any signal is nothing more than the sum of silence or side and the full homomorphic works Works.

B

Only with a multiplication and addition and like an all, the signal processing becomes to be a very good fit uh like you know, in homomorphic encryption because, actually, like you know you can before encrypting it, you can present it as a frequency domain and then, like you, know and then apply all sort of computation.

B

Okay, let's can we move forward.

B

uh So, like you know, when you're dealing with the language models, okay, any.

C

Language models.

B

The language models they like charge up, it is like a very popular like you know, everyone get excited about it, but the idea about any language models most of language models is to use something called engrams. So meaning, like you know,.

C

Someone else's questions briefly, they asked if you, um if you could explain just a little bit more about that about what someone.

C

Sort of sending certain previous foreign.

B

Like you know, any sort of like any sort of signal, whatever is just lower the signal, like any sort of signal into the sum of sine and cosines and, like you know, for instance, like you know when we're dealing with like you know the audio filters right. So the idea is just like you know to do a cut off of certain frequency, so meaning, like you know, when we actually transforming everything from a Time domain.

B

Like you know the signal into the frequency domain so in the frequency domain we'll have like you, know the magnitude of the frequencies and basically, but they will add up in the main signal, so what we can do, we can subtract them and then and we can subtract them in the homomorphic space. So meaning, like you know all this kind of sort of, like you know filtering like you know, we can apply in a different domain in a different domain in the frequency domain.

B

So, like you know, we can do like you know, let's say like in a filtration in a in a in a like, remove like a certain frequency or we can analyze voices because analyzing the voice is also kind of like you know, frequency related thing.

B

It's called Capital transform, but it's just one more level of Fourier transfer, but essentially it's decomposing like a signal in the sum of sine plus signs and like you know, and the the source of that uh like what you encrypt, you actually encrypt, not a Time domain signal, but you encrypt, like you, know the uh like those kind of like sand and cosines frequencies once it's encrypted, like you know, you can do whatever you want.

B

You can just calculate because keep in mind, it's just like you know signal is nothing more than just a sum of like you know, harmonics.

B

Is it uh I know it's just like I was there is quite a bit here to cover so I tried to explain as much as I said, but, like you know, the idea here is here: it's basically it's one of the it's one of the use like you know you actually can apply like you know this kind of hamburger encryption, not only in like you know so looking something in the database but, like you know, create a homomorphic encryption filter for like audio processing for like uh visual processing later we'll cover Quantum Computing, as well, so going back into the language models.

B

So in the language models like in all the language models comes into uh something called engrams engram like in this one. It's uh basically a language model is trying to predict the next World. The next word based on the previous work, and it's called one gram. So two grams, meaning like you know, based on two words we predicted like you, know the next word and that's basically, that's how GPT works. That's how modern black neural network uh work and ends like basically and uh what we do.

B

Basically, we just like you know, create, like you know, the neural network and the neural network is a symmetrical neural network. With a bottleneck in between and in the input would be, let's say two words: an output would be like a the next word, we're trying to predict and, like you know, and the neural network actually like you know squeezed into like you know something called like uh in the middle middle middle there, because there are many layers here and this middle layer, it's called like Laden in space.

B

It's a bottleneck, is it like you know, and also called embeddings, and the neural network is all about the set of coefficients and Minion like this kind of like latent in space. It's just nothing more, there's just a vector and, like you know when and charge GPT gives you like latent space too, uh like you know, as many of like you know, other language models, like you know at least using like neural networks.

B

So since it's a vector, we actually like you, know, start dealing with uh with um a natural language processing as a vectors, for instance like we can you can spot that men and women I sit in very close in a left upper corner and, like you know, kin and queen, they sit in very close too, because they actually have very, very similar in meaning uh and uh like it's, it's basically like the one of the most interesting thing about uh those vectors that, like you know, you can basically, you can add vectors and you can subtract vectors.

B

You can do like an operation on the vectors and meaning, like you know, you can do like, let's say um like, for instance, man plus man, plus woman, minor skin, it will be equal to cream.

B

So if you just like, imagine those kind of vector computation so and that's about that's by the way, it's a great uh Power of a charge repeat that I don't see much information like you know, people are talking like lots of charge GPT in form of, uh like you know, language models like generation, some advice generation with some kind of like you know, text or give me the advice, but the really really the true value of uh charge. Gpt is in this kind of embedding space. You also can through apis accessible.

B

So you actually can, like you know, do the same thing and, like you know, you actually like you can see like you know how many things are related one to each other and, like you know, on top of this one, you actually can do like uh some kind of new predictions. Like you know, by calculating those kind of vectors, the one they just sent like men plus woman minor scheme will be equal to Queen, and you know that's a great part of like a chat GPT.

B

So the reason this slide is here: all this stuff could be done in uh encrypted space. In a homomorphic encryption and the computation I just mentioned could be done in encrypted space as well. They're gonna, basically before doing anything you just encrypt, and then you calculate so it would be encrypted man plus encrypted woman minus encrypted kin.

B

It would be equal to encrypted 2m, so I know it's just like you know here is very a lot to cover, because it's a machine, it's a it's, a deep learning and, like you know, tragedy, pity and like those kind of language models, but the general idea, like you know, using like uh we, we still kind of get excited with shot GPT. But, like you know only here, like you know, very few sounds like you know about the data privacy yeah, we'll we already have data.

B

We already have a good models for like language models, but, like you know, we don't have an ability to share them confidentially.

B

You can move forward. Please.

B

uh So um this is a quantum blockchain safety safety and before we get into the quantum problems, so, like you know, in a Quantum things like you know, let's uh like for some reason, like you know, all the explanation about quote and computer is like you know, really tough. It's like like when I read into it. It's just like you're, not supposed to be that difficult so because a quantum computer was developed like like 50 years ago, even more, and the idea is like very simple.

B

So we have like you know we have a wave, so we have a wave and not one wave. We have a multiple waves. All we do like you know. We just shift the wave a little bit here and there we call phase shift and we just add them and if the way, if the shift like you know, let's say here is like you know, on a lower and a high level. So then, like you know, we have uh foreign effect.

B

So, if that, basically, if there's a in in Northern phase, if they actually can opposite phases, then, like you know, you have a cancellation effect because you you can see, those harmonics would actually like you know, would be canceled by the asset harmonics because they just shifted in the same time, if the signal is identical and you just simply add them together, so there would be amplification effect and whatever the quantum computer is doing is doing nothing else.

B

Just these two operations so get the harmonics and shift it and add it, and that's it that's all. That's the only thing is quantum computer here for so the physics behind it. Like you know, the implementation is very difficult, but the idea is, like you know, very simple: it's just based on the Fourier transform I mentioned before it's just like you know really sum of harmonics and some of her harmonics that are shifted, uh and if you can move next.

B

And, like you know- and here we actually talking about like you know, one of the algorithmic sure algorithm, so this algorithm is actually promising algorithm that breaks elliptic, curves and that's supposed to break the hyper laser privacy and ethereum. How, like you know, breaks Bitcoin and, like an all other stuff. So the idea behind, like you know this mechanism is very simple.

B

So we there are, like you know, um the ham, the elliptic curve, the encryption behind uh the let's say ethereum of Bitcoin is just like you know. Simply like you know, you have two prime numbers and one prime numbers acts as a private key another one uh acts as a like in a message, and basically we just do an exponential operation. Like you know and and just to revert it back, is very difficult.

B

Like you know, it's it's, it's not an easy problem, but, like you know, it's called kind of discrete logarithm problem so and, like you know how to break it. You're actually, like you know the Schnur, the person like the the scientist. They came out with an idea. So basically, you have to you can present uh those kind of schemas uh by through this, like you know, number of mathematical uh formulas or Transformations.

B

You present this problem as finding a period within the signal and, like you know, and as I mentioned before, finding a period it's just simply like you know, finding an oscillations and that's what essentially like you know, basically the the the the idea like you know, once we have it in that form, you see on the left side we just need to like you know, we need to do a phase shift and, like you know, we need to measure uh basically we we need to amplify those things we are looking for and we actually need to cancel like you know the other stuff we are actually not looking for and that's it like you know, once you have this kind of like period, you know from this period, you actually can say, like you know what what what?

B

What? What is my private key. So this kind of idea here is like again very simple but to get into it. It's just like you know about 10 pages. In a in a paper um yeah. We can move next.

B

So, like you know, it's kind of a summary: what makes a data privacy important in hyper laser fabric, so hyper religious fabric is uh not a single blockchain. Indeed, is a network of blockchain, it's actually the most advanced blockchain like you know, I've ever been developed because it doesn't represent the single blockchain and it actually, it's a blockchain or blockchains, and the privacy in in the in in hybrid religious fabric is implemented through that, like membership, service and membership service, meaning like and all this like.

B

You just Define like through ACL Access Control list, who is actually like you know.

B

Participant of this networker is not participant this network and it partially, like you know using, like you know the permission and the partially using like you know: cryptographical membership, minion CA certificate authorities, and it really becomes messy, like you know, when we actually need to exchange information for the parties that not a part of that that that's not included in this current uh uh current Channel membership and for this kind of, like you, know, people the hyper laser fiber, for instance, is offering something called side.

B

Channel Side databases, like you, know, private databases and it's really kind of like messy how to set it up and it's like you know, it increases complexity of solution.

B

A lot it's supposed to Simply, like you know, just get the homomorphic encryption or zero knowledge proof the one we actually will talk a little bit later and, like you know, and with that you actually can use any blockchain, you can use ethereum, like you know, unless it's a Quantum safe, like you know, you can use any form of a blockchain you and the the the the the the the the the the the the cost of development becomes to be much less because, like you know, you have to invest into the initial cryptography, but then the mean of communication doesn't matter so zero knowledge proof is the one we're gonna be talking in a in a bit like you know, it's slightly different from homomorphic encryption, it's more like a constrained system, as opposed to be a computational system as well.

B

Homomorphic encryption release so and uh AI is and blockchain like you know, really, like you know: zero knowledge, proof and homomorphic encryption in leverage you and AI market and AI Market is like uh is some place like as I mentioned before, where you can actually say- and here is my model- and here is my constraint and, like you know, on that data, it gives certain results uh yeah and the last one quantum computer on blockchain safe. He will talk about in a bit as well gonna go next. Please.

B

So um here is a slide about quantum computer again.

B

um We kind of like all scared that, like you know that once quantum computer is developed, all our Bitcoin wallet becomes empty.

B

It's um not necessarily true, like you know, because, like you know, you can protect this one in a in numerous ways. Like you know. uh First of all, like you know, keep in mind that, like you know, blockchain sorry, quantum computer has like you know, really it's a linear system. The quantum computer can only add and like do nothing else like you know, substitution becomes to be as, like you know, like phase shift and eventually it's the same as like an addition and like another robot.

B

If you just check Wikipedia like you know, there are about only 16, like known algorithms are known for quantum computer. It's like like the message here, a quantum computer. When it comes out it will be like a video card. It's something extra something nice to have. It won't ever replace it like.

B

You know the conventional computation things uh so, basically, just like you know something that, like you know, doing like an Australian algorithms and again, this number of algorithms are only about 16 and, like you know, they're very limited and the old winner, so like most of the solution is not linear, like even like you know, AI is linear, like most of the eye. Stuff is not linear.

B

So and, like you know, for a blockchains like you know, basically, once blockchain has a smart contract, you can Implement very simple things.

B

We can actually, like you know, submit, like you know, put in a smart contract, some hash of the random number you're going to hide and when you're gonna spend your money out from your blockchain, you have to provide this random number and, like you know, that's the noon by you, but at this point, like the smart contract could verify that, like you know that see, the committed value is equal to this hash and that's it and that's it.

B

That's a fix is known to be, like you know, post Quantum thing and, like you know this, this kind of little bit tricks here and there, like you know we still live in a more or less like you know, say world uh even like you know, when the quantum Quantum algorithm becomes practical. Okay, can we go next.

B

uh Yeah, this is like everything and I encourage you, I have like enough. You Publications, on a medium uh talking about different things.

A

But oh, it's like mine,.

B

So, like you know, I covered zero knowledge. Proof like very briefly uh but like zero knowledge, proof is uh uh kind of like you know what we have, uh but basically it's a private constraint system and, like you know, you can think about, like you know: zero Channel proof of homomorphic encryption, but with homomorphic Egyptian you can do a computation versus zero knowledge.

B

Proof like you know that what really gives you, uh like, you know, a constrained system where you can verify, like you know certain constraint, based on an existing data uh and, like you know, if you need more uh like you know, I have an article about it like you know, Hands-On of uh proof of ownership. So an example here is just like you know. We have a mutual fund and this mutual fund, like you know, has a list of uh assets uh and a list of assets are known.

B

What is not known uh that allocation for these Assets in a mutual fund itself and, like you know, you have a like in a practical example like you know in Python uh to to basically to see like you know how this stuff works, and there are a few methods. I'm not going to cover uh them so, like you know, and then I have like you know a couple of articles on, like you know, on the basics of encrypted computations.

B

So the main thing of unencrypted computations are like a computer science problem like graph Theory I would say what was presented because by itself it's like sounds very easy, like an encrypt calculator.

B

So but like you know to make it practical, because it's not fast, it's slow, because it's just dealing with just like you know certain type of cryptographical like this cryptography so and you have to like you know, move it from from the linear space into like a logarithmic space to make it faster and yeah.

B

Like you know, I have a quite a bit of I spent like you know many years of work trying to get like and all that blockchain like in different uh of blockchains working for like they're, not Enterprise and, like you know, we have a couple of successful projects. I put my kind of, like you know, as a developer. I put my kind of license in this running like lessons and, like you know, failures like mistakes.

B

I've made what basically, what not to do in this article like running the blockchain in production, what it takes yeah and I think, like you know, we can go into the the the the questions this.

A

A

Questions if there are any.

A

We can just please unmute yourself and then ask a question.

A

And in the room, I have a question.

C

Yeah, how can you build this on top of medical Legend s.

B

It's a really communication protocol right. So basically, uh basically it could be built as a as a part of the smart contract so because it's just a library right, so you just compiled it in you.

B

Just like you know, pass public key and we'll do encryption and basically and uh in computation and like you know, when you get data out, you just use your private key to decrypt it so in in other way, like you know you, basically you can just like in a publisher row data like you know, on a blockchain and get cloud like you know some form of API to do a computation, because once data like data ownership, is important.

B

So basically, once you put in a hyper Ledger you, the data ownership is guaranteed, uh meaning like it's a double spend problem. Obviously, but then, like you know, you add the cloud or like an any other system, like you know, can take data from like a from the proof source and do a computation, and let's say you can submit the data back.

B

It's like it's like the same as dealing with the raw data.

A

A

So uh in a real life example, like you know, something um I have been in blockchain in in quite a few years ago, so uh to understand how the security works correctly in the blockchain systems in a real life example. Can you give me something yeah, a work done or something you have seen where a security breaches happened or where uh um the encryption that you spoke of must be equal to you. Some some kind of such situations.

B

um So um like dealing with identity, it's like one of the case. So, like uh my meaning, like you know, you have a social insurance, number Social, Security number. Okay, so, and you look at you see like when you think about the driver license like a driver license case, you picture that could be a homomorphically transformed through that signal processing the one I mentioned, and it has an attributes like you know, validity date and blah blah blah.

B

So basically you, the the the the the the representation of like a driver license, is really a graph of data, and once it's a graph like you know, you can do computation on it once you can do computation it. Basically, once it's a graph now you can make an encrypted graph. Now the government, let's say, are like you know. The owner of this license can, like you know, encrypt this data, and now actually you can do like in a certain computations, for instance like check the you go to a bar.

B

Like you know, you don't need to ask 18 years old, you just have QR code and this QR code represent the thing and you just like you know, do a computation and gives you yes and no same with just like, let's say last name and same even with your picture, because there are like an algorithm of black people, like you know, face similarity, you can do thermomorphically too, but the beauty of it like when you start thinking like you know, when you're dealing with a encrypted data you're.

B

Actually, like you know, you not use a real data, so you have to come up with a data structure so and then, basically you have to come up how you're going to compare, because at this point like you know, your input is encrypted and your output is encrypted and then that's the data you send this encrypted to so.

A

That's that's uh one more follow-up question about that. um So uh so let's say we have two sets of I mean there's the source, encryption and there's the bonus uh result of encryption.

A

um So this the the decryption or or what you call as uh the check of uh the data in a validation.

B

A

Should be some kind of a key structure right I am supposing there's a public key private Keys kind of uh yeah.

B

Yeah yeah in common morphic encryption, there is a public key. The one yeah.

A

How I mean in your opinion, how would you would you say how, how secure uh that that process would be because consideringly non blockchain or the non.

B

A

Know the the the traditional systems that we can be 10 years ago: okay,.

B

A

Comparing to what we are having right now, like you, said, preventing something like this right. What would you say how secure I mean like if you give a scale of security? What would you say so.

B

You basically you the insecurity, you never trust anyone yeah. So don't trust me. So when you have any security concern, so there is a committee called nist, and this committee has exactly what you uh that answers like you know.

B

They have like a metrics for like different level of security, so main concern right now for what blockchain is a Quantum privacy, which I kind of like would not agree 100 on this kind of threat, but there is a threat for sure, but there is like you know, different ways how to solve this problem as well, uh but in the same sense, like you know, everything whatever you do has to comply and like comply with what complex standards and and the standards are published. Okay.

A

B

Like you know, they have like latest cryptography is not yet niece certified, not yet, but it's very, very close so and the committee I guess they're, like you know, I think like every six months and like you know there is a certification can made it, but this year, like from the business opportunity to think about like stock option right.

B

So when, like you know, when you say, like you know, two years from the maturity date, the stock option costs very little because, like you know, people actually don't see quite a bit of business because it's like loss of risk and like you know, but when uh stuff gets mature right, like you know, you have two off this, you have to hold it or you have to give it to someone else and, like you know, with this one, like you know, there is a market cap for everything and granted like you know, uh it would be really hard to execute, like you know, kind of like the fraud detection in a band in a way because, like you know the cyber security people, like the security people, would ask, show me the certification.

B

Oh, like yeah, it will be like tomorrow, but do come tomorrow, so the latest cryptography is basically but in the same time it's uh it's a kind of not many people doing it. Many times like you know, we hear just like you know to explore. Like you know, what could our life be in the future and how we can make a profit about uh how we can generate the profit out of future, because the future is the most expensive asset right?

B

C

Focus more personal Alliance is that many of these problems can at least be clear and will more efficiently with trust executed environments if I even know yeah.

B

Yeah there is a to like basically Intel puts a lot of effort on it.

C

B

C

And you have searches, the question is: what would you consider the reason.

C

B

Okay, basically, like there's a solution like an hyper laser fabric, it's hyper laser fabric. It's based on, uh like you know, the Asian is your knowledge proof as a security add-on. This is like an accord which is another flavor of a blockchain. They actually use like an sjx technology to do encrypted. Computations and like intellijex technology is like really a kind of a separate, isolated processor with just like you know the master key baked in it.

B

It's not extractable and it would allow to do any form of computation inside, uh but this year this one, like you know, yeah tend to have like a manual on their abilities in the past. I was like you know, so. I was so happy like you know, when I saw it first time uh because, like technically, you can apply uh the any sort of homomorphic encryption like uh it's a kind of very slow because, first of all, it's like you know how the the latest cryptography is the Dylan's matrices.

B

So that's a form of homomorphic encryption like just like you know like exponential. One number is another number, so the sgx gives you like. You know everything at the speed of, like you know more than CPUs, but like with this kind of so many leaks, uh so many hacking and, like you know so many stuff like that happen with the uh this kind of, and basically once it's compromised, you there's no patch.

B

You have to throw the the hardware away guess like and with this kind of thing at least you can re-encrypt, you can do something, but if you master is compromise for this processor, that's it like. You know you have to invest in like a new processor and your motherboard and like um kind of we'll see, but so far it's kind of to me.

B

It's not reputable yet and like, and the the hacks going, not from that Mark that behind it, the hugs going from, like you know, inner process communication from kind of some circuits like vulnerabilities from kind of some side effects not from the elliptical mathematics, not from the encryption they come from from like in a manufacturing process and I, would rather stay on academic side as uh it's just like in US mixing like you know, putting too much technology in it and and like you know, once you have this kind of becomes to be like iot problem.

B

So this kind of bulk, like you know, gets hacked. Everyone gets hacked through this bulb and the only way the fix is just like take it out and throw it away.

B

That's my view: I've done I mean, like at least doesn't pocs on it. It was a really good thing like to use in the azure, because that would makes everything like an encrypted, but, like sorry, I lost my confidence level in sgx.

A

A

Questions are they not disciplines at one um would be um projects, and you mentioned smart contracts, so have you in the process of building smart contracts? Have you given it to third party to audit the smart contracts or not.

B

Technically you have to because, like you know, basically a special like you know when dealing with icos and, like you know, when dealing with all the stuff um we don't have like you know, basically, when we're doing like like a regular development like so we rely on a pipeline. Part of the pipeline is just all different kind of automatic scans in unfortunately, for smart contracts. There's no such a thing. There is a sonar. Cube like you know, but it doesn't really yeah.

B

So, basically you, basically you would basically you would publish it and you would like pay about this like someone to hack it like. You absolutely need to have a reviewer, because this technology is still immature even with ethereum, blockchain, Bitcoin and Bitcoin. You can make smart contracts too. So I would not let smart contract going until it publicly kind of seen, not even privately, by independent company, but like publicly, because smart contract is here for a public to see the set of rules right. So the constraints.

A

So funny thing: what happened was uh before chat. Gpd came. We used to give the public contract for a third party company who specialize on auditing and well after different phases. We had different pipelines, features to release and we updated. According to that and funny thing what happened? Was we asked the chat apt to audit it? It did pretty well did.

B

A

Yeah, um so it it showed the mistakes, the human errors regarding the verbs and the number of items also, and so we were in dilemma like, should we just go with Chad CBT audit, or should we give the child.

B

A

Because the auditor will charge a huge amount uh to just audit a single page only.

B

So before again, and then, like you know, we see humans, like you know they can do stuff, like you know, basically tragic video. There is a parameter called temperature. Yeah temperature is a measure of uh entropy and the rental entropy is a measure friendliness.

B

So is this kind of like you know, temperature, like you know you can control like you know how good like you know how generic the charge repeat answer is, but in your case like you know what I would do first before getting to anyone else. So there are lots of like you, know, smart contracts out there and remember when I was talking about church GPT, you actually can access embeddings or like late in the space or embeddings.

B

This is a vector and like the way we and basically what that means you can take all the contracts from like, whatever you see at the contracts from like internet like anyone, you can find you can go through a charge, GPT, you can get a vector and you can run the charge GPS through a smart contract. You you get a vector and measure two vectors as an angle between them.

B

So like in a minion like you know, if you have like an obeak angle, so that it's because in charge, if you do those embeddings, it's actually brain like you know it's a big Consciousness like you know, behind the stuff, and basically, in this case, like you know, when you measure angle, you will see which smart contracts from the public it's similar to yours. So.

A

The wider, the better the.

B

More really yeah cosine is just like zero. Like one. It's like exact match: zero, there's no match right, it's opposite from side right! So basically, and you just compare everything to some everything and get this kind of like Matrix.

B

Remove like you know what is just like less than eighty percent and, like you know, they would allow it to to like you know, reduce it like you know, number of, and what is the Matrix Matrix is a graph, those kind of things, and then you look into a graph, and then you look in the Clusters membership algorithm or cluster algorithms, and then from there you actually can see which group of smart contracts you contract belongs to and also you should be able to see which voluntary Liberia sorry vulnerabilities, this contract May expose because similar contracts they have like.

B

You know this vulnerabilities too, and with all this information you just go into researchers and present it to them, and they just like you know, put their own thing because, like you know, you never eat like human is human right, especially like you know, people who are doing this when you're reading. Some papers or like man, you just like the paper, is five years or five five papers or five pages of paper, but you can see like it's covered here and there from this Dimension.

B

You can see you can you can you can read the people's life? You can here's one here here is another two years and the third page is another five years, so they kind of still like. uh Basically, uh uh um my understanding, like you know tragic, is good for something that generic for like something for ordinary something you like vulnerability. They should be ordinary, but they would never probably notice something that the charge GPT Network never seen of.

C

B

So simple example is a linear regression. So like linear regression is like you know, really it's a basically a it's like you know, coefficient multiplied. Oh wait, coefficient weight, multiplied by the number and you sum it. You call Dot product right.

B

So in this sense, like you know, basically you basically you get like models of weights and like and you just encrypt the weights first and then, basically, you take like a data and you just like multiply one to each other, and you sum everything and you got like in a final result.

B

If the question is like you know, if it's usable for training yeah because, like you know, training the model is not like a really a question, it's a question of regular like regularization, because you have to count like you know, entropy and, like I, don't think homomorphic encryption like you know, there are some papers and there are seen some projects.

B

The like you know they let you use uh neural networks in a in an encrypted space, uh but like, for instance, I work with uh like linear stuff, and basically you train your model on the plane data, and then you just use it on on on another part of the data, and you have pre like you know, result and you just send it out, so the decision trees could be implemented. That way too, because decision transitions like one zero. Yes, no, basically and neural networks is like nothing more than like bunch of submitters.

B

So but it's another topic for like uh for for quite a bit of discussions. So the GitHub is the best friend in that.

C

C

So I guess I'll do so. Thank you guys. uh Thank you for everyone coming uh everyone, everyone online. um Obviously it's been recorded now to go to hyphen Legend on YouTube and get more information. If you have any questions for Alex uh in the chat we've put, his uh is a LinkedIn account, so you can contact him and have any questions.

C

uh If there's anything else, you can also ask David and also thank you again, David uh you always put in the the extra work in and uh thank you for all students of them David and thank you. Everyone online and in person uh for coming I hope to see you guys in our next event and uh just just stay notice. It's on the meat event. Sorry the meet website just just uh keep on checking in and we'll be held. The next event. uh Let me show the comment all right.

B

Yeah. Thank you. Thank Alex thanks. Everyone.

C