The Future of Digital Identity: How Suremark Makes Verifying People Seamless at Scale

Joe: Well, Scott, you know, this is kind of like our part two because the first time that we were recording, I mean, we we were having a fantastic conversation, and then I think it was like my internet just decided to drop out of nowhere. And W. Scott Stornetta: Right. Joe: good ol good old Xfinity here in the Chicagoland area, they have just been having massive amounts of issues. My my neighbor actually installed a backup internet link with Starlink, and I am extremely tempted to do the same thing. W. Scott Stornetta: You know, I think that's becoming even more common. because you know, there's a certain simplicity and elegance with Star Starlink that says, well, as long as the earth is still where it is and Newton's law of gravitation still applies, you can pretty much expect Starlink to be up. Joe: Right. And it I mean like it it's it's such a cool piece of technology, in my opinion, you know. like I'm I'm going on a road trip here pretty soon with my family, just for a week. And I'm tempted to get Starlink just for the road trip, you know, because it provides a lot of advantages to just have a dependable internet link, you know, wherever we are. W. Scott Stornetta: You know, I I Joe: Yeah. W. Scott Stornetta: fly a lot on an airline which has Starlink Joe: Mm. W. Scott Stornetta: built into it, and they run a little Starlink ad at the start when they're also telling you how to fasten your seatbelt in case you Joe: Yeah. W. Scott Stornetta: haven't been in a car in the last 60 years. and so many of the shots they show are, you know, on the back of a camel, attached to a backpacker's backpack, inside the kayak. with you. Joe: Yeah. W. Scott Stornetta: And you know, it works. Joe: Right. Yeah. That that's it's really interesting you bring that up. I in in February I got on a flight and it it had started I didn't realize that, you know, I was just connecting to like the normal plane internet, whatever it might be. And my three-year-old was giving my wife a lot of problems, like like just participating in dinner and stuff like that, you know? And and so my wife Face timed me so my daughter could see me. And I mean she FaceTimed me and I'm on the plane midair. And and it came through like perfect perfect quality. You never would have known that I was flying on a plane at that point in time, you know? And of course, you know, my daughter sees me and she's like, hi. I was like, Hey, you gotta be good for mommy, like you know, and we had a full, you know, 15 minute conversation, hung up and she was good the rest of the night, but like That just like blew me away. I like, my God, I can't believe that this just happened. Like we're living we're living in such a great time, as well as like a very scary time with AI, I think, in my opinion. W. Scott Stornetta: Well, again, I I prefer to say we are living in a great time that has a little bit of some roughness around the edges. Because Joe: Some challenges for sure. W. Scott Stornetta: obviously, you know, I you know, my my AI experience goes back many decades. I I think I may have mentioned in the earlier part one that mine was the first neural networks PhD that Stanford issued, even though it was done through the physics department. So I've been an AI fan for for decades. But yeah, there's a little roughness around the edges, but that's what Shermark is trying to ameliorate. So bring it on. Joe: You know, I I have like I have two questions there. And so the first one is what's a neural network? Like just explain it, you know, to the to the layman, including myself, because sometimes I feel like I understand it and other times I feel like I don't. and then the other part of it is what is that like when you're researching something that is just brand new? You know, like And I'll I'll kind of compare it to my own dissertation, right? My topic is fairly unique. No, from what I can tell, no one else has tried to do, or at least no one has published them researching how to do what I'm doing with with the frameworks and protocols that I'm doing. They've done it in other aspects, other, you know, ways of doing it and whatnot, but applying it to a sensitive infrastructure like satellites, it's never really been done before. Right. And so like that's maybe the uniqueness of my research. But you know, take me back to when you're deciding I'm gonna do this thing on neural networks. What what does that look like? You know, how long did it take you to get through that process? Because I'm sure I'm sure that must have been a long time just to get through it if you're doing something so new. W. Scott Stornetta: Well, a couple of comments. First of all, it very much felt like the Wild West. these new possibilities and ideas were just kind of exploding like on a weekly basis. And I say this in the following context. Remember, this was a kind of very first surge around these AI related techniques. And there were many years of AI winter. intervening before we got to the stage that we're at today. And so another way just to put it in context before I explain neural networks is that what I did research on was with roughly on the order of a thousand weights or parameters, where today's large language models are using as many as a trillion weights or parameters. And so and yet the basic mechanism and paradigm is is still the same. And so to try to answer both sides of your question, what was it like in those days and what's a neural network? What it was like was there were some exciting work going on. And you know, my my PhD advisor, he he really I I think his finest quality was teaching me that the key is not solving problems, but having a taste for the right problem to solve. And, you know, whether my judgment was good or bad or just lucky, I just had the sense that this is the right kind of problem to solve. This is going to go somewhere. This has legs. And so I never really thought much about well, is this gonna be hard or easy or well defined or undefined? I just said, I, you know, I found what I want, I'm gonna go after it as hard as I can. I'm gonna learn what I need to learn, and I'm just gonna jump in. Now, what's a neural network? Well, I think you know, neuro maybe has Sent us off a wrong road because it it starts to conjure up, you know, brains and people, and then when these neural networks start talking, you think, is it alive and whatnot. And so a neural network is a computer algorithm, right? It takes some inputs, it produces outputs, and then it compares It's produced outputs to the training material and what the correct answer is. And so you give it inputs, it flows through the neural network, it produces outputs, and then it gets compared to the training. Here's the right outputs. And a neural network is able to perform a very simple algorithm repeatedly that says. Well, let's see how far off I was. And then let's push backwards through the network and see if I tweak these different parameters, whether the next time that kind of input shows up, it'll be a little closer to what I was trained on. And then the broader hypothesis is: if I've got a ton of training data and I gradually, you know, compare each. Each output and then try to tweak the internal parameters, over time I will gradually learn the generalizations about how to respond to new inputs and produce the correct outputs. That's in a nutshell what a neural network does. And of course, as you move from doing that with a thousand parameters to train up to Tens of billions and even a trillion parameters to train, the network can handle more and more complex situations with better and better fidelity. And that's what produces, and obviously this is a huge simplification. You know, that's why we can chat with Claude now, because it's been trained on what it's like to do that. And it knows enough about the generalization of, in the case of a large language model on text, you know, what English text is like and what people talk about. The fact that it's, in a sense, been trained on, or to put it differently, read every book in the world or every publicly available email and text exchange, means it has a pretty rich idea of how people. you know, interact and it it has a lot it can draw on to produce useful dialogue with us. Now I'm leaving separate, you know, neural networks also are the ones that are creating the simulated images and even the short form videos. But in a sense it's all based on that basic idea of you take massive amounts of data that is Sort of trained or tagged is a word that they use in that space. And you test out your machine and you see what it gives as an output, and then you say, Well, this is this would have been a better output. And it says, Hmm, okay, I'll propagate that information backwards through the system. That's why they call it back propagation. I'll tweak my weights or my parameters a little bit and then I'll have learned a little bit better the next time something like that shows up. That's a neural network. Joe: So I I I have I'm not quite sure how to ask it, but you know, all these different models you know, like you said, right? They're they're trained on all publicly available information, books and websites and you know, if it's a text that's publicly available in email, whatever it might be. You know, I I I remember a couple of years ago, maybe last year even, Sam Altman mentioned that they had already trained OpenAI on everything, ever everything that was publicly available. W. Scott Stornetta: Right. Joe: And so at that point, you're essentially training the model on the users. You need the users to interact with the model so that the model can become, you know, better, more efficient, whatever it might be. Do you think that there's There's like a slowdown in the evolution of it because of that. Because and I ask because you know, I'm thinking back just six months now, and can I really say that there was anything ver like extremely innovative with any of the models except for, you know, maybe Claude Mythos, right? That that not really came out, but came out and is causing havoc in cybersecurity teams and whatnot. But I mean, I use these models every single day, literally every single day. I don't even use Google anymore. And I couldn't tell you it's more efficient. Yeah. It's W. Scott Stornetta: Right. It's become my way of searching. Instead of Google search, I just ask the question. Well, it's more targeted, it's more pointed, and it goes and grabs all of the sources that you would have gotten with a Google search, but then it synthesizes it to respond to your very specific question. Joe: Right. I I guess, you know W. Scott Stornetta: is there a slowdown, I guess, is part of your question. Or is there Joe: Right. W. Scott Stornetta: anything new under the sun in the last six months or so? Joe: Well, there's a s there's the slowdown in innovation, at least it seems, from my perspective. But is there something that can be done to increase the speed of innovation at this point? Or are we kind of just stuck, you know, where we are right now until we get enough users using the platform, right? Because I'm thinking from a security perspective, that starts turning into like a weird scenario where you need your customers more than They may need you at times, W. Scott Stornetta: They need you. Joe: you know? W. Scott Stornetta: Right. Okay, well, let me answer it in this way. More and more training data makes the systems better. And part of what you're saying is they've kind of tapped out all the training data until maybe we get going forward, you know, more user input. But it's not the only dimension along which things can be approved. Okay. One of the dimensions, of course, is simply more and more brute force computing power, more and more parameters. And that really is an ongoing arms race as as we see, you know, energy use really skyrocketing. And there's a rich discussion there, but let me you know cover a couple of categories. The next is more efficient algorithms. Okay, same hard, you know. So hold the training data fixed, hold the amount of hardware you can throw at it fixed, but come up with a more efficient algorithm. Okay, something that gets to the heart of the matter faster. And there are both small variations going on there, but there are also big bets that are being placed, and in particular. There's something called world models that have gained a lot of interest lately. maybe the chief protagonist there is a very distinguished researcher named Jan Lacun, who used to be head of AI research for Facebook Become Meta, and now has formed his own startup in France. And it's the kind of thing where he just snapped his fingers and raised a billion dollars. And he's a big advocate of so-called world models that more deeply understand what's actually going on as opposed to just matching inputs to outputs. And, you know, that has an acronym that goes with a joint embedded predictive architecture, blah, blah, blah. But I think the short answer is. One, yes, the training data is kind of tapping out. and therefore companies that are really in the second tier right now, like XAI, you know, I sort of put OpenAI, Anthropic, and maybe Google, you know, they're in the in the top tier. But XAI does have one advantage in that it has access to all of the Twitter. Joe: Mm-hmm. W. Scott Stornetta: material. So it's constantly getting new material. But it's not just training data, it's also compute power. It's also improved algorithms, whether slightly tweak improved or completely moving away from the so called LLM, the large language model, into a very different kind of model. So all of those are avenues for progress. But it still has that interesting point that you make. It sort of needs you more than you need it in some respects. I think the other thing, though, if we're trying to have a candid and an interesting engagement, is my experience is different than the one you've just described. I see meaningful improvement over the last six to twelve months. And, you know, one way you can say, well, You know, if this is the base and this is perfection and it's here and it's moved to here, you may think, well, that doesn't seem like much of a growth. But if it's an improvement from this much to that much, it doesn't seem large. But if you're reducing the error from this much to that much, well, it is kind of large. And so, you know, this progression from it being kind of an interesting toy and then a tool. To really being something you can engage with, it's all in those last few percent. And so just to give you one concrete example, I've been working on a particular concept for a long time and finally decided I really needed to start writing a paper about. And so there's a lot of thinking that gone into it long before I I put pen to paper. Okay. But being able to dialogue with, in my case, Claude really accelerates the process of putting the concepts out on paper, getting the logical structure right. you know, it's it's like you have a what would you call it if you Like a boxing partner. What am I a sparring partner? It's like Joe: Sparring partner, yeah. W. Scott Stornetta: having a sparring partner. And I took a flight from Washington, D.C. to San Diego on Tuesday morning this week, or I guess it lasted throughout the morning, a six-hour flight. I started the paper at the start, and by the end of the flight, I had a rough draft of something that if I'd been working by myself or even with a collaborator, It would have been at least days, if not a few weeks, before I'd reach that point. Now, is it the same thing I would have reached after those few weeks? Well, maybe not exactly, but it's still my thinking. But having the sounding board, and that's something that if you make just you know, if you improve from 93 to 94 percent, well. That may seem small, but if you're ch changing the error from seven down to six, you know, that's bigger. So I've been impressed with what I've seen in the last six to twelve months. Joe: Hmm. Yeah, it's interesting. It's a really good explanation or way of of looking at it, I guess. and you know, my my own experience, right? Like I I have terrible writer's block. I I mean, like it is bad. This you know, thinking back, like it goes all the way back to like sixth grade, the papers do the next day. I'm just getting started. You know, I'm staying up all night. It's 40 pages long. As a as like a sixth or seventh grader, like that's super long, you know. and same thing with my dissertation, right? Like, I'm sitting here and I can't put together a sentence, because I'm just overthinking things. And I have this idea, I have this topic, but I I need the structure. I just need something to get me started, you know, and and you know, get me going with it, right? Because like I don't know, I I just get caught in my head too much. And, you know, my model of choice is actually Grok because at the time I was trying OpenAI and Chat GPT was just completely useless to me. It was hallucinating full blown research papers and, you know, hallucinating like dynamics of satellites and just all this stuff. And I couldn't get anything reliable out of it at the time when I was using it. And so I switched over to Grok. Grocks seem to be a whole lot more reliable to me. And so, you know, it's it's definitely assisted me in like getting my idea together. Here's the papers that you should be reviewing. You know, here's a synopsis of what makes sense. I checked it against your topic. You know, here's what here's what like this introductory paragraph would look like for this section, right? Like all that sort of stuff because you You know what a lot of people don't understand is like when you're writing a dissertation, there's a thousand templates out there, right? And W. Scott Stornetta: Yep. Joe: unless your school is telling you, this is our template, you have to use our template. If they don't tell you that, like for someone like myself, that's like an impossible feat to to surmount, you know, like it's not gonna happen. W. Scott Stornetta: Well, you know, a comment or two about writer's block, because I also wrote a dissertation and I've also had to write lots of papers. And to me, writer's block is often viewed as like some kind of weakness in someone. But I just think it's part and parcel of the creation process. And we try to get unstuck. I hear that, but part of it is. It's only in the process of pushing ourselves to get unstuck that we finally realize what we've been trying to say. In other words, it's not like the idea is already there and it and the problem is just getting it out. It's that the process of trying to get it out is what gets the idea fully formed in there. we are I have this sort of illusion that I already know what I want to say, but I only I find out that I really learn what I want to say in the efforts to say it. And so, you know, how to push through the writer's block, how to push through that stage, is to me the real question. And all I can say is the reason that so many researchers have whiteboards sitting behind them. Is that they, you know, kind of game things out, and it's even better if there's a second person, you know, I hold the red marker, they hold the blue marker, and we kind of go back and forth on this, and then then you say, now I see. So this really becomes the important point. And then you erase all that and you say, So what we're really trying to say is this, then this, then this. And so that clarity comes as you think through it. Joe: Mm-hmm. W. Scott Stornetta: And so to me, the idea that these current models serve effectively as a sort of search on steroids in the one case, but also a sounding board that I can kind of play what I've just described up at a whiteboard with another person, I can kind of engage in that same sort of activity to help. Get me through these things. And then the other thing, of course, is that we are better at recognizing something that we like than creating something that we like. And so if you can get it to do some of the writing tasks, and then you read it, you say, no, that's garbage. But yeah, this is this is getting closer there. And then you can build on that. Momentum as well. So I think they're great tools. Just great. I guess Joe: Hm. Yeah. W. Scott Stornetta: the problem, of course, is that, and I apologize if I'm talking too much, but the problem is, and I see this in the dimension where I assign students to do these kinds of tasks, is that they let the LLM take over, and they're no longer taking responsibility for the final outcome. And that's a separate issue to deal with. But my feeling is when I put my name on something, I'm taking responsibility for it. Whatever tools I used in the preparation, whether that's just a word processor or you know, a full-blown LLM that's helping co-compose it, it doesn't matter. What matters is when I finally put the thing out, it's saying This is me. This is what Scott Stornetta has to say and think, and I take full responsibility. Joe: Yeah, I mean it's it's so easy to have it just write, you know, the paper, the section, whatever it might be for you, and think, okay, this is attached up to this huge computing power, right? And it it definitely knows more than me. Why am I going to adjust it or anything like that? Like that's like that subconscious, you know, rabbit hole that your mind goes down immediately. but There's been countless times where, you know, it gives me a paragraph or a section or whatever. And I mean, I I still go through every single thing that it gives me because in in my opinion, when I'm defending it, I need to know it inside and out. Like it can't be like a well Grok told me this. It has to be a no, it is this because of X, Y, and Z. And I went down this path and that's what I found. you know, more of that, right? So I'm still going through and even like, you know, checking every sentence and the structure and everything else like that. so even though even though the tool gets me around the writer's block, it's not writing the paper, so to speak. And W. Scott Stornetta: No. Joe: that that's the key differentiator, right? so, you know, to okay, we have enough time. So I I wanted to ask you Right. And this is something that I recently thought about. So and and you mentioned it yourself. Grok has kind of slipped back and Elon Musk has even mentioned it where he where Grok has fallen behind the competition in some areas. But they do have the compute power. And Anthropic is renting compute power from the supercluster that Elon Musk put together. So my my thought is that they have what everyone else is already basically in search of, trying to build in some way, shape, or form. Once they get the right people internally together to actually start utilizing the rest of that compute power that they're that no one is touching. Do you think that that would accelerate them beyond I mean, maybe not beyond everyone else, but kinda level the playing field? Okay. That's what I was thinking too. W. Scott Stornetta: I think it's actually a very real possibility. Okay. And I'm not talking I'm not talking just about the arc of bringing up these terrestrial systems, but you know, they're aiming higher than that, right? Literally, right? They want to put you know orbital compute to work. And they're talking about compute at the terawatt level, which is You know, here the constraining factor is getting enough bringing enough energy resources online and the permitting and the zoning process here on Earth. And Musk is saying, let's just rise above it all. We're gonna just, you know, ship so much compute and solar energy power up into space that you know their aim is to be an order of magnitude and more outpace the current Tier one players. So do I think they can catch up? I absolutely think it should be viewed very seriously. Do I think they can even exceed the situation? I think that's also a very real possibility. Joe: Yeah. Because it it seems like they can scale faster than anyone else. Y you know it W. Scott Stornetta: Well, I think that's sort of a theme for almost everything that Musk has done across a whole variety. They scale faster than anyone else. It's like one Joe: Yeah. Yeah. W. Scott Stornetta: of the core assets. Joe: Yeah, absolutely. You know, I I've I've read the two like the two primary books on on Musk's, you know, life overall, right? And I mean like that's a prominent theme is where he does a really good job of cutting the fat in a process in a company with headcount or whatever it might be. You look at you know, when he took over Twitter, what he laid off something like eighty percent of the company within thirty days or something like that. W. Scott Stornetta: Yeah, we went from eight thousand or something down to two thousand roughly. Joe: Right. You know, and w I didn't notice any outages or slowdowns really. In fact, a over a little bit of time, it actually got better. It got more efficient. It it seemed to just work better overall. and mind you, I didn't use Twitter at all when before Elon Musk took it over. I actually I I decided that I basically Had to get on Twitter slash X at the same time when he took it over, mostly because of the podcast. And, you know, it's like, okay, I'm doing this thing. I gotta take it a little bit more serious. Let's, you know, get a social media presence, right? W. Scott Stornetta: Yeah. Joe: But it it's a really it's a controversial way of doing it, right? Because it causes a riff with a lot of people, you know. But the way that he does it though, I think is probably the most ethical out of a lot of companies. He gives very advantageous severance packages. he really gives you every opportunity to, you know, like say that you did great work at that company. You just weren't a fit. They didn't need the role, whatever it might be. so from what I un from what I understand, you know, like that it it usually, yes, it does suck, but it, you know, it's it's a little bit better than some other situations. W. Scott Stornetta: Well, I think it is, and I think it's a necessary, quite frankly. If you Joe: It it absolutely is. W. Scott Stornetta: really want to talk about growing rapidly with scale and achieving scale, you have to be pretty I mean, it reminds me a little bit, and I've dealt with this issue, you know, is he a nice CEO? Is he a bad CEO? Am I being a good CEO? Again, understand, you know, he manages trillion dollar companies. I'm trying to get Positive cash flow. So I'm not trying to put myself in the same league, but I deal with the same kinds of issues. And the thing that has really struck me as a guiding metaphor is a response to something I've heard several times. When people are on a, you know, people are employees at a company and they really feel good about the team and whatnot. And they say, Wow, this just feels like family. Okay. Because of the close relationships and a host of good things. And I think there's a strong corrective that needs to go there. A business is not like family. A business, and I the analogy that I heard from someone else that I think is far better. Think of a business, it's like a winning baseball team. Okay? Everybody there likes to play baseball. It's fun. You have good relationships, you enjoy what you're doing, but The goal is a win-loss record as a baseball team. And if that requires, you know, moving one person to a different place or hiring a new person and sending the old person off in a trade, no one no one criticizes baseball managers for making those kinds of decisions. They have a job to do, they're trying to execute as best they can. I really think that's a much better. starting point for people to think about what businesses do. They're trying to build a winning baseball team. And so, you know, it's not a family. Your family is like a family, not the company that you work with. Joe: Right. Yeah, that's a really important distinction, you know, because then people start taking it more personal when the business makes business decisions and whatnot. And you know that that's a thing, right? It's it's like, is it is it uncomfortable? Yes, it's not a pleasant thing. I'm sure your manager hated talking to you and you know, letting you go and whatnot, right? But at the end of the day, the business is exit in existence to make money and support its shareholders if it has shareholders or investors, right? and if it's not doing that, it has to do whatever it has to do to make more money. Including if the CEO needs to be removed and put in a new one, they do it, you know. W. Scott Stornetta: Right. And even going to the make more money, if I could just nuance, you know, offer a nuance Joe: Mm-hmm. W. Scott Stornetta: on that, it's the purpose is to achieve the mission of the company. And what people often fail to understand is it's very difficult to achieve the mission of a company if it's not profitable. So profit, yes, you have obligations to the shareholders, but profit. Isn't really what drives entrepreneurs. What drives them is they want to achieve the mission. But it's hugely more difficult to achieve the mission if you're not profitable. Because what you want is a self-sustaining business. And so you've got to make the set of decisions that allow you to move towards the vision that you want to achieve. You know, the whole changing the world. dynamic. And so to me, the dollars and cents really for the entrepreneurs that I know most closely and that I admire the most, you know, it's really not about the money. The money is a means to an end. It's achieving the mission. It's changing the world. It's having a positive impact. And so when I hear, you know, the rejoinders about, well, you know, He's a billionaire. Well, it's like you've put him in a new subspecies or something, as if, you know, he's just It's not what drives a lot of people. I mean, there are some that are just saying, well, you know, my mission in life is to figure out something I can do to become a billionaire. But that is a very small subset of entrepreneurs. It's, I want to bring this to pass. I want to see a world where X and Then it just sort of follows that if you want to do that at scale, that there have to be meaningful revenues, you need to be profitable, and well, there you are. You own ten percent of that company, and so, you know, you wind up in that space, but that wasn't the goal. Joe: Right. Yeah. No, that's a really good point. And I I I feel like people always miss the point, you know, when they start saying like, he's a billionaire, they're a trillion trillionaire, whatever it might be. Right. They they try to like, you know, separate him into a new species, like you said, right? And you have to look at it like how many how many employees do these people employ, right? W. Scott Stornetta: Absolutely. Joe: In their in their position. that support families, that support children, that, you know, benefit other organizations, right? Like they W. Scott Stornetta: That's right. You're talking you know, you're talking about exactly Joe: you know, like the impact is absolutely worth, you know, that that person being worth X amount of billions because he did the work and the marketplace is rewarding him for that work based on the product and the mission that the company is delivering on. I mean that's It's literally as simple as that. And I know there's a lot of people that I probably just lost like two thousand subscribers just by saying that, right? But that's a simple fact. You know, and and that's how the world works to some degree, at least in some, you know, in some situations, I guess, right? But W. Scott Stornetta: Right. And the other thing I'd say is, okay, if you want to judge the person, then judge them not on how much money they have, but what they do with the money. Okay. you know, I'm such a big fan of Andrew Carnegie, Joe: Mm-hmm. W. Scott Stornetta: you know, of you know, real you know a century and a half ago, you know, he had this, you know, essay he wrote where he said, you know. I had huge sums of money come to me, but my goal was to end up with zero at the end of my life because I needed it to put it all back into the system so that you know people would prosper. You don't Joe: Yeah, he gave it all away. W. Scott Stornetta: yeah, so you don't you don't say, well, he was a billionaire, you know, or for his, you know, a time adjusted billionaire. He he wanted to accomplish something in the building of his businesses. He wanted to accomplish something in the distribution of the profits of his businesses. That's how I, you know, that's how people should be judged. Joe: Right. With with AI becoming More at the center of everyone's life now. You know, I I I feel like a solution like Shermark really benefits the people. You W. Scott Stornetta: Mm-hmm. Joe: know, because and you know, you can you you know obviously, you know, speak to it a little bit more because this is again part two and we haven't mentioned Shermark just yet. But having the ability to verify who I'll actually give you a real world example. I this happened literally this morning where I I have a friend from high school. We stay in touch every once in a while. He messaged me on a platform that he normally messages me from, but he asked me a really weird question. Hey, I have these encrypted files. I need you to unencrypt them. I don't know how to do it. All right, well, I'm not going to be able to help you, but two, that's a really weird message to send me, right? And then he texts me from another number that I would have expected him to text from. So now I have two correlations, same person, right? A normal person would say it has to be him, I guess, right? Maybe a maybe someone that isn't very questioning of things, right? But I'm in security, I question everything. W. Scott Stornetta: Yep. Joe: And so W. Scott Stornetta: And and the n our nation is grateful to you for your service. Joe: Yeah, I don't I don't know if I serve, but you know, I I I just asked him some qualifying questions. You know, where did we first meet? What year did we meet? What what was the exact location that we met? Something that only me and him would actually know. And, you know, of course, like they couldn't get it. But without a solution like Shermark deployed, I have no other way of identifying if that's a fraud because it literally came from two places that we have talked before continuously and they're two different platforms, so to speak, right? And so it's like, all right, well, how would a normal person verify this sort of stuff? W. Scott Stornetta: Right. And I think the one way to think about Shermark is we have this in intuitive notion of, you know, well, we have a shared experience. We have a kind of shared secret. And it goes back to the example that you gave all the way at the beginning of the first episode, where you were able to prevent something like a forty million dollar mistake on the part of you and the CFO because you had agreed on a kind of shared secret. Joe: Yeah. W. Scott Stornetta: just as a final check. And the point that Shermark is trying to make is, yeah, that's the right idea. The problem is to try to do that at scale, if you needed a shared secret between you and every other person in the world, it would get a little taxing. Joe: Yeah. Just trying to store W. Scott Stornetta: But with with our peer-to-peer system for establishing credible identity through you you prove who you are through showing who you know, you can just boil it down to each of us just needs to keep one secret, and then we can cross-validate with everybody else. It's it it does at scale what your natural instinct is. as the best way to verify identity. Joe: Yeah, it's fascinating. That that is it's interesting and it's very timely, you know, because like that is something that we absolutely need in this modern day with deep fakes becoming, you know, more believable and more real. I have an episode coming out where, you know, this company protects you with deep fakes via like voice, you know, over the phone and whatnot. And they deep fake they deepfake my voice. And just as an example, and I mean there's There's no way that my mom would ever not know that that's not me. You know? it it's completely unfair. Like completely unfair. especially because you could just spoof a phone number, you know, like it's probably W. Scott Stornetta: Yeah. Joe: pretty easy to find my phone number, unfortunately. Like these are all things W. Scott Stornetta: Right. I mean, Joe: that you can do. W. Scott Stornetta: you know, when your wife calls and it's got her phone number on the cell phone is the incoming and it sounds like her and she says, I just got in an accident, I really need your help to immediately do this, it's like Mm. Joe: I'm not asking qualifying questions. I'm it's W. Scott Stornetta: Right. Joe: already qualified. It came from the phone number, I see the face, like W. Scott Stornetta: Right and yet it may be not your wife. yeah, so well I'll and let me let me bring it home to your podcast, which is when we finish this second of the two, I'm gonna circle back and I'm going to sign both of these podcasts. And then I'm gonna invite you to co-sign them. And so That way, when it appears on these different social media platforms, anyone that has installed the open source no-cost software, they'll be able to see right next to the you know the video recording, they'll be able to see a kind of pointer that says, you can verify that this really was Joe and Scott. And so It's gonna even come back to these podcasts that you're gonna see fake, you know, fake interviews. And to me, the way to shut it down is to lock down those pieces of content. And I I know I've used the example before, but just as the HTTP to HTTPS transition was about locking down the IP address against the assigned domain name, now we can lock down an individual piece of content that appears with an individual person. It's the same mechanism that took us from HTTP to HTTPS, but a applied on a much finer grain scale, applied to individuals and individual pieces of content as opposed to just this website. Joe: Hmm. That'll be really interesting. That'll be a good a good demonstration, you know, of of the solution. I think that'll be great. W. Scott Stornetta: Yeah. I th I think your your subscribers will get a kick out of seeing that. And it also gives you another a kind of secure channel Joe: Mm-hmm. W. Scott Stornetta: to communicate in a sense a little bit more intimately with your followers. It it reminds me a little bit of you know, I grew up in a day where we had record albums. Okay. And part of the benefit of the record album was you'd open it up and you could Play the record, but it also had liner notes, you know, little additional commentary on the different tracks and and whatnot. And this this allows us, the participants, to add sort of liner notes on the podcast. It's a kind of, if you know the X Twitter concept of community notes, but it's sort of secure community notes, community notes that aren't just crowdsourced from anonymous. as a group of people, but the actual participants get to add their own secure commentary. Joe: That's interesting. Okay. Well, Scott, you know, unfortunately we're at the end of our time here, but you know, I r I really appreciate the patience with getting part two here scheduled and W. Scott Stornetta: No, again, this has just been a lot of fun. And you know, i it's enjoyable because you're knowledgeable and you're engaging. Okay. Joe: I try. W. Scott Stornetta: And so, you know, I'm happy to do what I can help. And I also, you know, I s I still leave open that, you know, if I can be helpful on the dissertation side, you know, don't be don't be a stranger. I know what Joe: Yeah. W. Scott Stornetta: it's like that you're going through. Joe: Yeah, absolutely. I'm I'm starting to quote unquote wrap it up. So hopefully by the end of the year I'll be done, but I don't I don't know. It might be it might be next spring that I'm done. W. Scott Stornetta: All right, well good luck regardless. Joe: Yeah, yeah, absolutely. Well, you know, Scott, before before I let you go, how about you tell my audience where they can find you and if they wanted to connect with you and where they could find your company, if they wanted to, you know, learn more and probably even install the software. W. Scott Stornetta: Sure. Sure. So SureMark.digital. A lot of people ask if it's dot digital.com, but no, digital is now a top level domain. Just SureMark dot digital. Sure is in sure and certain. S-U-R-E-M A R K. And then you can follow me on LinkedIn and I'm not a huge Twitter poster, but I yeah. Happy to connect with people that, you know, wanna Have an idea or s or something, liked something in the podcast and wanted to follow up, you know, it's what we do. We're trying to help each other. Joe: Yeah, absolutely. Well, you know, everyone watching, you know, I really hope that you enjoyed this episode. I'll be sure to put all the links that Scott mentioned in the description of this episode. And for now, you know, go and check out the software. It's pretty cool. You'll see the demonstration of it, you know, on the video if you install the software. Thanks everyone. Have a great one. W. Scott Stornetta: Great. Thank you.