ZK-Proofs in AI and US Defence? - Ismael Hishon-Rezaizadeh. Ep630

Transcript

Welcome to Epicenter, the show which talks about the technologies, projects, and people driving decentralization and the blockchain revolution. I'm Sebast and I'm here today with Ismael, CEO of Lrange Labs. Lrangee believes that there's a gap in cryptography within national security and defense. We realized there's an opportunity to take things that were developed in crypto that are frontier hard tech and to purpose that outside of crypto. Zero knowledge proofs provide this way to glimpse inside the black box to determine that the output that you got back has come from the correct model with these set of correct inputs which allows you to determine why the decision was made under what circumstances decision was made and under what

circumstances decision should be adjusted or things should be changed going forward. And I'm here today with Ismmail, CEO of Lrange Labs. Hey, how are you? >> Thank you so much for having me, Sebastian. I was on Epicenter recently and there's been so much exciting progress we've had as a company since then. I'm so excited to be back and talk about a lot of that today. This episode is brought to you by Nosis, building the open internet one block at a time. Nosis was founded in 2015 and it's grown from one of Ethereum's earliest projects into a powerful ecosystem for open userowned finance. Nosis is also the team behind products that have become core to my

business and that of so many others like Safe and Cow Swap. At the center is Nosis Chain. It's a low fee layer 1 with zero downtime in 7 years and secured by over 300,000 validators. It's the foundation for real world financial applications like Nosis Pay and Circles. All this is governed by Nosis DAO, a communityrun organization where anyone with a GNO token can vote on updates, fund new projects, and even run a validator from home. So, if you're building in web 3 or you're just curious about what financial freedom can look like, start exploring at Nosis.io.

Okay, Ismail, welcome back. you've been on uh just a couple months ago, but so much has been happening over at the garage that I wanted to get you back on to talk about some of the interesting military and industrial applications for ZKML that you guys have been working on. A lot of like exciting partnerships have been announced. But I guess like you before we before we start like kind of bouncing off of that conversation from from July where we talked about deep proof uh for folks who who have not listened to that episode who should go back and listen to it. Let's maybe you know set the context here for like what is Lrange building and what is deeprove and um what are you building in the space around ZK machine learning?

Yeah. So, Lrangee likes to think of itself as the preeminent company for frontier research in applied cryptography across both commercial and national security and defense. Um, and what that means is we are uniquely positioned to build zero knowledge proofs, things like fully encryption and then consensus in a way that no other commercial company has the capacity to do really across anywhere in crypto or or defense right now. And so the the tech we build the core of this is something called deep proof which is a zero knowledge machine learning library.

Effectively it takes a model think of an arbitrary uh machine learning model all the way from an LLM to a computer vision model used in a drone or to a simple MLP used decision the movement of assets on chain. And we're able to do two things. we're able to firstly prove that that model has executed correctly um by by effectively generating a zero knowledge proof of the correctness of the execution of that model. The same way that a ZK rollup proves that all the transactions executed correctly, ZK machine learning proves that AI executes correctly. And secondly, we can do this over a configurable set of private inputs. The model can be private or the input data can be private. And this is a very powerful feature because it inbakes

privacy into how AI could be used across both commercial and then obviously government, defense and national security settings. Privacy of AI is a very very big question and a lot of the incumbent attempts to add privacy to AI are based on you know airgapped systems um sequestering data sequestering cloud environments but or hardware based security but they aren't based on first principles innovation in cryptography. Lrange takes the approach that it's going to go to first principles. It's going to rebuild the mathematical fundamentals of how we can bake privacy into AI and then we're going to purpose it as we do across dual use um applications. Let's let's maybe just

kind of talk about one of the use cases that people are most familiar with which would be like chat bots right talking to uh chatbt or Gemini and um how how the inference is generated in in those cases versus say some of the you know other um more private or privacy preserving uh chat bots like um like Venice for instance and and contrast that to some of like to a first principle approach because it sounds like by what you're saying that those systems don't necessarily use a first first principles approach. Is that >> correct?

>> Yeah, I think Venice uses TE's. I mean, Venice is like a they take open source models. They deploy that on a network and they make it private, which that's great. I I think that's fine for some people. You're you're not going to be having a drone swarm that needs to maintain privacy over computer vision inputs that's using Venice, right? You're not going to have an LLM used for uh you know by a field operator to determine a plan of action for some type of um defense purpose that uses Venice. These are interesting curiosities that I think are built by crypto companies, but they tend not to be commercial applications.

I would argue that for something like OpenAI or Enthropic or DeepMind to actually take one of their closed source proprietary models and add privacy on top of that. It requires a solution that is low touch, doesn't degrade performance and doesn't result in the underlying model IP being leaked. Right? If you deploy OpenAI's model on Gemma or sorry on on Venice, you're effectively just going to to just lose all the IP on that. the weights, the biases, the architecture all to be public, right?

It's it's a very different question, right? So, if you want to use frontier closed source models, you have to be able to bake cryptographic security in at the level of the model developer. And that's what zero knowledge machine learning does. It lets you add privacy on top of models and lets you add verifiability on top of models wherever they're they're deployed. So ju just so we understand this correctly um a a model that has zk properties has to be trained so from the ground up it has to be trained as a zk model. You can't take an existing open source model and >> make a zk.

>> No that's not what I'm saying. What what I'm contending is that you don't want to use an existing open source model because the performance of existing open source model is not good. The majority of commercially interesting models are not existing open source models, right? Most people when they they use chatbot, they're using Claude or they're using uh you know '05 or they're using uh you know Gemini. They're not using Gemma and GBT2, >> right? Um, those are interesting applications if you really want to ask your LLM something you don't want the world to know, but you, you know, you so care about privacy, what you're asking that you you you probably shouldn't be asking these questions to begin with.

Um, generally where I see ZK machine learning useful is in applications where these are frontier proprietary models that need to protect both the two aspects of privacy. both the privacy of the person who wants their input data sent in and the privacy of the model. Okay. Interesting. So what you're saying is that a in order for a ZK model uh to be any model any model right any model uh um the efficiency of that model uh and its ability to be to be a private model is very much tied to how it's trained or is that are those two things uncorrelated? Yeah, I think for any AI

model in the world, right, if you develop the best AI model to do something, you're not going to open source it, right? Unless you run a cherry, but you know, if if you develop a great generative model for video, right, VO3 or Nano Banana 2, you're not going to open source that. You're not going to send it over to China. You're not going to give it to your competitor. You're gonna deploy a a website and you're gonna have an API and behind that API your model's going to sit on a server and people are going to come in, they're going to ping it and you're going to charge them per credit. You're not going to give it to them for free. Um, so you effectively need to protect privacy of two things when you're talking about AI. You need to protect privacy of IP, intellectual property for

the actual model. And you also in many cases want to protect consumer privacy. So the the data that someone is sending to that model. A lot of the private AI solutions are just taking open-source models that are publicly available and deploying those in trust execution environments or in a cloud somewhere that they they call airgapped and they are not actually adding privacy on top of commercially relevant models. >> Right. Okay. >> They're they're not taking the frontier models nanobanana 2 v3 clawed and they're not adding privacy on those. taking Gemma 3 with a billion parameters that nobody's going to use for anything and deploying it as a curiosity.

>> H how much of uh ZK machine learning and the the ability to do privacy within models is hardware dependent? None of it none of it you hardware it's it's not I mean you can you get faster inference on GPU and you get faster proofing on GPU right >> um so there is obviously some questions of hardware >> what I meant by hardware dependent is is t like how much of that is dependent on the availability of t tees none of it >> not zk machine learning is entirely just mathematics got it okay which which is arguably why it's more interesting uh because it doesn't require any specific hardware uh to have provable AI that's also private. Precisely. Interesting.

Okay. Um, so a lot of the focus around Lrange and what you guys have sort of like been focused on in the last couple months uh and certainly the like announcements that the company have been around military and industrial applications. Why is there such an importance here within these sort of very highly sensitive uh applications for AI um and privacy? what what's the uh what's the kind of edge that you guys have there? >> I love this question. So, Lrange for the majority of its history has been a commercial focused company. We've developed privacy solutions and and cryptographic assurance for AI usage as well as other types of computation that

that were relevant database computation early in our history um for solely commercial applications. There's a belief that I have though and this is something that I've seen talked about by Alex Karp and the technological republic. It's talked about publicly by Palmer Lucky a lot of businesses pursuing incremental and trivial applications of technology that marginally improve consumer life instead of materially influencing national security and defense. Right? This is a generation of businesses that have been, you know, developing chat apps. Uh it's a generation of food delivery and you know marginal convenience improvements in in the bay, right? It it's the the generation of engineers who were brought

up to build a slightly better delivery app instead of trying to close the hypersonic missile gap with China. It it it it effectively has separated arguably the most competitive and technologically advanced area of this country from all areas of national security and defense. And this has been like what US tech was for the better part of 15 years right before Anderil and Shield and some of the the more American dynamism and and re-industrialization companies started raising very large rounds the last 3 four years there was very little venture financing that deployed into anything in traditional sectors that was doing aerospace defense or national security and that's not a successful state of a

country right what we require as a nation is for the most innovative people of that country, the people who build the best technology, who push the state-of-the-art, who compete, to to develop the things that make our nation special, to purpose that not just to improve the the quality of life of people in Silicon Valley who want their food delivered on a drone, but to purpose it to increase freedom, to increase uh the hegemony of the United States globally. And to do that you you require dual use applications of uh frontier tech. And so when I look at crypto I see very much the the state of Silicon Valley 10 years ago, right?

We've had an entire industry that put a billion dollars into frontier cryptography development for zero knowledge proofs, fully homorphic encryption and consensus or well over a billion if you include consensus. probably a billion for uh ZK, a couple hundred million for FH and funding and maybe you know five six billion for things that are raised with novel consensus maybe 3 4 billion but all of that was purposed for one thing in crypto how can I launch a more efficient token and so we've put effectively the the America's best in cryptography consensus mechanism design distributed systems all of not to building things that make America

safer, make America more secure, allow us to dominate internationally, but we've instead purposed it to things that just flip a quick buck for investors wanting a token. When we looked at this, we realized there was an opportunity to take things that were developed in crypto that are frontier hard tech and to purpose that outside of crypto, not just to commercial but for dual use for defense, national security, military and government. Um and and that's why Lrangee started doing this. We have three professors in applied cryptography in the US. Our research team is led by Babis Papanthu who chairs our cryptography department at Yale. We have

eight PhDs in applied cryptography on our team. We have um multiple patents we filed in the US on novel proof systems. We have five or six papers we've authored on the company's history. And yet like many other companies all we were purposing this for was launching tokens. Now I think there's a tremendous commercial opportunity to instead purpose that not just us but the whole industry to things that are a lot more important. Yeah. I mean I uh I think I it resonates with me this uh this idea that a lot of our a lot of the innovation in tech in the last 20 years has been around increasing convenience. Um maybe this is also sort of a reflection of

>> just like how cheap capital has been in the last you know 20 30 years. Um and also like a relatively peaceful state of the world uh in the last 20 years. I mean certainly since the cold war >> debatables. >> Yeah. >> Right. I mean debatable but I mean like people's everyday sort of like considerations were not were no longer whether or not you know a nuclear bomb would be dropped on their heads where which arguably was the case you know for like a generation before us. Um now that that has that that the world dynamic is definitely different now and I think that the that is reflected in the fact that there's more capital for um uh for

military and industrial applications. Um but the the return profile on those I think are much different than uh the um you know the delivery app type of things for investors. How do you think that's going to change? I mean already I think we're seeing that that change in in in capital funding in crypto. >> Um how do you think that's going to change the entire sort of VC landscape? uh where you know the the time to return is probably a lot longer in a defense company that's building you know drumm technology over uh you know delivery app that can be sold to private equity in like a couple years or you maybe uh sold

off on like the next funding round. >> Yeah I I would actually disagree. I think that the reforms in procurement have allowed uh defense and national security focused companies to see substantial growth at rates that are comparable to uh even the delivery apps. Right? You look at public markets, Palunteer has been one of the fastest growing companies on the equity markets. I mean Palunteer is a comp of you know what you want your public company to do. and you compare Palunteer's performance over the last, you know, let's say 48 months compared to Uber, Door Dash, and other uh more convenienceoriented companies, there there's there's no question that you would you would you

would want your money parked in Palunteer. Um it's it's a hard tech problem, right? You look at another dual use company, Tes SpaceX, right? SpaceX is IPOing. It's going to be one of the largest IPOs of all time, right? I'm hearing people saying one and a half trillion has to be the valuation, right? I mean, these are these are incredibly hard to fathom outcomes, right? I mean, Uber's what, like a 50 billion market cap uh business and, you know, Palunteer I think's over a trillion.

SpaceX could be over a trillion and roll is going to going to come out probably in within 24 months and it's a once again going to be a fantastically large business. Their growth has just been tremendous. I I I think the the fact that the prime defense contractors, the the Lockheeds, the Rathons, the Northrup Gruminss, the General Dynamics, the General Atomics, they have much more competition now. And there's a much larger willingness within the Department of War and other agencies to purchase technology and to to to solicit bids for technology from non-traditional defense contractors. And that is going to mean significantly more innovation, significantly more alignment between US

private sector innovation and US military and US defense uh capabilities. Um I think that just starts with with the willingness of any administration to to purchase defense relevant technology from from uh non-prime contractors and we've seen that that that that is a trend. So I do think the return profile there is going to be uh very very strong and I think that's why we've seen so much investment in this right I think last week it was I can't remember the name hypersonic missile company raised that was closing the gap you know 3 400 million um and you don't see rounds like that for things outside of American Dynamist that are just you know foundation models in uh web 2 venture right now

>> as as as you were talking I was looking at the market cap for lucky Lheed Martin and Rathon. I had no idea what these were but um yeah, Rathon's like 240 billion. Loheed's 100 billion. Um kind of interesting though that uh that software companies in the space like Palunteer are, you know, chasing valuations at like over a trillion or I think 450. I just looked it up. I was wrong. >> Okay. 4 450 billion for >> Yeah. 450 billion. Yeah.

>> For Palanteer. Yeah. >> Um while the you know the guys building jets um are are far below that. Like I guess like um in terms of the the uh you know the geopolitical aspect here you you wrote somewhere I think it was like on a coin market cap um article. I've got the quote here. I thought it was interesting. So superiority in software has become the most efficient way to sustain US leadership worldwide. >> I mean I think that's true. I mean to a large extent a lot of that software superiority has gone into things that are maybe more convenience-driven and and you know now that the world is changing or like people are there's a reaction to that and that is going now

into more uh military and industrial applications you know to the extent that that has been possible because of uh the availability of chips um due to you know somewhat uh friendly commercial relationships with China and you know by extension Taiwan the ability for the US to be superior in software has been dependent on its on access to chips and that access is now you know possibly compromised in 2027. Yeah, I I I I would say broadly that the the ability to be superior in software does come down to chips, but it also comes down to energy independence. Um, which which China also has risks if they if they move on

Taiwan, right? The majority of Chinese oil moves through the Malaka Strait. um which you know effectively if if if Chinese energy was was significantly impaired having large amounts of of chip access or would not necessarily improve their their global position software or AI development. But yes, it would substantially harm the United States position in uh AI development if if chip access from Taiwan was significantly reduced or cut off and chip manufacturing in Taiwan was cut off. And I think there's there's a recognition of this and there's efforts that that both private companies and then obviously the US government is taking to uh improve domestic chip manufacturing and build

foundation for US electronics chip manufacturing. But it's very unlikely that this will catch up to the capacity that ch that that Taiwan can produce um uh within a reasonable period of time. Uh there's there is there is the I mean there is like the time it takes to reach capacity which I think you know we're probably talking about you know in the decades decades um but there's also access to the rare minerals that uh make those trips possible to produce and to I think mostly are Chinese. >> Yeah. Um is is the west kind of by that uh by by that fact doomed long term unless there are there are uh you know sustainable uh business

relationship with China, commercial relationships with China that allow them to purchase these uh these rare metals and and also manufacture chips at a scale that allows us all to grow. No, I I I would I would go the other way and I would say that if China is not able to um affect a military action against Taiwan by 2030, and they they've publicly said 2027's the the the point that they're going to do this. Um, and most intelligence analysts expect 2027. Uh, China has a substantial risk of never being able to and therein significantly reducing its prominence internationally. The Chinese naval

effectively bottled up in the South China Sea. They're blocked off um by the ring of islands consisting of Japan, Taiwan, and um the Philippines. Uh all of their oil moves to the Malaca Straits. their population is aging. Within 15 years, the average population in China is going to be the same as the average population in Japan, which is not very good. Um, uh, Chinese equities have not been doing well. The problem with the SEO system, um, with stateowned enterprises that that that China has is that they are they are not as competitive at implementing and developing new technology as US counterparts. They don't have an open and free market economy. There is brain drain from China. Um there's a lot of

intellectual talent in China that would rather study and and work in the United States in a free country than they would within uh the hierarchy of a stateown enterprise in China. I actually don't think that China's position is necessarily very strong. Um we also have seen Chinese military equipment fail. Um Pakistan was mostly uh using Chinese military equipment during the recent conflict with India and most that equipment didn't work. um uh with the recent uh strikes by Israel on Iran, most of the anti-aircraft systems that were being used by Iran were Chinese and most of them didn't work. Um so there's levels of grift and corruption and inefficiencies in the Chinese system that are intrinsic to how the Chinese

government structures uh control and enterprise. Um, and I think a lot of that boils down to historical issues, uh, issues over freedom and issues over free markets. Um, and so I I think that the re-industrialization narrative that, uh, the US private sector has been financing very aggressively. So building US domestic manufacturing capacity, building AI manufacturing, building the capacity to produce to mind refine rare minerals in the United States, building chip production in the US, building drone production in the US, building decentralized manufacturing so that in a contested time there's a very hard to knock out manufacturing base. Um

plus uh energy independence, small modular reactors, US is very far ahead I would argue um in the economic structures that will allow it over significant periods of time to innovate and outperform China. One area that um that I think is is is is kind of interesting is this is is um well I mean I guess a military development that has really accelerated it since the war in Ukraine is um is is drone technology and and uh and particularly the the use of lowcost drones to achieve um like all all of the uh sort of like um u all all the positive adv advancements we've seen in in in um in Ukraine there and and it

seems Like I mean from where I'm sitting the US approach has been to kind of like throw lots of money at this problem and with with like no boots on the ground, right? Whereas like Ukraine has been actively pursuing this problem and and demonstrating that they're able to like utilize uh cheap uh commercial drones for all sorts of military um applications. >> Yeah. What's the like to what extent is the bottoms up approach that we're seeing in Ukraine a more effective approach to uh to generating outcomes over like spending billions of dollars

on like new drone technology that may or may not be I mean maybe used in conflict but like by the time that technology gets used in conflict there's been like billions of dollars spent on that technology. We we'll only know if it's effective right like when it's been used is the thing. Um, yeah, I I would say it's slightly different than that, right? So, a lot of the drones being used in Ukraine right now are repurposed civilian drones, right? So, DJI drones, like the DJI Mavic. Um, and a lot of these are FPV drones that are being, you know, flown by an operator, like the racing drones and FPV view that they strap an RPG on top of and in contested environments strap a fiber optic cable on. Um and so these are like very unsophisticated but very effective improvised munitions, right? Uh which

are I think in many ways a great piece of evidence on how warfare is evolving and how battlefield strategy is evolving. And I would argue that US drone doctrine has historically prioritized very large platforms. Right? The US Predator drones and the Reaper drones have been highly effective at u pinpoint strikes on targets in the Middle East on the war on terror. Um you know the the the the general Atomics Reaper platforms and um uh Predator platforms. Now, we're seeing both Air Force and Navy prioritize new CCA initiative, the collaborate collaborative combat aircrafts, which are effectively um unmanned UAVs that pair with existing US fighters. So,

you'll have an F-35 and you'll have five CCA drones, which are very large combat aircrafts, flying alongside of it, which is going to hopefully increase the capacity of of of US aerial programs to defend and to uh affect missions in highly adversarial environments uh both in terms of combat with other aircrafts and then combat on on ground strikes. Um but so it is true that a lot of the the US platforms are are much larger than the platforms that are being used in Ukraine right now. Right? The the majority of the the programs the the shieldai Xbat and VBAT the Anderil um uh Fury the andel um Omen these are like group three group five

UAVs. You know the US has groupings of what they've considered UAVs. Most of the ones used in Ukraine right now are the group one and the group two UAVs. So very small managed by a single operator. Now it is true that the US has less of a of a small manu less of a manufacturing capacity and small drones than China does and also has less competitive swarm technology that China does, which is actually one of the problems that Lron is is currently working on, which is how can you have swarms that operate in highly contested environments when base links are knocked out? How do they effectively form consensus and reorganize themselves to be able to affect the mission outcome even if half the drones get shot down, lit on fire, jammed, etc. But I would say that US battlefield doctrine is changing. Um,

and there's going to be a prioritization of both very large platforms and small platforms. Heg Seth has talked publicly a lot about that. I think there's needs to be catching up that the US has to do here, but I I wouldn't say the US isn't learning from from the Ukraine war. I think it's probably one of the the areas that that the US is studying most aggressively. It's also worth mentioning that that probably one of the most effective and undermented uh pieces of of battlefield technology in the war in Ukraine is uh electronic warfare jamming um things that knock out drones. And the US has substantial work that it's doing in this end, right? So, how can you effectively knock down large amounts, large swarms of of of unmanned vehicles as they try to affect uh some action on

the battlefield? So, let's talk about verifiable AI for defense and um some of the work you've been doing with uh with Andreil. Um you you announced recently that um that the Graange was uh building within their Lattis SDK, which I'm not super familiar with. Maybe we can talk about what what that SDK is and um and also how Lrange is is integrating with an SDK and and and and how how relevant is ZKML in that context. >> Yeah. So to take a step back, one of Lrange's core thesis is that there is a gap in cryptography within national security and defense. Effectively the the the modern paradigms for

implementing cryptographic security and defense systems are based on air gap air gapping the system. So just private clouds, private networks, private links, and the assumption that um these these communication links can't be broken since they're encrypted, which broadly is true. Where we believe this type of framing falls down, is in highly decentralized combat environments, which is what we see in modern drone warfare. When you have hundreds, if not thousands of different drones across potentially different manufacturers that need the capacity to form um uh consensus or form agreements on the state of a battlefield and then be able to affect action

subject to the onboard AI or whatever command and control system provided them with with a mission objective. And so when we look at that problem, it's very analogous in our mind to the problems we see in crypto, which is you have 10,000 100,000 nodes globally that need to reach consensus in the state of a shared ledger very very quickly based on uh a set of transactions that could be submitted at any single node at any point in time. It's very similar in our view to to the question of how do drones reach coordination in test environments.

And so a lot of the technology that Lrange is building orients itself around adding enhanced cryptographic security to the use of autonomous weapons and drones in highly adversarial environments. And so that's things as simple as proving the computer vision models that run on top of drones on edge devices um or proving the correctness of command and control systems and AI that runs on command and control systems. So uh Anderoll's Lattis SDK is an open SDK that is developed um by Anderoll to coordinate across assets and drones that are built by Androl as well as by other companies wherein drones can interface,

coordinate and um uh allow an operator to control those drones in complete downstream mission objectives. Um, this is everything from their Seabed Sentry and UVs all the way through Omen uh and Fury and a lot of their their larger airborne platforms. Um, and so what Lrange built was a proof of concept demonstrating how a command and control system that made relevant battlefield decisions with AI could be made accountable and provably correct. We've open sourced this. We've released it publicly and we think this is hopefully the start of enhancing how cryptographic security is going to be used in um defense relevant programs uh that

involve autonomous systems. >> You guys have written about um accountable autonomy >> and uh why that's important. Um >> what is that? What does that mean? And and and why is it relevant to >> to the context of AI and uh and defense? >> Yeah, there's a few places where accountable autonomy really matters, right? If your baseline assumption is that in a contested environment, you only have crash fault tolerant um requirements. So there's no Byzantine actors, then you you assume that the drones onboard AI can't be compromised.

So if the drone's onboard AI can be compromised and in theory someone can jam the thing, they can they can uh spoof communication with it, take it over, update the software, then it's very very important that you can prove the correctness of the model running there such that the output that you're getting back if you know another UAV is communicating and and interacting with with the the first asset is provably correct. Right? It's the same thing as as crypto, right? You have how do you verify that an AI model was was run correctly on the blockchain? Well, you can't have everybody rerun it. So, you have to have the person who ran it prove it. That's very true if you're assuming Byzantine assumptions. Now, it's debatable if you should assume Byzantine

assumptions in a war zone or not. Um, we're not going to get into that. There's some people who think you should, some people think you shouldn't. Um, someone once said to me, we can trust the hardware, the hardware blows up. Um, so there you go. uh but now if you have a command and control system and you want to ensure that the outputs that are being consumed by the end device or the edge device are correct that's another place where you need zero knowledge machine learning effectively how can you ensure that one system that's controlling a bunch of downstream assets or one system that an operator is using and one AI model an operator is using to control a bunch of downstream assets is in fact only making for

example killchain relevant decisions as a result of uh the correct AI models being used in the command and control system. Um that's one of the use cases that we think is the most relevant. Right? So assume you have an LLM that's job is to coordinate some mission objective. It's supposed to take in a bunch of sensor input from a bunch of uh UUVs and then coordinate some subset of UUVs to go and uh you know affect some some battlefield action. uh in this case you would require that the UUVs that are affecting that battlefield action would only affect that battlefield action if there's accountability that the correct model using the correct inputs has in fact recommended action otherwise killchain decision shouldn't be made

unless the correct chain of command the correct chain of um of both data custody and inference custody has been followed this is what we talk about accountable autonomy that autonomous systems are not just operating as these black boxes making, you know, random battlefield decisions, but that they're made entirely with an operator's decision or entirely by the decision of a model wherein there's hard constraints put in place by an operator. Sorry, what's a kill chain decision? Uh, a decision to to affect, let's say, a kill a kill action to to to fire a missile to to to to use some armament for a battlefield purpose.

>> Okay. And in and in this context, I guess the assumption is that the AI is providing that killchain decision and then yeah, an operator is deciding to execute that decision or not. That's often how it works, right? So it's a structured multi-stage process describing how adversaries can conduct attacks. Um so you know under what circumstance uh under what assumptions under what sensor feed in inputs should a autonomous system um uh engage in an attack >> right >> should an operator in a loop should just entirely be AI if it's entirely AI this is a place where we highly believe that verifiability is paramount

>> right I I guess like um so I'm I'm not super familiar with this but I I I suspect fact there are some sort of Geneva Convention laws that relate to the use of AI in military applications. >> Um I'll have to check. Um I don't believe >> yeah I like let's let's assume that you know there are or there may be like in the future some Geneva Convention rules around like the use of AI in war.

To what extent is provability and um and um to what extent is provability and accountable autonomy important then in ensuring that decisions made by AI are in fact lawful >> and I guess this this can also extend maybe to >> uh to domestic use of AI in in policing for example you know it's not only in the military context but also in like national like sort of policing. >> Mhm.

>> Yeah, this is a great question. So, I would say that the the the use of autonomous weapons is very very contentious right now, especially autonomous weapons that don't have an operator to loop. So, effectively, let's say drone that can make a kill chain decision without an operator saying, "Hey, yes, I approve this to occur." The US uh I believe doesn't have autonomous weapons that can make kill train decisions about operators being in the loop. That that as part of US military doctrine there always must be an operator in the loop.

Um that's my current understanding. I don't know if that's changed. Um but you know to be honest and talked about this, Shield has talked about this the use of autonomous weapons is not a new thing. It just feels new, right? You know, when when Caesar was was fighting versus Getter, he dug a spike pit surrounding um uh the the Golic army, right? That is nothing more than a tonous weapon. You walk and you fall into the spikes, right? They don't have to be there, right? You can go anywhere you want in the middle of the night and there's a there's a spike that you can step on, right? World War I, World War II, the use of mines. Um all of these are autonomous weapons. you get too close to a mine, you could be a school bus, you could be a family of of of refugees and

horrific there's horrific implications. So the the idea that you know an armament now is has a computer vision model that says hey if this is a Russian tank yes explode if this is a school bus uh don't that that's a net positive that's substantially better than the state-of-the-art of a of a mind that doesn't have any type of AI usage right but just the question over hey we're going to we're going to to outright ban the use of AI and weapons I is is actually much worse than the alternative. I'd much rather that a missile has onboard guidance to to reroute and avoid hitting the wrong

target. Right? If it's if it's firing at a target and the the comm's link is jammed, I'd much rather correctly rearm and and hit the correct target instead of, you know, hitting a school. All of this stuff, I think, is super important. the more precise weapons can be and AI is a is an incredible way to increase precision of weapons, the the the less collateral damage and and casualties you'll you'll you'll you'll see. Um and so that that's something I do firmly believe. But I I to your point, I do believe that there has to be con uh conventions and organization around what how these decisions can be made such that they're lawful. How is auditability conducted? How can we quickly audit to make sure that all of the use of

autonomous weapons are following a predefined set of rules? Right? Is the AI operating correctly? Right? If um if uh a weapon missed, is it the fault of an operator? Is the fault of the onboard AI? Is it the fault of um an adversary taking over the system? Is the fault of the system being jammed or having a kinetic interference? All of this stuff is super important. And when you add AI, you open up a bunch of questions over what verifiability and auditability of AI means. And that's where zero knowledge machine learning is super super important. Um, and that's been one of our very large pitches that we've made to to almost everyone we can talk to in defense that zero knowledge proofs provide this this way to glimpse inside

the black box to determine that the output that you got back that indicated that uh uh uh an action should be taken has come from the correct model with these set of correct inputs which allows you to determine why the decision was made under what circumstances decision was made and under what circumstances decision should be adjusted or the model should be adjusted or things should be changed going forward. Yeah. Yeah. I was thinking about this like as with a lot of things with AI, I think it quickly becomes a race to the bottom where you know if we have increasingly more more and more AI in military decision-m then like at the bottom it just becomes about taking out the other

person's AI, right? uh if if we're narrowing every military decision to like this is is this a military target and at the same time you have sort of like increasing use of AI then at the end you're just trying to take out the other person's AI or the other country's AI and you're no longer trying to take out anyone other like military or you know uh infrastructure it's all about just taking out the AI but I I would say taking out AI has a lot to do with infrastructure right I mean I think that's the whole industrialization it's the energy grid it's the the where you train the AI.

It's where you run the inference and data centers. It's it's it's where you >> um I meant infrastructure, but oh yeah, I got you. Of course, like yeah, there you know electricity infrastructure, IT infrastructure, but like at the end of the day comes down to that where >> um then I guess like you know having having decentralized AI that's able to function autonomously in in a more of a mesh system becomes more interesting. It does. Um, I want to talk about the token here because, you know, I know you guys still have some applications and kind of products that are geared to more towards the more crypto application uh, crypto use cases, but you know, when you're talking to Andre or uh, or Lockheed about implementing Lrange technology and

the token has a lot to do and a lot to sort of uh, do in the in the decentralized nature of the of the uh of the ZK verification uh aspect of the product. How does that conversation go? And I mean does the token have anything to do with the military applications or is it strictly for for the cryptoreated uh use cases? >> Yeah, so the token is staked into Lrange's prover network with it which is a decentralized network of operators that can generate proofs um within an auction mechanism that that has been designed and developed by our team where work is assigned into the network. The network generates proofs and sends those proofs back to whoever requested them.

The decision of whether or not you want to deploy Deep Proof or any of Lrange's technology in decentralized or centralized capacity just depends on whether or not the person in question would prefer one or the other. I would say for all of our crypto use cases, which is a lot, we work with probably every AIX crypto company in the space right now. We've announced stuff with Sensient Gaia Mirror Zerog Open Ledger, uh really all of these cryptoxai companies. It all is going to be running through and runs through our prover network. Um where our proven network powers all of the ZK machine learning that we do crypto. Now, the defense applications, it'll just depend on the on the decisions that our defense counterparties want to make. Um I can't promise one way or the other. We have

decentralized deployments and our software also works on centralized places as well. Lrange Foundation is the foundation that issues a token and Lrangee Labs is a development entity for the the Lrange prover network as well as deep proof and a bunch of other technology in cryptography. Um and we are very excited to see growth of our technology across dual use purposes, right? Think of Palunteer. There's very different deployments of Palunteer for public sector and for private sector. Um, you know, Gotham, uh, for example, is a is a platform that Palunteer develops entirely for police forces. Um, something like Foundry and AIP are really strongly commercial, but also have government applications as well. Um, and I think that Lrangee is

committed to being a dualuse company, which means that we will continue to expand our adoption in crypto, we'll continue to expand our adoption in enterprise, and we will continue to expand our adoption in government defense. What are your thoughts on Europe in the sort of military geopolitical landscape specifically when it comes to AIdriven military technology? >> Yeah, I mean look, I think there's some there's some great military and defense contractors out of out of the EU um or out of Europe, Rafal uh Tales, uh in France, BA systems in the UK. I would say that these are more legacy style primes more so than they are, you know, upand cominging uh highly competitive manufacturers of of of new drone

platforms, new AI platforms. But, you know, we know the teams, we met the teams at at at these companies and we've, you know, think they're fantastic, frankly. Um but I I would say that you know where I think a lot of the innovation is happening in Europe and in in modern military doctrine around drones is obviously in in Ukraine as we touched on before right Ukraine uses AI and drones. Um one of the very useful things you can use for AI and drones is is allowing those drones to operate in jam environments. Right? So, you know, assume you have a bunch of electronic warfare jammers that are trying to knock out an operator link, right? The the the one way you can do that, the very naive

way is you just attach fiber optic cable to this. This is what a lot of the FPV drones are being done. And you know, now now Russia's launched these like big mother ship drones that have like six kilometers of of of fiber optic cable attached bunch of smaller drones. You know, I personally don't think that that's a very future proof strategy. there's, you know, kinetic issues when you have a bunch of fiber optics attached. But AI provides another alternative, which is you can still jam the operator link. You can have a bunch of electronic warfare jamming it. But if the drone can complete its objective in the last mile without needing an operator link because it has onboard AI, um, then it can then it can do this.

This is if you remember the uh the Ukraine operation where they attacked the Russian jets. it took out to Russian jets. Um, all of those drones I believe were using onboard AI. Um, since the Russian air base was obviously jamming and so these models were trained to recognize weak points in the Russian jets, they got close enough and the jet was able to kind of loer as required to then be able to to affect um uh detonation on top of the Russian jet at the weak spots and that wouldn't have been possible without onboard AI. So, I I think that there is substantial military innovation in in Europe right now, and I think a lot of it's driven by the war in Ukraine. But I'd say like the the world leader right now, I'd argue,

in um drone innovation is the US and China. With the advent of of AI in military technologies, it it feels like a huge turning point to most of the military technologies that we've had up until now that have been mostly kinetic. To what extent do you think there's a sort of like getaway phenomenon or a runaway phenomenon where when one uh uh sort of country or or military alliance really takes off, the other ones can no longer catch up just because their their superiority that the AI superiority just makes them so much more powerful. Is is there a new sort of dynamic at play here that really changes the trajectory of

military advancement in a way that we haven't seen pre previously? I mean I think military developments I I think warfare continually evolves right um and I I think the the the strategies that are employed by nations to counter the strengths of other nations militaries just continually change right I mean the entire US power projection platform often predicates on the US air force US aircraft carriers um wherein you know US carrier group can effectively go very deep blue waters and and project US authority and US power anywhere in the world. And so what has China done to try to combat that? Well, they've developed, you know, cutting edge hypersonic

missile technology. And in the conflict with China, and Pete Haggus has talked about this, so this is not me making it up. It would be very very hard for for for US carrier groups to be able to combat just a large barrage of hypersonic missiles that were fired by China at uh US aircraft carriers. And so if the 16 US aircraft carriers could be knocked out in the first 20 minutes of a conflict with China, it would substantially harm the US's position, right? And so there's obviously a lot of work being done in the US to improve the US domestic hypersonic missile manufacturing capacity as well, right?

You look at, for example, the role of tanks in the war in Ukraine, I would argue is analogous to this, right? Um uh a lot of these have been effectively uh uh uh left as to become artillery more so than they are now kind of these forward moving um heavily armored uh platforms since uh FPV drones that cost $1,000 in RPG attached can disable and take out these tanks at will. So, a lot of these tanks have been now both on the Ukrainian side and the Russian side set back under uh effectively like mosquito netting to stop drones from getting through where they can lob long range uh effectively artillery shells, but they

aren't kind of these forward advancing units. And it's an entirely different paradigm. I think this stuff continually changes. So I I you look at like what I would argue makes the US so special and what result in the US being the strongest country in the world and for for hundreds of you know couple hundred years is effectively the the the market. It's the fact the US can commercially compete um by having a tight alignment between both public sector and private sector. This is the same story of the US military manufacturing base, US manufacturing base during World War II.

It's the same story of of DARPA funding research across across, you know, Massachusetts and and San uh San Francisco areas during during the Cold War. And it's what we're seeing now with with the manufacturing of of of weapons and autonomous weapons in um uh these kind of re-industrialization and American dynamism companies. um when you align these things and you have what is the largest most productive and most competitive free market in the world that is putting resources behind it you you're able to produce stuff that other countries cannot right this is this is why the the you know I tend to be very short on the the whole state of enterprise durability long term I don't think you can centrally plan an economy long term I agree and we'll have to

leave it there thank you so much is for coming back on and uh sharing your thoughts on on this like really interesting topic which um I think is only going to get more interesting as time goes on. >> Unfortunately, I mean, fortunately or unfortunately, uh you know, war war continues to be uh something that happens across the world. Um and I don't think that's going to go away anytime soon. And uh if if the Western world can continue to project uh its values um through through these technologies, uh then I guess that's a better state of affairs than than the alternative.

>> Yeah. I I don't think anyone in crypto wants to live in a world where the West doesn't win. It just is not amenable to the quality of life, the standard of living and and the the the freedoms and civil liberties that that that we've grown accustomed to. And I think there's a moment where there's an opportunity for companies in crypto like Lrange and hopefully others to assist um with ensuring that all of us as well as all of our descendants can benefit from the same liberties that we have. >> Thank you.