Bitcoin’s Biggest AI Test Has Arrived

0:00The combination of domain expertise and AI is going to be unbeatable. Historically, I feel like defenders have often suffered from their domain expertise, like at least to a certain kind of domain blindness. There cannot possibly be a bug here because I wrote that code. That's the explanation for a lot of stupid things that get overlooked, right? Humans just inherently have bugs in software. Like, even Satoshi has bugs that he wrote in the original Bitcoin client. Pre-AI code is from kind of another era. Everybody is going to leverage these tools. Leveraging of this technology is actually one that empowers individuals with ambitions in a greater way to enable freedom that was not accessible before. Welcome to FYI, the four-year innovation podcast. This show offers an intellectual discussion on technologically enabled disruption. Because investing in innovation starts with understanding it.

1:02To learn more, visit arc-invest.com. Arc Invest is a registered investment advisor focused on investing in disruptive innovation. This podcast is for informational purposes only and should not be relying upon as a basis for investment decisions. It does not constitute either explicitly or implicitly any provision of services or products by Arc. All statements made regarding companies to securities are strictly beliefs and points of view held by Arc or Podcast guests and are not endorsements or recommendations by Arc to buy, sell or hold any security. Clients of Arc investment management may maintain positions in the securities discussed in this podcast. I'm just super pumped to be back. Wow, a lot has happened in just a month. I know Kathy, I don't know what the term you call it in AI here. You know, there's dog years or whatever. There's like AI years or AI days and it feels like a lot of- New year's, yes. Warped days. Unbelieveable. It's unbelievable. Yes, it is unbelievable.

2:02Is there any- When you get the cost of something dropping 99.99% per year, that's, you know, inference costs, you're going to get a lot more of it. And, you know, a lot of people don't need to be right at the frontier model, you know, which is the more expensive way to go about it. They can be n minus 1. And, you know, we're getting a lot more of that because corporations are trying to figure out, okay, you know, we don't- we do need to be right at the frontier to answer a simple question. So, you know, I think that's one reason a lot has happened in a month. I couldn't agree more. You know, one of the big things, and I love Rob just to explain Rob Hamilton, who was on a couple months ago with us, just explain what has happened in the Bitcoin space. Because I feel like us bitcointers, we just got the first foray into the AI versus AI kind of battle.

3:03And I think you, you especially in the team, have said a really nice precedent and blueprint for others potentially to follow. The way I like telling this story, as I've been trying to- In reflecting back on it, the important timeline was, Kimi K3 went open-wait, meaning that anyone in the world could run the model if they had the infrastructure and the, you know, the dedicated compute host it on July 27th, which was a Monday. Over the past, you know, year and a half, two years as AI has gone better and better, I very regularly when a new model comes out would look at my own internal code and understand whether any deficiencies, whether any issues that should be fixed. Humans just inherently have bugs and software, like even Satoshi has bugs that he wrote in the original Bitcoin client. It's just a very natural thing. And this is just part of hygiene and maintenance and running and writing software. And what was so drastic about Kimi K3, which is a ZAI open-wait, an AI research lab and

4:07company based out of China, was for the first time, I saw a massive divergence in what the open-wait models were willing to do versus the frontier. And also in this way that it almost has like a personality, it is tenacious. It gets in there and it starts really looking for software vulnerability issues. I'm not sure if the model was specifically trained for this behavior. Just inherently, its aggressiveness and going through and finding problems was something I had never even seen from the front American frontier labs. And so Monday, July 27th, it went open-wait, started scanning my own codebase, fast forward a bit to Wednesday, July 29th, that evening, about I think at that time, a thousand Bitcoin had moved from a particular set of users in the Bitcoin ecosystem who were using a specific vendor called the Cold Card. And it very quickly reports started surfacing.

5:08We're now like Thursday morning. Reports started surfacing all across Twitter, Reddit, you know, just conversations about people were seeing their funds move and these devices were special made so that you never had to talk to the internet, right? Myself and a couple other engineers started poking around and looking at this, understanding that for an attacker very unlikely went through hundreds of wallets and remote attacked hundreds of people. So it must have been some sort of foundational issue. And sure enough, the nature and how the Cold Card specifically, their older model from 2021, the Mark III, had an issue with how the entropy was being derived, where traditionally if you want to secure a Bitcoin private key, you want to pick a number between zero and two to the 256. And that's kind of like a cosmic needle and a haystack of the asymmetry of your power in being able to protect your key. Unfortunately, with some software bugs in the firmware, it was generating a much smaller space somewhere in the ballpark of 32 bits of entropy, which is trivial for computers,

6:14graphics cards, like consumer hardware to be able to reverse engineer it. We know that after the fact, but in the moment, I loaded the Cold Card firmware with Claude code using Fable and said, Hey, can you help me find this bug? And it instantly downgraded me to opus 4.8 because of cyber concerns and even then would not answer it. Open AI code X was more open, but it was kind of indirect and coy. I know people who spent a little bit of time talking to it were able to get an answer, whereas when I used Kimmy K3 to look at that same exact code, it instantly just printed out a full vulnerability report. And this was something that was very revealing to me was like I mentioned this divergence and how the models behave. And quickly after that, it turned into a community grassroots effort of going through hundreds of Bitcoin open source repos and looking through and finding issues. The team, and it's about 25 of us. And that's just like this loose group. I think everyone has since this kind of warning, starting gun has gone off, has very aggressively

7:20started looking at all of their own code. And we found around 10,000, I think it's maybe up to like 12,000 findings now across these 600 repos. And this has been, while it is initially a Bitcoin story because Bitcoin being money, if you're an attacker, that's like the lowest hanging fruit of what you would want to go get. It's a software problem. Like this is across all software. And the bugs always existed, but it would previously take a dedicated team of engineers to be able to figure this out. And to your point about inference costs and dropping, we're now removing the intellectual capital and cost to be able to do this either proactive surveillance for cybersecurity or to execute a black hat attack. And so that's like a very high level starting place. I know everyone else on the call here has plenty of other thoughts that could probably want to add to it. But that's kind of what's happened over the past month. It's been a very busy month in software development in general. You mention it right at the tail end. Something huge that shifted here is you've always been able to get audits.

8:23You've always been able to get very smart people to look at your code and validate that it's working properly. And that has been the normal path. The problem is that it takes weeks to months quite often. If not longer, you often have to wait for a while until whatever the auditor's queue is currently filled with is empty. They can get people on it. Then you have to spend tens of thousands, sometimes hundreds of thousands of dollars depending on the complexity of the audit. And then you get the audit findings back and then you go fix them. You send it back and forth. It's this whole lengthy process. It's still a good process. I don't want to say that we don't need that actual certified auditor path anymore. But that was the only path before or just hoping that you happened to see the bugs that were in the code. Now I can take something and that was what really shifted with Kimi K3 is I couldn't really do it with Claude. I couldn't really do it with Chatchee PT when Kimi K3 came out even before it went open. A switch flipped and I was immediately able to audit all of our repositories properly and get all of the info I needed without having to go to an external auditor. And that switch of me being able to spend a couple hundred dollars and get a good read

9:25out of the repository. Now I still need a good understanding of how our code works to be able to figure out what's false, positive, what's real. But that shift between weeks to months and tens of thousands of dollars to a day with our own internal review on top of what Kimi provided and now many different open way models provide. That was just a massive shift and was a shift in favor of attackers initially. But it's also something that now gives defenders a unique tool set that we have never had. And that's been I think something that often gets overlooked is people panic over bad people can use this stuff now. Instead of also realizing that good people can use this stuff now too. Like we have to be on the cutting edge as people building in the Bitcoin space leading the charge in terms of securing our own code with all of the amazing capabilities the AI model spring now rather than on the back foot constantly. And that's going to be a test of kind of who can keep up and who can make this work in the long run. But like Rob said, I think everyone's picked up and started doing these cuts a lot. It's now thankfully. Why are these open weight models able to do this and closed weight not the closed weight

10:31ironically have more guard rails, more censorship, more something and it's getting in their way. Ever related question at Seth and Rob where your company signed up for the various like vendor or specific security versions like code security and other things? I'll quickly answer for for us. So we had applied to both cloud and chat GPDs one a while back. Both are still pending. We still haven't been able to get into either program because I have access to good open weight models now. I haven't pushed on getting that because I don't think they provide a lot of value on top of what we already have. But I know that varies from organization organization. I've had access to open a eyes cyber program since May. And then I think later May early June, I got access to anthropics to your question like to that point. It's been a frustrating couple of weeks because I would take findings from my open weight models and say, Oh, there's a vulnerability that's confirmed from the person who writes the software. Now I am trying to fix it. And I would take it to my properly gated like I filled out the application.

11:35I scanned my face. I went through all these steps. And I would say there I have now noticed a bit of a difference. There's a couple it's a fast moving space. So a couple things have happened. Initially I was getting rejected on both sides like neither anthropic nor open eye were helping me. This then open AI has come out with a model called daybreak, which is their GPT 5.6 soul, but without these guardrails removed. And just yesterday and the topic came out with fable 5.1. And that one now if you're in the cyber program is way more accommodative to your question Kathy unlike why it's the American frontier labs absolutely see this information like their models are more advanced. What happens though is you send your query. It gets a response. It looks at the response and says, Oh, we can't share this. This could hurt someone. And then they either modify it or just outright deny sending any information back to you. And so I think in response to these open weight models like I mentioned open AI and anthropic

12:40now have become a little bit more permissive and letting you look at stuff. I think they were naturally nudged into let forcing their hand so to speak to let you see more of this information. I so far with both fable 5.1 and GPT daybreak like these new cyber aware models. They are on par with the good open weight models. They're not I'm not finding many advanced findings that I could not find with the open weight model. So if seems like they just nudge the guard rail enough to be at parity with the open weight models at the moment. I tend to agree. So I think our team's experience. We had been on those programs for a number of months before this incident and we'd actually had great experiences that those tools seem to work just fine for us. I know that's been inconsistent that's part of why I'm asking is it's not every team that is at that experience. But ultimately it's not that the models as we've just discussed it's not that the models from the private companies cannot do these things. Of course they can. The company just don't want us to use these tools in those ways and they've been forced to recalibrate over the last several months as many people have been publicly saying why

13:42is it that your tool won't do the thing that I need help with when this open source tool will do it or worse why will your tool introduce a bug that then it won't itself fix because it now thinks that fixing the bug is dangerous. So this is kind of just an example of like the calibration that's required here and in general I think most of us Bitcoin or something this call would probably agree that at the end of the day giving people the access to the tools is the safest way to make the tools ultimately safer for everyone. And I think it's at least been nice to see the companies move a little bit closer to us over the last couple months in response to some of this. It'll be very interesting to track whether the closed models become better than the open models because what you're saying is they're not really now. But one of our hypotheses is that this is going to become a huge business opportunity for these frontier models. You know, just pushing the envelope and for that reason they're going to that well for that reason and because it is the right thing to do they're going to push hard in this

14:45direction. But as you say, there are also risks and I do understand them. Competition just such a good thing. I love it. The more competitive these folks can be like me, Rob or Lorenzo, did you like on that Monday, July 27th? Did you feel like you had a butter knife to like in your entering a gun fight? You know, when you were using what's the analogy here? It's very similar to that. It was it was very interesting to see the like I said, the divergence and maybe to be to Drew's point, we had found things with the cyber tools that were available. It was just a step function increase in all of a sudden where I rapidly scanning these tools regularly with these trusted cyber access programs and then Kimi K3 walks in and finds a whole new thing slew of things that had never been seen before. Right. And so it felt very uncomfortable being like, Oh, interesting. I was paying top dollar for what I thought was the leading stuff and it turns out that

15:46there's a whole multiple layer here of kind of improvement on the models in general. And so the silver lining is over the past three weeks. The American Frontier Labs now are definitely in line with the open weight models. So since, you know, I think GPT Astra, the new ones about to come out, we're not going to know like is that going to be a whole new step function again in this happening all over again. The good news though is that everyone's now way more aware of this and more proactively looking at their code and thinking through their security processes. And I think while there may be more another wave of this, it will become with a lot more sense of preparedness as opposed to a blind signing people. There's so many things that come to mind because I feel like these are conversations we've had like in the last weeks. Robert, let me ask you a question. I actually don't know the call card team. So from what I was looking into this, I think cold car is like a minority vendor like within the hardware wallet space. I think they have like, sure, one or two percent market share probably. So like nowhere near.

16:47I think we're ledger or trezzor are right now. Do you think to me, it seems like they just hadn't like they haven't been using any AI model to scan this like from what you're telling me it seems like Keme K3 was just better. But from my perspective, it seems like they were just not running any of these models on their code base. Like what do you think before Rob answers that? Can I just like to even more Lorenzo? So I actually like drove so Rob doesn't do this full time, by the way. He jumped into this and like joined like started this red team and so on. And it was it's been simply amazing. And like I actually delivered some food to his house just because he hasn't been sleeping and so on. And I was like, Hey, how is Bitcoin? And he's like, Oh my gosh, Bitcoin rock solid, Bitcoin's amazing. I just wanted the audience to know that this is an attack on an application or a company that was providing a service or a product.

17:50And it was their code base that was attacked. Bitcoin still beautiful, still monetary network, still thriving. Okay, Lorenzo, to your question, I don't know what they were specifically using or not using. I don't have that information to the point about the whole ecosystem. I think myself and many others, we've put tens of thousands of dollars specifically looking at LibSecP, which is the foundational cryptography library that Bitcoin uses. And there's, you know, there's no smoke there. Like there's no catastrophic like, Oh, wait, actually all the signatures are bad, right? Like the protocol itself is strong. Right. And I think an important thing to call out to, I had customers with Anchorage who used cold card, drew it unchained. Like people used cold card and there weren't losses, right? So like the idea of like multi signature and diversifying your risk, they're, they're like, this is just a concentrated risk of ruin of a single point of failure because it was specifically people using the cold card with the default entropy on a single signature.

18:53Right. So it is the most concentrated view of the risk vector you could get. And, you know, you're able to still keep your Bitcoin safe. And I think that's just an important thing to evaluate in the context of all this. Like to Rods Point, Bitcoin's not broken to your point about Cull card. It's the one specific vendor. I don't know what they were specifically using or not using. But I think this is a stronger learning. And I think everyone has a aggressively sweeped all of the hardware wallets since then. I think those were one of the first things everyone started looking at. And there's been nothing else like this that's been found. That's like been a critical issue. You would think the team had spent like at least like two or three K of, you know, like Fable like to scan that. I assume that was the case. I don't know. I would say like, I think it is unfortunate. Obviously, if you look at just like the hardware wallets space, like it, to me right now, it's kind of, it's, it's a bit broken, right? From the product perspective, because like Rob, you mentioned like this was very narrow

19:56within Cull card, but obviously, you know, it kind of, it goes on on media. People are scared. I know Trezzler has had like some problems like in the, in the past. And Ledger has had like many, many data leaks. Obviously of, you know, the clients kind of buying the hardware. And so it's, it seems, to me, it seems like the three biggest vendors. And today, frankly, like it's, it's just like they've, they've lots, they've lost so much credibility, I think. And I think it's probably a space that needs to be rethought from the, the ground up, right? Because it's, it's not only like you said, like this was like basically using a bad random organizer for like the seed phrase, but like Ledger is more like the supply chain kind of attack where, you know, they get, they get your address and your number. And then they connect you and then like Trezzler was another one. It feels just like it's, you know, there's so many angles to this.

20:57And I feel like today, like at least for me, like I would have a hard time recommending a hardware wallet, honestly, provider to me. I feel like what you're expressing is what a lot of people are feeling. I, it's natural, I would say to feel that way. But probably I would imagine Rob would agree. I have heard him say this recently. It's a, this is one problem does not the entire category invalidate. And it's important to realize that there's just as many of not more terrible things going on on the other side of that, you know, equation, right? That if you are afraid of self-custody because of this specific issue, that doesn't mean you should uncritically embrace just trusting another organization. For some people and some organizations that might be the right move. But there's many, many instances of third party custodians just disappearing with Bitcoin or getting hacked and it being lost and other things. So I really don't like the frame that self-custody because of this incident is now untrustworthy. It's, although it's reasonable to say I have uncertainty as to who to recommend as

21:59a hardware wallet provider. Right. Because of all these issues. I think that's very reasonable. I think that's why Rob and, and maybe Seth and others might agree, well, then don't just take one, pick multiple of them, glue them together, right? Leverage their weaknesses against each other and build something stronger than anyone can be. The only painful downside of that though is the other point that Lorenzo mentioned, which is the physical risks that come with giving hardware wallets. And you actually only multiply those risks when you do multi vendor multiseg as well. Unless you can get them all in person or get them shipped to a box that's on your name or FedEx pick up something like that. Like I think that that always has to be part of the consideration. I don't think it's why interim solutions, things where you can turn like an old phone into most of a hardware wallet. Yeah, maybe it's not 100% of the security you're getting from hardware wallet, but it gives you an interim solution are really, really useful for more people than I think realize it because there are so many downsides with hardware wallets. Now, I say that as somebody who has worked in the hardware wallet industry, like I worked for foundation before I joined Cakewallet, I will understand the pros and the cons of both sides. That sounds like a downside of the current state of the distribution model for hardware wallets.

23:03More than the technology itself, that you probably agree. If we're in the future and I can buy hardware wallets at the gas station alongside a pack of cigarettes, well then it becomes a lot easier to obtain these things without having to worry about supply chain attacks or my identity being revealed. So we're just not there yet. There's a tiny number of vendors and cold card was itself a tiny company, but very important in the space. Ledger and Treasurer ultimately, these aren't companies the size of Apple or anybody else that's going to have devices in every single retail location across the country. I think it's impressive that Ledger even has devices at all available in best buys and places like that. It's actually been awesome. I know a lot of people, Rob, maybe you have similar stories, clients who are scrambling, hey, where the hell do I get a device to replace my cold card? Well, turns out you can just walk to best buy. That's pretty cool. I think what you're saying is basically just like the UX, right? And the user experience, obviously, of having several hardware wallets, several seed phrases, multi-sig, where do you store them and all that stuff.

24:05So obviously, it's not only the tech, right? It's like distribution, like the user experience and everything. So yeah, I agree. As you said, you're like maybe it's a space at least on distribution and supply chain that needs to be kind of rethought from the ground up. Like I was doing the exercise and maybe you guys don't agree, but it's like if you had to recommend like to a new user and I would, you know, Bitcoin is in a category itself because I think it's much more robust than crypto, right? Like the less the rest of crypto. So Bitcoin is really in a category of its own, but like, would you recommend like self-cost city, put it like in an exchange, put it like in a spot ETF? Like you have other solutions today, right? And I was just doing the exercise the other day because of the, you know, of some of these news, basically. As I've said that as long as you're Bitcoin is safe and you're happy with that, I'm happy for you, right? Like as I was a general principle of everyone protecting their wealth.

25:07And I think as a general model, like mentioning the ETFs, like I think it's way easier if I'm managing my estate, if I'm managing my assets. And it is way easier for me to hold the Bitcoin ETF. And to sell that and get some cash, then to use my geographically distributed multi-sake spread all over the world to, you know, buy a coffee, right? Like obviously in the capital stack of these things, there's trade-offs and how you manage that. And that's for each individual. And I think the uncomfortable, like, less and walking away from this is there is no one size fit all answer. Like risk is an individual, like risk management is an individual person or a company's decision. And they have to understand the many different paths like between liquidity and security and all of these different things in privacy. And like, it's a very complicated matrix that you have to kind of walk through and there is no silver bullet. And so I think that's just like an important, like, distinction in all of this is that you would want to make, be thoughtful about these decisions for yourself. And I think on, for the self-custody side, I think people get complacent.

26:08They sit like these wallets that were stolen from were set up in 2021 through recent. But a lot of Bitcoin moved, that hadn't moved in years and people were, oh, my security is set and forget it. And security is an ever-evolving game and not something that you want to say, well, I made this decision four years ago. And now I'm just going to never worry about it ever again. Like if you're going to be responsible in the sense of owning your own private keys and doing that stuff, that's a thing that requires constant vigilance. I, you know, it's comparable to driving a car or owning firearms where like if you're going to take on this responsibility, there's duties that are associated with that that you have to fulfill and being vigilant is a critical part of all that. I love the firearm analogy, right? I think I forget who it was that has said that. It's like, you don't buy guns and leave them at the firing range, right? Keep them at your house because that's part of what the value of owning a gun is. I think similarly, there is value in being independent from existing financial systems, but then it puts a lot of burden on you to do it well. And as we've all learned, that is not as drop-to-sets, not set and forget, something that requires work. You need to have good partners, you need to do your research.

27:09That's scary than yes, working with third parties and relying on them can help and be a part of a solution. But also on a highlight, we just got done talking about how AI is like real now and is attacking all sorts of things. And it just happened to get to Bitcoin in hardware while it's first. It is 100% coming for the social media websites, it's coming for a tax website, it's coming for banking websites, it's coming for everything. It's just harder to steal things in those places. That's why it's coming later. And so in a world where AI is manifesting and has powerful the way that it is now, the idea that, oh, this is too complicated for me to deal with, I'll just trust that brand. Well, you're putting a lot of trust in that brand then to not make a mistake as AI continues to get more and more and more powerful. And yeah, you can kind of criticize coal cart and say, oh, they were a small team, they didn't do a good job, they should have done better fine. But we'll be having the same conversation within an ear about a large custodium that got attacked and all their funds were stolen. I guarantee it. It's happened before even without AI. It's definitely going to happen now. And in that picture, I think you're left with, if you really, really want something that's

28:12robust, then don't pick either of the extremes. Right? Leverage their strengths against each other, do an individual analysis of your own stack and your own risk tolerance and what you think you believe in and don't. And then focus on it that way. To Rob's earlier point, there's a great, I think economics joke about this about like two economists are walking down the street and one says, hey, is that $100 bill on the floor? I mean, the other one says, no, I can't beat. If it were someone would have taken it, right? Then they kind of move on. I think that's kind of what I think about Bitcoin, right? That I know even before Rob and the red team go after the Bitcoin source code and inspect it with Kimi and other tools, I know it's probably secure because if it were not, people will be stealing Bitcoin right now. Like we'd all know about it. And I think what's so remarkable about the cold art attack is it did not take like months for the attacker to find the attack vector and then start to exploit it. They began to exploit it immediately. And something I just want to toss out for conversation is that this is related to a very interesting idea, which is the collapse of the market for zero days.

29:14All right, today, like the actual economics of computer security rely upon market forces that help govern the sale and discovery of zero day exploits, zero day exploits just to be clear, these are bugs in computer code that attackers know about, but defenders don't know about yet. That's what a zero day is. And typically if an attacker finds a zero day today, what do they do with it? The answer is almost nothing, right? If I find a vulnerability in Zoom or a Microsoft word or in an esoteric Siemens Westinghouse machine control program, it's very hard for me to make money from that. Well, what I can do is I can list my exploit for sale on the dark web and I can hope that someone buys it. And who might buy it? Well, someone who can take that exploit and weaponize it, ie, the US government can weaponize against the Iranian regime, this bug that you found in this microcontroller in the center of use, right? That's pretty cool for the US government. Maybe they're willing to pay to buy your bug. More commonly, it's companies themselves that purchase their own zero day exploits. That's the largest purchaser in these markets. People use the black cat ecosystems as a way to do security research.

30:16And when people find these bugs, they then purchase those bugs. So in a more normal industry, if Cole Card was a, you know, a fitness tracking app or something like that, right? They would have found the bug. They would have posted it either through responsible disclosure if there were a white hat or in a black cat context, if they weren't. And eventually someone would have noticed it. And maybe they would have bought the zero day and then it could have fixed it potentially. In this new world where there's a thing called Bitcoin that you can just steal, none of them makes sense anymore. Like you don't spend months listing your exploits on zero day markets. If you're an attacker, you just exploit it because it's immediately going to get you money. And this is both scary and positive. I think it's scary because it means attackers have much more willingness to attack us, especially when we're close to Bitcoin. But I think it's actually good because it kind of means that zero days don't exist anymore eventually, right? Like if as the world's software becomes closer and closer to Bitcoin and Bitcoin is used within more software and more software has the capability of moving Bitcoin around, zero days become much and much harder to exist. That's good for defenders, right? Defenders don't like zero days.

31:17And so I think we're kind of already hitting that point now finally in the set of software and hardware that is adjacent to Bitcoin, namely hardware wallets and other things like that. I'm curious what Seth's experience is here as a wallet maintainer. I think this is just a new era. And I think it's actually really exciting and kind of healthy. It feels scary because we're in the middle of the transition. But on the outside of this is the knowledge that it's not like there could be a latent bug in your tool for four or five years, like there was with coal card. It's more like if the tool isn't being hacked right now, it's secure, right? Being able to get to that point is very exciting for me. And for Seth, that someone who maintains wallet software, I've listened to your Freedom Tech Friday podcast. You are living in the future because rather, you're hosting all this compute locally and you have like developed this entire DevOps system that is basically vigilantly constantly sitting on top of your code base now, which I think is going to be the model for software companies in general. Like if you want to just talk a little bit about your journey on that because you're

32:17so deep into it, you've massively accelerated where maybe I am off calling an API to anthropic or open AI. It's all local for you. Yeah, I mean, I don't want to go too deep because I'm in my dear a little bit, but the short answer is as we started looking to this, I mean, we started when Kim and K3 dropped before open weights, like when just the initial model dropped and you really start using it for auditing and realized like, oh, this is actually incredibly useful. Like we need to start this and build this into our processes right away. The downside was immediately when Kim and K3 launched, it was so popular that it was almost unusable via API. Most providers could barely serve it because it wasn't open weight yet. A lot of providers literally could not serve it because they would they would host it themselves. And immediately just started hitting issues where the actual usability of the model upstream was really, really painful. And I guess I we didn't have access to a cloud or open AI cyber programs at that point either. That has snowballed quickly into realizing like not only is this something that's so essential that we need processes built around how do we do regular cyber security work around our

33:18own repositories. But there are a lot of advantages to owning the compute yourself. Now, when I say owning compute, we have about if I open the door right behind me, we have about 40k worth of Nvidia DGX bark. So just a little tiny like knuck or mac mini size box that you can cluster together. And they serve all the compute that we need to run scans on every commit that's pushed to any open PR to run against every PR at the moment it's merged to run against the whole repo at least weekly to run against all of our upstream dependencies and validate all of them. So really nice thing about running locally is obviously a we don't have usage limits. We don't have to run into that. We don't have things where they can just change the billing or change the price of models. We don't have to deal with any of that. But more importantly, none of the findings that we discover or the actual exploits that we develop against those findings ever leave our premises, our network. That's amazing. And so we can ensure that if we discover something, it can't be exploited by someone else, some attacker in the path, clawed in open AI, can't, I mean, I wouldn't weaponize it, I

34:22would hope, but they can't. Some of the more sketchy upstream providers that are maybe operating out of China are not someone that we would want to be shipping exploit information to so we can keep that all in house. And I think something that's often forgotten in all of the the red teaming and I know like Robin, you your team, you all have had pushed on this hard, but is often forgotten by others is just finding a vulnerability. It's not proved that that vulnerability is actually useful. And all like if you look at our security disclosures, since people realize the AI is kind of good at this stuff, so many reports from external security researchers, but none of them validate their findings, essentially against an actual running version of K-Quala. I'll be honest, the first couple of days, I was just handing things over at it of abundance of caution. And then quickly it turned into and this is like a conversation around harnesses and like building tools, like if these models are really powerful, how you direct them with these harnesses is really impactful. And part of what I started doing with my initial scans were opening K-Quala 3 with a fresh

35:25instance saying, are there bugs here? Hit enter and it would come back with some stuff. And then quickly iterating on if you find a critical or a high vulnerability, you're required to do a full proof of concept to show that it's real before surfacing and making a final, to try and add some teeth to it with steps on logs of how to reproduce it, the exact patch set and how to fix it, and kind of guiding that as part of the use of these tools to make them more impactful. And it's interesting how you can just have a model out of the box that's pretty useful. But if you're able to direct it on a harness level, you're able to get more impact. And I think ultimately every company or software vendor is going to have their own harness relative to the context of their own project. Because if every software project has a unique way of doing things and their threat model and how everything works, it makes sense that you and I as different proprietors of software are going to have different needs to be able to actually make it usable. And so that's a whole software market that had not previously existed and being able

36:27to kind of make custom bespoke harnesses tailored to each software project to kind of really maximize the value you're getting from these models. Yeah. So I'm doing it on that a little bit more to like that's one of the advantages that defenders do have is you I would hope have the best understanding of your product compared to some outside black hat or something like that. And so you can build better harnesses, better prompts, better verification, better exploit chains than any attacker would be able to. They're not going to spend all of their time on just your project where all of your time is going into your project. And so when you do actually put in the time to customize and build around it and build up a wealth of data around what false positives are, what true positives are, what works and what doesn't when exploiting proof of concepts, you continually improve your own tooling so that your project becomes even more secure in a way that attackers just simply will not have the ability to do at the same speed or the same depth. Specifically on that point, and I observed this multiple times scanning these many code bases, we would I would do a generic scan on a code base that I wasn't a domain expert

37:31because I got to a point where I was just mass scanning code bases. I would hand it over to someone who's maybe a domain expert in that kind of software. And this happened multiple times where if you've ever had a conversation with someone and they say something that causes like a cascading sense of many realizations, I would take this library, hand it to someone, they would have a couple thoughts, like a couple half sentences and look at this one line of code and I would copy like they're like their casual like comment into the harness and then the model would say, oh wait, actually, oh wait, actually, oh wait, actually, and all of a sudden you would get a deeper and deeper sense of the vulnerabilities there. So as much as this is lowering the cost, like a slot machine, yeah, like you hit a jackpot, the differentiation of labor and the specialization and understanding the stuff deeply, you get to see a really wide difference. And if me generically scanning code bases versus someone who's a domain expert in that, it's an interesting tale with AI people talk about like, you know, like lowering the cost of labor and like things that used to be really time intensive and cost intensive are growing

38:32down. Being a domain expert and understanding these things deeply, you get an even further compounding advantage leveraging these tools, which was a very interesting realization and kind of from the defensive tech side, since you understand your code base more than someone else, you're going to know exactly where to poke in prod to get even more out of this. So I think to your point, it does benefit the defenders. Yeah, it really does. And we've seen this wear an attacker would get lost on rabbit trails and false positives that I can just tell my CTO or one of my devs, hey, can you double check this? Like this is validated. I did a multi model validation, like everyone agrees that this is legit. And in five minutes, one of our amazing devs will just be like, that's absolutely false. There's no way you can actually exploit this. That doesn't make sense. He'll give me one line of why it wouldn't work. And again, I would send that back into the model about, oh, yeah, you're so right. Like, yeah, there's no way that would work. But if you were just an attacker and you didn't have that, that ability, that deep knowledge of the product, you may spend hours or days following this rabbit hole, building an exploit, it's trying to do it.

39:33And then realizing, oh, no, that actually doesn't work in practice. This is the first time I've heard. And we had this hypothesis. We said the combination of domain expertise and AI is going to be unbeatable. The domain expertise is critical and gives you an advantage because you know your code so well. And you know your space so well. And hackers probably are all over the place, right? And I might just add that historically, I feel like defenders have often suffered from their domain expertise, like at least to a certain kind of domain blindness. Like there cannot possibly be a bug here because I wrote that code. Sometimes I think that's the explanation for a lot of stupid things that get overlooked. And I think Lorenzo, you were asking earlier, hey, wasn't the Col card team scanning or didn't they have any practices to look over their code? Well, I'm sure they did. I'm sure that they were. Sometimes domain blindness means that there couldn't possibly be a bug here because all we're doing is we're calling out to a piece of hardware. So that's not our fault.

40:34Right? Of course, I think that's where the AI can become very helpful because it's sort of like a little attack dog like, you know, zealousness for going after every single thing is a way to combat domain blindness and force you to relook at something with fresh eyes, namely its eyes. But then of course, and this is an analogy, I think I want to continue on a little bit maybe in the conversation if you guys will let me. I think there's a certain sense of like computer is turning into biology here. We have this metaphor of locks and keys, which is a little too physical, I think for the coming era, like we think how does your body secure itself against external threats? It's a very different kind of mechanism than a lock or a key, though kind of molecular biology wise that's still going on somewhere. I think software starts to look and feel a lot more like that. Like I think of, you know, the machines in the back room of Satsang over there, that's like a weird immune system that he's built, right? And I think of like what Rob, what you were just discussing, that sort of uncritical AI reporting of vulnerabilities without sufficient validation, it's a little bit like inflammation,

41:37isn't it, right? Like it's just like, hey, you get more and more crap-tastic code layered on top of something that didn't need to be fixed in the first place. It's almost like an autoimmune reaction. So I think as we start to learn these tools better, like we have to take our domain expertise, our knowledge of technical systems, but maybe also a little bit of inspiration from a world that is maybe somewhat distant to many of us, which is literally the world of biology, living and evolving things, and see what we can do to make our computers a bit more like them. You are absolutely right. And Rod and Lorenzo know how excited I get when we talk about convergence of technologies. Now this isn't direct convergence, but neural networks are patterned after the brain. So I think you're absolutely right. And the other, the other, the earliest kind of aha moment, I would say, for us or for some of us, was the example of AlphaGo, the ancient game of Go and the Korean expert just

42:38believed no machine was going to beat him. He was number one. He was the champion, and he went in very confidently to compete. And he lost, and he was crushed. He was absolutely crushed. And then he recovered, and now he's just playing against machines, and he is becoming so much better than other human beings that are not competing against machines. So. And if I could jump in, I would just say, that's where I think we just all level up. I mean, Seth, could you imagine right now how much hardware you have in your closet now, and being and doing what you've been doing over the last, I want to say, one month, like building this immune system. I think to Drew's point is like the exact point. I actually think Kathy, like so many now businesses are going to like read, like look at their operations and say, how much do we need to invest in our own physical infrastructure or there are immune system such that we can keep up with this new frontier world.

43:40And we need to reorganize, rethink how we're doing this, right? Absolutely. And that's where I, you know, someone to you, Kathy, just a extreme optimist here. I think the domain expertise, I think there's going to be a massive opportunity of just human capital, new jobs, creation, a number of new innovations on this frontier. And it's just going to be absolutely insane. I mean, it's going to be tumultuous. Don't get me wrong. But all of these things, and at one point now, I just want to bring up, you know, we got probably 10 more minutes or so is, you know what's so cool about Bitcoin is nobody just tapped Rob Hamilton on the shoulder and said, Hey, why don't you quit being CEO of Anchor Watch for the next like, you know, month? And why don't you just organize a red team to work on this? And Rob, you know, mentioned at the beginning, there's 25 people plus, and there's some people we don't even know that are just actively working to defend Bitcoin and other code bases out there. How cool is that?

44:40And one other thing, Rod, this brings into the discussion something that David Poel on our team has worked on. And this is the quantum threat. And you know, some new flavor of quantum neutral atom caused a big stir in the community. I think we talked about it on Bitcoin brainstorm. And the breakthroughs there were caused because researcher were caused by researchers leveraging AI. What has happened since then is that now the core developers and others in Lorenzo maybe you can talk to this a little bit. They are open minded to using AI themselves to protect against the threat. Pre-AI code is, you know, from kind of an other era, right? And I think everybody is going to leverage these tools. You know, I think from a government perspective and national security, even, you know, having a budget and helping out like Bitcoin, you know, to remain secure, I think in the long

45:45run. It is going to be critical. I think it's going to harden a lot. I think there are already like many teams like working on this and like this, you know, for profit, non-profits and, and, you know, it's very, there's a lot of people working on it. But we talk about this a lot like internally. I think like code that was written like, you know, two or three years ago is definitely something that teams now have to look at it like very, very carefully, right? Not to mention 10, 15 years ago, right? I left to bring up the quantum thing that Kathy just addressed, if you don't mind, Rod. I thought there was a huge, huge parallel for me between what it might feel like in the future when quantum computing potentially becomes real and indeed potentially does so in a sudden way. And the experiences that we had during the coal-card incident of what feels to me like four years ago, but I realized was last month. There was a certain aspect during that attack of, and we don't have to get into all the details of how we, how this occurred, but there were certain people who were using, let's

46:45say, multiple coal-cards and were, if they didn't spend their Bitcoin, actually, okay, at least not under immediate threat that no attacker was necessarily going to be able to steal their coins if they didn't take any actions, but then they also couldn't really spend their coins safely because as soon as their coins would be pushed to the mempool, the transaction that they would author to spend their vulnerable coal-card generated coins, soon as that would hit the mempool, it would be potentially able to be attacked by an attacker, especially if the attacker was monitoring it, if they knew about the balances and the advance, if they amounts were high. And that is so reminiscent of what we have theorized in a sudden quantum computing attack to look and feel like that if your funds are sitting in, quote unquote, quantum safe addresses like P to S H, there's no way for the quantum computer to attack them, but as soon as you attempt to move those funds, you reveal a public key on the Bitcoin blockchain in the mempool, and suddenly this gives the attacker an open into use of their quantum computer to come after you. Now, for the coal-card attack, it wasn't a quantum computer, it was a stack of GPUs, it wasn't due to inherent differences between classical and quantum computation, it was due to a bug

47:48that caused bad entropy on a subset of devices, but nonetheless, during the moment, I was like, man, this is what it's going to feel like. It's going to feel like a bunch of people who don't understand what's happening, who aren't technologists, who are affected, their life savings are affected by some weird technology thing that happened. We're working with them to try to rescue them, but merely broadcasting Bitcoin transactions now suddenly feels dangerous. What are the solutions to that? What are the ways that we can bridge past this? For unshained, we wound up building a solution which allowed our clients to directly broadcast their transactions to a minor, skipping the mempool entirely, and that turned out to be very helpful for recovering many, many thousands of coins across our platform. I don't think quantum computing is going to happen anytime soon. I do think it will happen eventually. I don't think it will happen suddenly. I think we will have warning. Those AI researchers are not keeping that stuff secret. They're talking about the advances that they're making in quantum computing as they make them. Nonetheless, in the back of my mind, during that whole event, I was like, man, this is what it's going to feel like. It did not feel great, but we'll say that.

48:50It reminded me of how important it is for Bitcoiners to develop a strategy around quantum computing. At some point, I'm not one of the ones who's saying we got to drop everything else and focus on it right now. I think there's many things we need to be focused on. This is one of them, but it's definitely something I'm tracking more and more into your point, Kathy. AI is accelerating timelines in quantum computing, which means it should be accelerating our timelines as defenders. Yeah. And as we always say, half of the solution is understanding the problem. So the fact that so many people we're talking about quantum gave me some confidence. And we too share your point of view that this probably isn't going to happen for a decade. You know, one thing that's interesting is on the quantum side as well as just on the AI advancement side. There's one treasure chest that is just out there waiting. And that's Satoshi's coins. And so not to get spicy to close this out, but I'd be curious to hear everyone's take

49:54is like, if you see Satoshi's coins move, what's your gut reaction? The end times are upon us, I suppose. I think it increases the likelihood that something is happening with a quantum computer. Satoshi may also still be around and may finally have gotten to a point of moving them. So it's not strictly 100% certainty, like to be precise about it. I think it would be interesting to see how you put that key. He finally just got around to it. Like he had an idea. Yeah. He's a busy guy. He's probably working on a lot of different things. He just forgot about it or he found his old USB drive from 15, 18 years ago. Right? Yeah. I think it'd be interesting. It's been one of those, even before the quantum discussion was front and center, it was always one of those things of what would happen one day at Satoshi move this coin. So I just think the quantum computer conversation loads in extra context around what would happen around that. And ultimately, the Bitcoin protocol still works. And I think it's just incumbent on, there's been even over the past week, new hash-based

50:57signatures that are being tested and worked on for an eventual migration path. So I feel confident and good in the long run, but it may cause some short-term volatility. From an investor's point of view, seeing the big whales moving, I was thinking, as this quantum controversy was, surfacing was a little unsettling. Inside or selling in the equity market is a very good signal. If it's off some kind of plan and not well advertised and so forth. So I have to say it was a question in my mind. But as we were doing more and more of our AI and quantum research, I began to feel like you drew. But it was uncomfortable. If Satoshi's coins moved, I'm going to leave that to Lorenzo. Is he moving the coins just because he wants to move the coins? Is he moving the coins because he obviously knows that quantum is a threat and it's

51:58kind of time to do it. Is it a quantum attack itself that was able to move the coins? My understanding is you need to transition to new signatures. So at some point, if he's dead, the community is going to have to decide what to do with those coins. And I think there's a lot of smart people working on that. I know that several kind of solution if Bitcoin is forked. So yeah, definitely I think the market wouldn't take it very well, honestly, but I think there are solutions to this problem. It's a nice bug bounty to have though, because it will give us a good heads up and be the first warning bell potentially in that future where quantum computers are real or some other break is. Even if those coins are kind of hacked and accessed by someone, I do think in general, people underestimate what can you do after that with those coins.

52:59Right now, the ecosystem is much more mature in terms of secure. The exchanges would blacklist, same for brokers on and off-frames. I think it'd be much more difficult that maybe five or 10 years ago, which makes it harder in terms of bounty. So I think that's kind of one thing to consider without obviously the legal aspects and everything. But I do think back to the earlier part of our conversation, the willingness of core developers, at least some of them, to face this threat with AI is, I think, critically important. Because guess what? They have the domain expertise. And to that point, I've directly worked with over the past month several core devs in finding things and having them go through and filter and upstream and fixes. And I've been involved with Bitcoin for 13 years. Over the past month, I now have multiple commits in Bitcoin Core that were shepherded by

54:02core developers. But that originally came from my AI-based findings. So I'm very optimistic in their seriousness and understanding that there's plenty of things to fix and do for the maintenance of the project. So maybe just to close, where do we go from here? I mean, it's so funny because I'm going to say that again. And then life is just going to just go, I know who knows, especially with all these new models coming out with Bitcoin. I'll say for one of the my, or maybe a better question is like, what's one of your biggest takeaways? At least for me, is the Bitcoin community is so amazing. It's so resilient. And I think it's now going to even be more focused on the core development and what are the things that we need to focus on versus some of the noisy stuff that may have just taken up a lot of oxygen. We believe that AI is highly deflationary and all of the technologies out there, highly deflationary. And because costs are collapsing in AI, we think that real GDP growth is going to accelerate

55:07far beyond where people believe it will accelerate. It's been roughly 3% for the last 125 years since the Industrial Revolution. And we think there will be a step function here to 7, 8%. Nobody believes this, but that it will be accompanied by much lower than expected inflation. Why am I mentioning this here for a couple of reasons? If we're right, then interest rates are going to go up not as much as they might, if inflation were a problem, but just to adjust to the market. It's the market working. Interest rates go up as your growth rate accelerates. We think that's going to cause a lot of problems and create counterparty risk. And that this is going to be another... Most people think in terms of Bitcoin hedge against inflation, what are you saying? This deflation will be bad for Bitcoin. I don't think so. I think that the counterparty risks are lurking out there. And we're seeing some of this with private equity.

56:10They fund their companies with short-term debt. And we think short rates are going to go up more than long rates. And so we think there could be a world of hurt taking place there. And banks with exposure to that world could be in question. So I think this is Bitcoin's role as... It's not only a hedge against inflation, it's a hedge against deflation. Because the counterparty risk is not there. Anyone just want to go around with some final thoughts. You know, we can just wrap up. Maybe we'll start with Seth. I think the big thing for me is, like I mentioned throughout, this is a huge paradigm shift, but it's not as scary as it seems at first. Like the cool-cut thing was a wake-up call, but it's telling companies to do stuff they should have been doing all along. And there is going to be a lot of good security hardening that comes with it. But ultimately at the end of the day, these tools help defenders more than they help attackers. And I think that's a really important piece to keep in mind so that it becomes less of a... The end times are here and more of a... How do we adjust to this paradigm shift when we're thinking about this throughout?

57:14And the rate of innovation, the rate of change, the improvements in local AI and the things that you can do on-premises is just immense. It is moving so quickly. And really good people are at the forefront of that. People like Rob, the rest of the Bitcoin Red Team, that are doing amazing work. And I'm sure a lot of people in all of companies on this call who are currently neck deep in AI stuff, trying to secure their environments and staying on top of it. So it is an exciting time. It is a bit scary. But I think in the end it's going to be really good. Lorenzo, do you want to jump in? We had a huge symmetry between attackers and defenders in the space. Pre-AI and I think between what Seth said and Kathy, with the cost decline, and these tools being much more available and just better frameworks, I think it's going to be net good for defenders. And I think the space, probably in the year or two, is going to be in a much, much better place if I had to guess. Drew, how about yourself? Man, I think I've still just caught in metaphors and analogies.

58:15Just thinking, what does it actually mean if computers become more like living things? It doesn't just mean that we talk to them. It means that they evolve. It means that they defend themselves. It means that heterogeneity is very powerful. And monocultures are very risky. We said at a work the other week that like, AI can bring about a world that Bitcoin was created to prevent from happening. And I'm so glad that Bitcoin exists because AI needs Bitcoin, just as much as Bitcoin needs AI. These are part of a larger historical trend that will only be understood hundreds of years from now that, during this century, a whole bunch of stuff started to go the other direction, which is to say the pendulum that had been swinging for a thousand years or more towards centralizing power and control, that started to break and started to go in the other direction. And I think I'm very excited to be at the forefront of that. Rather than when I will close with Kathy. It's been a very busy month, and I understand a lot of people are concerned or scared. I'm ultimately an optimist. I just inherently am an optimist.

59:16And I think the leveraging of this technology is actually one that empowers individuals with ambitions in a greater way to enable freedom that was not accessible before. The security landscape is going to probably have some volatility and all of software, but we're going to come out stronger on the other side because of it. I think it's such an exciting time to you can just do things. Like I was not a security researcher three weeks ago. And now I'm helping out a lot of people working on these things. Like I'm now have contributed software to the Bitcoin project, which is something I've always wanted to do. I think it's an incredible empowering time for people with ambition to go out and seize the opportunities that are there. And I just encourage everyone to think about your day-to-day life and things you want to achieve and do. I think it's like this technology is just such a liberating force. And I'm very excited to be living in this moment in time. Every time we do a Bitcoin brainstorm, my confidence increases in Bitcoin, right?

1:00:17And you all are working in it, rolling up your shirt sleeves and all of that. And you are helping us, certainly on the investment side, maintain our confidence, but maybe even increase it. The reason I went on that little riff is, if we see deflation, there may be some people out there who say, oh, what then? Why do we need Bitcoin or what have you? But I think that we haven't known, not since the Great Depression, what deflation can do. And this notion of counterparty risk, I realize there are some centralizing forces or vectors, but there is a lot that, I mean, individuals in this movement still have a very powerful role to play. And I think that's what we've been talking about here. And thank you, Rob Seth, Drew for all you do.

1:01:19And of course, the great Rod Rudy. Kathy, you're the best. And honestly, every month I look forward to this, because selfishly, and I think Kathy would say this, we get to ask questions and then kind of opine and bring some of the best people in the space. And I'm really thankful for you guys. I do think, as a closing thought, I think, like custody of Bitcoin is going to even be more robust, stronger, and better for the individual as well as businesses. And I think there's going to be so much innovation happening, whether it's on the distribution side to actually access these signing devices, as well as more easy use of multi-sign, multi-vendor solutions out there, because you're just going to want to learn how to preserve and hold this amazing bearer asset. So again, I cannot thank you guys enough. I mean, I don't even talk about the price, but I think the price was like way down like last episode, just a few weeks ago. And now it's been up like, I don't know how many percent,

1:02:19but it's all good in the Bitcoin world. So again, thank you everybody for joining us. We'll see you next time. ARK believes that the information presented is accurate and was obtained from sources that ARK believes to be reliable. However, ARK does not guarantee the accuracy or completeness of any information. And such information may be subject to change without notice from ARK. Historical results are not indications of future results. Certain of the statements contained in this podcast may be statements of future expectations and other forward-looking statements that are based on ARK's current views and assumptions and involve unknown risks and uncertainties that could cause actual results, performance, or events that differ materially from those expressed or implied in such statements. Bitcoin is a relatively new asset glass and the market for Bitcoin is subject to rapid changes in uncertainty. Bitcoin is largely unregulated and Bitcoin investments may be more susceptible to fraud and manipulation than more regulated investments. Bitcoin is subject to unique and substantial risks, including significant price volatility, lack of liquidity and theft. Bitcoin is subject to rapid price links, including as a result of actions and statements by influencers and the media,

1:03:21changes in the supply and demand of Bitcoin and other factors. There is no assurance that Bitcoin will maintain its value over the long term. ARK strongly encourages any investor considering an investment in Bitcoin or any other digital asset to consult with a financial professional before investing. All statements made regarding Bitcoin are strictly believes in points of view held by ARK and or podcast guests and are not recommendations by ARK and buy cell or hold Bitcoin. Historical results are not indications of future results.