00:53:51 · David Senra, Nikesh Arora
[00:53:51] S0: I completely agree with this perspective. Out of all the examples I've read about, I love what Sam Walton said in his autobiography, which he wrote when he knew he was dying, had cancer all over his body. And he says, I know I'm a fairly overactive fellow, which is the most understatement of a century. He's like, I don't put pressure on my kids. This is like, let them Why would I want them to be like me? [00:54:09] S1: But it's kind of interesting. If you do it the right way, your kids will watch your behavior. They watch more behavior than you think. They know more about how we are, who we are than we possibly can ever understand. And sometimes they will want to do certain things like you. But if you try and force them to do certain things like you, you're never gonna get there. I can see my kids would like to do certain things like me, that's great. I'm very happy that they want to. And I'm sure sometimes they're looking and saying, I never wanna do that thing that my dad does. So that's fine too. As long as they can find out what parts of what I do they like to do and what parts of what I do they don't like to, they're fine. I'm okay with that. I... [00:54:46] S0: I'm thinking back to, like, all the other people I've had on the podcast. And I'm like, I just feel, yes, AI is changing every single business. Toby Lütke was on this this this show earlier this year. [00:54:54] S1: I'm going
00:54:54 · Nikesh Arora, David Senra
[00:54:54] S1: I'm going [00:54:54] S0: to see him again in like two weeks to record again. And his whole point was like, we're going look back on the year 2026 where because of AI, like every business is up for grabs. So every business is being transformed somewhat, right? But I feel like yours might be the most out of anybody else I've had on the show.
00:55:10 · Nikesh Arora
[00:55:10] S1: No, no, no. Let's break it down. Cybersecurity is because I don't come from that, I told you it was two different words in my mind eight years ago, I'm still going to treat it like that. It's somewhat straightforward. Cybersecurity is the idea that bad actors trying to get into infrastructure and exfiltrate something important. Let's take that for a premise. If I know it's bad, I can stop it at the perimeter. Bad guy shows up at the door, we know he looks bad. He's got a hoodie, he's got a gun. This is bad. We know how to stop them. Half of cybersecurity is stopping bad guys at the perimeter, bad things at the perimeter. That's half of cybersecurity. 10%, 20% is analytics how to get hygiene right once they're inside your house to make sure shit's sorted. The remaining 2030% is if somebody gets in, what do I do? Things will get through your perimeter inside your infrastructure. How quickly can I find the bad guy? And how quickly can I inject them out of my infrastructure before he or she can do harm? That's cybersecurity in a nutshell. How does AI impact this? Let's start with your comment about vulnerabilities. AI says, I can find holes in your perimeter faster than anybody else in the world. It used to take days to find it, it take forever. I can find holes to enter your perimeter. Okay, great. My perimeter just became more porous. Now, AI is not gonna patch the perimeter. I have to still go look through the infrastructure, find where the porous parts are and patch them. I could be a software company, I could be a infrastructure provider because everybody's IT stack is made up of tens and thousands of vendors. Vulnerabilities So exist in our software and software that's driven by our customers themselves. That's in terms of step one. That's what we're seeing right now. The likes of Mythos, likes of the new models that OpenAI has, the likes of GMM 5.3, they have immense capability of finding vulnerability and bad actors can rent them cheap and throw them at you and find everything that they have vulnerability around. What is the call of the hour? The call of the hour is to find a way to patch these things swiftly so these bad guys can't get it. Guess who's building the service? We're building a service which is called virtual patching.
00:57:12 · Nikesh Arora, David Senra
We write signatures, we put bandage around the entire perimeter, we know where the holes are, these people can't get in. Over time, our customers will patch the underlying systems and they'll be fine. So that's one problem. Good news is our business has gone up since Mythos came in. [00:57:27] S0: That's what I'm trying get to. [00:57:29] S1: Holy shit. Can you help me patch it? I guess what? Yes, I can help you patch it. I inspect everyone's parameters. So you've got a laptop somewhere, I live on a laptop, I charge you $25 a year. And I watch 150 megabytes a day, give or take on your laptop to make sure nothing bad is coming in. [00:57:46] S0: You told me you're hacking my phone calls. [00:57:48] S1: I was just kidding. I wouldn't know how to do it. Remember cybersecurity is too worse. The marginal cost of inspection of traffic in laptop by me is zero. I don't charge you more the more traffic that goes through. AI is set up to charge you every time you run traffic through AI in terms of tokens. So if AI tried to do my job at your endpoint, it'll cost you $5,000 a day or pick a number, dollars 500 a day, you'll pick up $50 a day. They can't do what I do at $25 an endpoint. If you go back inside and if the bad actor gets in, what happens? Yes, AI can help you analyze all the anomalous behavior quickly, but you still need harnesses, still security expertise, you need data ingestion, you need infrastructure. So in a nutshell, our business is not at risk. [00:58:37] S0: Yeah, from outside, that's what I was assuming was the case.
00:58:37 · David Senra, Nikesh Arora
[00:58:37] S0: Yeah, from outside, that's what I was assuming was the case. [00:58:40] S1: What I was like always assumed that CEOs are trying to maximize their businesses from the outside. Generally, that's a good rule. [00:58:46] S0: Yeah, when we talked about no one wakes up wanting to lose, but a lot of people lose every day. [00:58:50] S1: This is what I [00:58:51] S0: was trying to get to. Who do you go to when you want to learn about like the frontier of AI? Who do I go
00:58:56 · Nikesh Arora
[00:58:56] S1: to when I want to learn more about AI? There are different people. I spend, I want to say at least four or five hours a week with AI founders of some sort. Because these kids are young. They possibly worked at OpenAI. They probably worked at Google somewhere. They got these companies. They're very kind. They spent time with me. They try and tell me what they do. I learned about weight models. Learned about training. I learned about training data, I learned about guardian agents. I go around, I talk to tons of people. I try and I talk to probably five or 10 cyber startups a month, maybe 15. So I'm trying to absorb how they're thinking about it. Because in a bizarre way, this is a new and developing science for lack of a better word. There are no real experts. People know how to do their shit. Everybody else is trying to imagine the implications to everything else around them and they're trying to build a mental models, they can build product and go win in this category. There's no concept of data labeling before you started having self driving cars beyond label data. Now let's turn into reinforcement learning, simulation models, neural networks that are trying to simulate data so we can train AI. So literally have to stay in the conversation with tons of people to understand and absorb what people are talking about. You just have to spend a lot time on Twitter, spend a lot of time talking to people, spend a lot of time reading stuff. You're like, half I understand, half just gets through the other side, can I absorb everything? That's how I learn.