Cybernomics Radio
Welcome to Cybernomics, the podcast where we break down the latest enterprise innovations and challenges shaping the Information Security industry. Whether it’s AI, cloud computing, or digital transformation, we dive deep into the forces driving businesses forward.
Join host Josh Bruyning as he engages with industry experts and technology leaders to explore how businesses are leveraging technology for growth. From cutting-edge advancements to the economic impact of tech decisions, Cybernomics delivers insights that keep you ahead of the curve.
Tune in for expert analysis, compelling discussions, and a front-row seat to the future of Information Security.
Cybernomics Radio
#26 “75% of My Company was Built with AI,” Randy Blasik, Co-Founder and CTO @ComplianceAide
In this conversation, Josh Bruyning interviews Randy Blasik, the CTO and co-founder of ComplianceAide, about his journey as an AI bro and the use of generative AI in building his company. They discuss the power and potential of generative AI, the challenges of bringing a product like Compliance Aid to market, and the skepticism and misconceptions surrounding AI. Randy shares his experience of using generative AI to automate various aspects of his business and highlights the importance of embracing AI and understanding its capabilities. In this conversation, Randy Blasik discusses the challenges and benefits of using AI in compliance aid. He mentions that while hallucinations can be a problem when it comes to numbers, the hallucination rate is very low for other types of data. Building trust with AI is a challenge, as people are worried about job displacement. However, Randy shares a positive experience working with a compliance firm that was overwhelmed with business and found relief using AI. Randy's legacy is focused on helping people and providing advice to those navigating the AI landscape.
Josh Bruyning LinkdedIn
Randy Blasik LinkedIn
I mean, imagine what you can do if you can interface directly with ChatGP. I know it sounds scary to a lot of people, but imagine how powerful it is now. How powerful would it be if you can directly interact with it.
Speaker 2:It's going to be. You know it's going to be insane. We're going to start spawning agents. If you're spawning them with your mind, you're going to be able to spawn a team and they're going to start communicating with each other. I mean, we're already doing that. Now. There's so many different technologies, like Microsoft has Autogen and Autogen Studio.
Speaker 1:I am talking to Randy Blasick. Randy is the CTO and co-founder at Compliance Aid. So, randy, I've been so excited to talk to you and I'm so excited to get your story out there. You are what I've learned that the Twitterverse or I guess now you call it the X-verse calls an AI bro right.
Speaker 2:Thanks, josh man, thanks for having me. Yep, I guess I'm a self-proclaimed AI bro.
Speaker 1:Yeah, you know, okay. So for those who don't know, what is an AI, bro, and how do we get ourselves wrapped up in this whole you know thing?
Speaker 2:Dude? Good question. Okay, I'm a helpful guy. You know, I've been in the industry for so long. I've been a computer hacker since 13 years old and I'm 40. So I just try to help people. I built this generative AI-based company. I actually use generative AI to build the company. I'm probably one of the first to do it. I'm building this gen AI-based company, which is a super hard task to do. I'm out there on Twitter just trying to help people out. I'm always recommending gen Twitter, just trying to help people out. I'm always recommending Gen-AI solutions. It's Gen-AI this, gen-ai that. And people just sometimes get at me so they're like, okay, they call me AI bro. They're like, okay, ai bro. That's how I got the name, so I'm going to stick proudly to it.
Speaker 1:What is the lead up to where you are today in terms of what you're doing with generative AI? How does a hacker go from your background into the world of generative AI?
Speaker 2:Okay, good question. I started young. I wanted to be a programmer at first, back in 1994, there wasn't a lot of formal education, you know. So I actually forced myself to hack, you know. Hack so that I could steal software that would help me learn to be a programmer. And just in doing that I just uncovered a lot of different technologies along the way, you know. I mean, you know, I stumbled back then on some really early AI stuff in, you know, in GPUs I did a little download withabbled with, from a hacking perspective, stealing audio. At one point I think I had every song in Western America in 1994. Started across a couple hundred AOL email boxes.
Speaker 1:And the statute of limitations has expired on that Allegedly.
Speaker 2:I allegedly may or may not be right, oh come on, you're in good company okay.
Speaker 2:So I was doing those things and I just stumbled across all these technologies and I just I just checked with it. Man, like I followed like a lot of the gpu stuff with ai was like really honestly, like they were doing, they were doing some stuff really early on and I kind of just lost touch with it. I got until I got I worked for like a Fortune 500 company there for a little while and I actually didn't like do a job and then, you know, I always kept up with hacking offensive hacking, defensive hacking. I was global cyber security. Compliance was like a cool thing for me. I'm such a nerd Like I like seeing the compliance side on the global side and then from like the offensive hacking side, I like kind of going back and forth.
Speaker 2:And then you use Income's generative AI right and so obviously what I just said is probably a lot People are probably barfing in their mouths. They're probably like what is this guy? He's like a freak right. And so this generative AI stuff kind of helped me. When it came out Gosh, this is before ChatGPT I was starting to see I could really automate a lot of things with it. My little hacker brain, my offensive hacker brain, I could automate some cool things, and then on the defensive side I could automate some things, and then the generative AI stuff just made me do it a lot faster and a lot more accurate, and that's kind of what made me really go right all into it. I guess I don't know if that helps or not.
Speaker 1:Yeah, thank you. This ties into the hook that really made me interested in your story, which was and you know this is not about the product, necessarily, but you can't really separate the product from the co-founder, from the creator, but you can't really separate the product from the co-founder, from the creator, and in this case, the creator is not, from what I understand, 100% human. The creator of this company is, you said, 75% AI. Can you explain that a little bit?
Speaker 2:Yes, one of the cool things about me being an hacker and automating things and using this generative ai stuff I was like you know what? You know I? I want to use generative ai to sort of be my fingers right. I I was like really forward looking with this company and I was. I wanted to look really far in the future you know two, three, four, five years and I was like how, how are companies in the future going to be built? And I think there's people out there like Sam Altling, for example.
Speaker 2:He's famously said they think they're going to be like a one person's billion dollar company in the next 12 to 24 months, and that's because one person's going to figure out how to build a team of AI-based agents that are basically working on behalf of that human Back in gosh. Maybe November of 2023, when I really started building the compliance aid I took that concept to heart and I started building AI agents to write code for me, to build my business strategy, to write me support documents on how to set stuff up. You know how to recommend technologies. You know I basically use generative AI agents to be my company essentially, and so it wound up being that you know 75% of the company code, tech stack strategy images, some of the support videos and marketing material. All that stuff is all basically generative, ai, recommended or built did you feel a specific call to do this?
Speaker 1:did you feel like you were the only person on the planet?
Speaker 2:oh my god yeah, you need to do this.
Speaker 1:Yeah, yeah, strangely, I felt like yes yeah, if it wasn you, it wouldn't be anybody. You know, and I've heard that from so many friends who are co-founders and technologists, and I think Mark Cuban also famously said you know, if not me, then who? So what was that call like? What was that moment when you said I have to do this, I've got no choice. This is something that needs to happen.
Speaker 2:I have to do this. I've got no choice. This is something that needs to happen.
Speaker 2:I was answering the request for proposal on some bank across the globe. It was a long RFP and I was like I knew ChatGPT 3.5 had just come out, just like that day, I think, and it was like that afternoon and I was so interested in it already, I was so drawn to it already I started using it to answer some of those SOC requirements and like it really knew the requirements and it was doing almost as good of a job as me and it was like towards the end of the day so I was tired, so it was doing such a good job and I was like this is kind of a big deal because you know, everybody knows nobody really wants to do all this compliance stuff, you know, and I was just like, oh my god, this is what everybody needs. And so right in that moment I was like we need to do this. This is when I need to, like, leave my job and get right into the compliant building, building the compliance aid.
Speaker 1:What were some of the other barriers that made you maybe second guess doing this, and what are some of the things that you had to give up in? Order to pursue this journey.
Speaker 2:Okay. So it's not an easy journey. Like you really genuinely have to commit a hundred percent to it, like every minute of the day you have to commit to to build in, you know, a company to really be successful at it. Every minute of the day you have to build a company to really be successful at it. I had to give up. I'm working seven days a week, vacations, holidays Thankfully. You have to have a good relationship with your stylist. I think you have to think about your kids where they are in their life. You have to take on a lot of roles job roles, thankfully. I had done those roles, um, you know, previously in previous careers and, you know, and my family really supported me really well throughout throughout the process, um, you know, so they were super supportive, um, but, but it's, it's a long journey, for sure, scary.
Speaker 2:Yeah to like use as a as a significant resource, strangely gave me an overwhelming sense of confidence in that. That, um, that's a strange thing to say aloud, but that's like the truth it's powerful.
Speaker 1:I mean, I've I've heard elon musk say that, and I don't think that he's trying to be cute either. I think he's. He's absolutely correct that we are in some way shape or form, depending on how you define it, but I think this is true we are cyborgs when we use technology. We interact with technology in a way that it extends, is an extension of ourselves. It grants us powers that we didn't have before. I mean the phone, the computer. They've really extended the capabilities of the human individual, and so AI is another iteration of that and an evolution of it.
Speaker 1:And so you know, back to what Elon Musk had said, the only issue that we have to solve in becoming true cyborgs is the interface and the latency issue. Right now there's very high latency because we're using the thumbs, and you know, the more you reduce that latency and that's his whole pitch for Neuralink the more you can interact with the technologies and exponentially increase your powers. And so what we're seeing and what you're doing is just the beginning of that. I mean, imagine what you can do if you can interface directly with chat GPT. I know it sounds scary to a lot of people, but imagine how powerful it is now. How powerful would it be if you can directly interact with it?
Speaker 2:It's going to be. You know it's going to be insane. You know we're going to start spawning agents. Or if you're spawning them with your mind, you're going to be able to spawn a team and they're going to start communicating with each other. I mean we're already doing that now. I mean, look, you know, there's so many different technologies, like Microsoft has Autogen and Autogen Studio for building a team of AI agents. I mean, connect that to Neuralink or Microsoft like a service, like Azure or AWS, a cloud-based service that already has deployable architecture and infrastructure via script, these little AI agents, because those manuals have been online now for 10 years. They already understand how to do all that stuff. It's interesting how close we are to really doing some really interesting things and building some really interesting companies.
Speaker 1:It's crazy to think about. It's absolutely nuts, and you know you would think that this is all science fiction, but the truth is you're doing it and others are doing it, and it's changing the world, isn't it?
Speaker 2:Yeah, yeah, I mean yes, you're doing it and others are doing it and it's changing the world, isn't it? Yeah, yeah, I mean yes, yeah, I mean I was down in san francisco about, um, maybe a month ago now, uh, for a multi-modal hackathon, you know, with it down in the cloudflare headquarters, um, and some of the, some of the technologies that these people, these brilliant people, built, were like mind blowing. I mean, you know they were doing like. You know they're using multimodal AI to generate 3D images in Apple Vision Pro, to use generative AI to adapt personal, identifiable information on meetings like this to sort of, you know, reduce your liability in saving, you know, sensitive information. You know, but I mean, there were like 26 other amazing companies or amazing ideas that these people just built in an afternoon and every single one of them were using generative AI to build their suite of technologies. It was just incredible.
Speaker 1:I know that you've kind of come across some folks who don't get AI. You know they don't really understand it, and so can you explain a little bit of your experience with that, you know, and why do you think that some people don't get it and some people get it intuitively? Good question.
Speaker 2:So they're mean, like definitely on X. They just they're like really mean, they're mean. I don't know why. I wish they were.
Speaker 1:Do you think some people are just against it philosophically and they're you know. They think that this is going to doom humanity and maybe that's a resistance to it, or do you think it's? Do you think some people are just built to embrace technology and maybe others don't get it and they're just jealous?
Speaker 2:I don't know like I don't know man, I hope not jealous, because, like, honestly, anybody can, anybody can, can use it. I mean it's like a natural language, right. Just you just kind of have to open your heart to it, right, and just take like an hour out of your day, just like you know, kind of by yourself. You don't have to do it like a group of people, but like open your heart to it and like really try to try to use it and try to get something meaningful out of it. And it's like like the people that I see, like that are mean, it seems to me, that are mean mean to me or just in general against ai. They to like haven't given it a real chance. You know, like they, they, they were, they were like looking almost immediately to get like all of their problems solved.
Speaker 2:Like this AI drove. Basically you're sitting here telling you that it solved all those problems, it built 75% of this company, you know. But I'll tell you, guys, like it didn't, I didn't do that in like an hour. Yeah, I was like you know, I was like hundreds of hours working with, working with AI, kind of learning about it, learning about me, and they're just being patient with it, with the technology and just approaching it kind of with like an open heart and open mind. I guess that's a spooky thing, you just said.
Speaker 1:You said I'm learning and AI, and the more you can understand that concept, I feel like the world opens up to you and there's more that you can do. And there's probably somebody listening to this right now. I guarantee it. Someone is listening to this right now thinking why, why didn't? I think you know I had the idea to spin up a company, the idea of machines, making machines or using AI to spin up a company, or AI to do X, Y, z, but you know what? I'm going to leave that up to somebody else. Or that's not something that I can do, or maybe that's something that's too big of a task or whatever for those who embrace AI and let it understand you and you try to understand it.
Speaker 2:It somehow rewards you with a sort of power.
Speaker 1:I mean, is that crazy to say?
Speaker 2:Yeah, yeah, man, because of all the negativity that's been cast upon me. But yeah, man, I think it really does.
Speaker 1:What are they saying to you? I mean, what are these? What can these people possibly say?
Speaker 2:They don't believe it. You know they don't believe. They just don't believe that it. You know they don't like a revolutionary new discovery that you're getting.
Speaker 1:And you classify it as a new discovery. That's important. I mean it's it's. It is really a new discovery. And for those who are, I mean, the proof is in the pudding. You've done it. So let's pivot over a little bit to compliance aid, and again, you can't separate the product from the creator, right. So what are some of the challenges of bringing a product like that? So, describe what compliance aid is in maybe 30 seconds or so.
Speaker 2:Describe what compliance aid is in maybe 30 seconds or so, and what are some of the challenges of bringing a product like that to market. So we're trying to mimic cybersecurity auditors for small businesses. Cybersecurity compliance is something that we're seeing globally. We're seeing a lot of people get new compliance regulations. We're seeing globally, you know, we're seeing a lot of people get, you know, new compliance regulations. We're seeing a lot of businesses that are out there, you know, that didn't have to follow regulatory requirements, now have to follow it, and there's service providers that are out there that don't have the right you know amount of staff to actually help these little businesses adhere to this stuff.
Speaker 2:So I'm like, you know, using my little hack and nerdy self, I'm like this is a perfect gap to fill with generative AI. So that's basically the premise of the compliance aid. We're just mimicking cybersecurity compliance auditors for small businesses, you know, and we're trying to do that through service providers, because those people are the ones who have been asked by their clients to, you know, for this task and they're right now and some of them are having trouble doing this stuff and I think generative AI would help these businesses. And it's just like me, right, you know, if you look into the future, you know businesses are going to have teams of AI agents that are. You know they're helping them and you know, because I did it for my business, I'm now ready to help service providers implement um.
Speaker 2:And then what are the challenges to bringing that to market? Gosh, I mean, early on people were like, obviously, like you know, wondering what generative ai was. Um they talked to me about hallucinate, uh, hallucinating a lot. They were always asking about. You know, how often does it hallucinate? And, quite frankly, like I didn't see any of the hallucination issues that people were saying, even during gpt tree.
Speaker 1:Yeah, number one. Someone heard the word hallucination and they wanted to sound smart, so they used it. There I said it.
Speaker 2:Thank, you for saying that, because it's like over and over again. I'm like guys, like I haven't had this issue like since, like, like for like November of last year, like, and it was my problem, like I figured a fault and it was the reason why I was hallucinating, because it was doing what I was telling it to do, which is spelling incorrectly, because I spelled it wrong, right.
Speaker 1:So I think that the hallucinations are more of a problem when it comes to numbers. If you ask it for certain kinds of data and you ask it the same thing over and over, it may give you some weird stuff, but when it comes to everything else, it really the hallucination rate is very, very low. Yeah, and especially you're working in the world of policy, so it's words, and so the hallucination rate for something like that is extremely low. It's not like it's trying to solve math problems necessarily.
Speaker 2:So there's that, and then that was the biggest challenge is just people building trust with AI to bring it to market.
Speaker 2:Yeah, perception. Definitely, people are definitely worried that it's going to take their jobs at first. But then the team that I'm working with now I can't really say their name, but there's a really cool North American-based compliance firm that does healthcare for some big hospitals. They're so overwhelmed. They have like so much business and I just like this morning I just finished, like you know, doing this crazy high trust assessment for them and this thing takes like six weeks for them to do and I did it in like 24 hours with our AI team and the relief is like, oh my God, and they helped me build the prompts for it, you know for the last few months, and me build the prompts for it, you know for the last few months. And so they're like reading the responses and like this is so cool, you know they're so excited.
Speaker 1:So yeah, yeah. Well, I can't believe there are people coming at you for this, people who were, you know, and then ai bro thing is supposed to be. I guess it's meant to be derogatory, but you know what? I own it, you own it, we're ai bros. Whatever you want to call it, that's fine, we are, we are the future, and the future is ai. Okay, so, okay, let's, let's. Let's close with this. What do you want your legacy to be? Do you want it to be ai and compliance aid, or is there some larger mission that you're on that you want to be remembered for?
Speaker 2:man, okay, look, um, I'm probably like one of the first out there, like in the world to, like you know, basically use ai like for like a majority part of my life, like I plan vacations for me.
Speaker 2:It's tough to meet with, you know. So personal advice and definitely very professional advice, and I think it's done really, really well for me so far. You know. So, like I just want to help people um, like moving forward, like so they, you know, if there's anybody out there like business owner, you know cto, cto, you know entry level person, you know young person, like if there's anybody out there that sees this and you're like, you're feeling like you're an iron. You know, like I kind of felt, you know, about this ai thing, like reach out to me, you know, and you know I'm happy to help you with any advice, but that's really that would be like my legacy, more so than like the compliance. I just, I guess I just want to help people kind of like, you know, you know, get over this little learning curve, which and let's, let's just see what kind of exciting possibilities all of us can you moving forward excellent, andy.
Speaker 1:This was an absolute pleasure. I'm sure you're going to be back and I'm sure the story will evolve over time and anybody who wants to get a hold of randy he is very accessible, like he's just said. If you need to call him, call him. As a matter of fact, I'll tell you his cell phone number. Right now it's no, I'm kidding. What's the best way for people to reach you?
Speaker 2:uh, sure man, uh, so uh, it's on twitter, uh, at place of randy, pretty, pretty easy, or linkedin same deal and I think I could probably send those to josh. Maybe he can post that up on on the video yeah, so I'll post it up.
Speaker 1:You guys, get a hold of Randy, talk to him for yourself. I mean, if you're interested in learning more about compliance aid, you're interested in learning more about generative AI and sort of the applications that will be inevitable in the future and will change the face of the earth, you know, get a hold of us, talk to us. We're happy to have those conversations with you. So you can reach me on linkedin. I'm at linkedin and it's at josh bruning j-o-s-h-b-r-u-y-n-i-n-g. Or you can send me an email, uh, josh, at bruningcom. Or, if you've got my phone number, call me, let's talk. And if you're interested also in being a guest on the podcast, hit me up. I'm excited to hear your story. I'm sure that your customers, your prospects, your audience, whoever they may be, they are dying to hear your story and to connect with you. So with that, randy Blasick, thank you so much and thank you for listening to this episode of Security Market Watch.