WEBVTT - A Conversation with Mike Privette from Return on Security 0:00:00.290 --> 0:00:02.780 All right, Mike, welcome to unsupervised Learning. 0:00:03.730 --> 0:00:04.899 Hey, thanks for having me, Daniel. 0:00:06.040 --> 0:00:08.170 Very cool. So I've been wanting to get you on 0:00:08.170 --> 0:00:11.080 the show for quite some time, and we've been talking 0:00:11.080 --> 0:00:14.500 about doing it. I really love what you're doing over 0:00:14.500 --> 0:00:18.639 there on Return on Security. I kind of see you as, like, 0:00:18.640 --> 0:00:24.280 the Nate Silver or the 538 of, uh, the security industry. Like, 0:00:24.280 --> 0:00:28.750 you just kind of seem to know what's going on, and, uh, 0:00:28.750 --> 0:00:32.530 just really impressed with what you're doing. Can you tell us, um, 0:00:32.530 --> 0:00:35.230 about what? All you have, uh, sort of in the 0:00:35.229 --> 0:00:37.780 hopper and going on in different projects you're doing. 0:00:38.350 --> 0:00:41.350 Yeah. Awesome. Uh, well, thanks for that intro. And, uh, 0:00:41.350 --> 0:00:43.690 and the kind words, too. I love that, uh, the 0:00:43.690 --> 0:00:47.680 concept of, uh, the 538. Um, you know, I really 0:00:47.680 --> 0:00:50.440 am just trying to at return on security, capture the 0:00:50.440 --> 0:00:54.610 entire industry, uh, of cybersecurity. Um, and I've, you know, 0:00:54.610 --> 0:00:56.470 through the years of working and I've seen a bunch 0:00:56.470 --> 0:00:59.200 of different variations of people looking at the industry from 0:00:59.290 --> 0:01:02.050 either a product centric kind of view or from an 0:01:02.050 --> 0:01:05.290 investment banking kind of view. Uh, or for like a 0:01:05.290 --> 0:01:08.380 public equities, like, you know, the big public companies that 0:01:08.380 --> 0:01:10.840 get traded and talked about a lot. Uh, but I've 0:01:10.840 --> 0:01:13.480 never seen anybody kind of pull it all together, uh, 0:01:13.480 --> 0:01:16.000 in a way that you can understand if you're not 0:01:16.000 --> 0:01:18.940 an investment banker. Uh, and so that was kind of 0:01:18.940 --> 0:01:21.280 one of the things I was trying to do return 0:01:21.280 --> 0:01:23.980 on security was to kind of catalog all these things, 0:01:23.980 --> 0:01:26.919 but speak for the voice of the practitioner, uh, and 0:01:26.920 --> 0:01:28.510 the CSO. And that's, that's kind of what led me 0:01:28.510 --> 0:01:33.070 to this. And just, um, most of my experience, uh, 0:01:33.069 --> 0:01:35.290 through security has been on the engineering side. So I 0:01:35.290 --> 0:01:37.420 was always kind of front and center with buying tools, 0:01:37.420 --> 0:01:42.940 deploying tools, operationalizing tools. Um, and I had a lot 0:01:42.940 --> 0:01:46.330 of opinions because of that. Yeah. So it just started, uh, 0:01:46.330 --> 0:01:48.610 kind of writing them down and, uh, capturing this stuff. 0:01:48.610 --> 0:01:50.920 And I thought, uh, maybe I can help other people 0:01:50.920 --> 0:01:52.750 understand this space better, too. 0:01:53.350 --> 0:01:57.370 Yeah, that's really cool. So who is the main audience 0:01:57.370 --> 0:02:01.210 like the ICP here? You mentioned CSOs. That definitely makes sense. 0:02:01.210 --> 0:02:04.420 So security leaders, but would you say also like senior 0:02:04.420 --> 0:02:08.110 practitioners as well like technical people or what do you find. Yeah. 0:02:08.410 --> 0:02:11.350 I would definitely say that. Definitely security leaders. Uh, and 0:02:11.350 --> 0:02:13.570 then a lot of the senior practitioners who are often 0:02:13.570 --> 0:02:17.140 the ones informing or advising the leaders on what to buy, 0:02:17.139 --> 0:02:19.750 what to what kind of direction to go in, what 0:02:19.750 --> 0:02:23.440 kind of strategy to put together. Um, and that I 0:02:23.440 --> 0:02:25.330 kind of made this also for me as I was 0:02:25.330 --> 0:02:27.910 missing kind of these aspects of what, uh, you know, 0:02:27.910 --> 0:02:31.120 the weekly newsletter could kind of give like, what's the 0:02:31.120 --> 0:02:33.010 market even look like right now? And what should I 0:02:33.010 --> 0:02:36.250 be thinking about ahead of time? Um, so I'd say 0:02:36.250 --> 0:02:39.820 that that audience of security people makes up about 80% 0:02:39.820 --> 0:02:43.690 of the audience. Um, the other 20% is actually a 0:02:43.690 --> 0:02:46.870 healthy dose of cybersecurity founders. Uh, they use a lot 0:02:46.870 --> 0:02:49.480 of it. What I produce is like market intelligence for 0:02:49.480 --> 0:02:53.320 themselves against competitors or whatnot. Um, and then a lot 0:02:53.320 --> 0:02:56.110 of the rest of that is also, uh, VCs or 0:02:56.110 --> 0:02:59.470 private equity funds, uh, who are also doing similar kind 0:02:59.470 --> 0:03:02.109 of market intelligence and trying to keep track of startups 0:03:02.110 --> 0:03:03.790 where they are in their journey and, you know, when 0:03:03.790 --> 0:03:05.169 they might want to raise more money. 0:03:06.010 --> 0:03:09.430 Yeah. That's perfect. I mean, I don't know how much 0:03:09.430 --> 0:03:11.650 that was an accident, but you really did create the 0:03:11.650 --> 0:03:14.110 best resource for all those different groups. 0:03:14.650 --> 0:03:17.050 Well, thank you. It was kind of an accident, but, uh, 0:03:17.050 --> 0:03:19.750 just kind of was born out of, uh, just like, 0:03:19.750 --> 0:03:21.790 just trying a bunch of different things and getting some 0:03:21.790 --> 0:03:25.660 kind of outside help, uh, from friends who were in finance. Um, 0:03:25.660 --> 0:03:27.460 so I kind of wanted to make it sure that 0:03:27.460 --> 0:03:29.620 they could digest it, too. So I asked some friends 0:03:29.620 --> 0:03:34.839 who do, uh, investment banking or, uh, M&A activities and say, like, 0:03:34.840 --> 0:03:38.080 how do you normally want to see financial data or 0:03:38.080 --> 0:03:40.630 money data, things like that, like what makes sense? And 0:03:40.630 --> 0:03:42.550 then I kind of try to apply that in security. 0:03:42.550 --> 0:03:45.160 And uh, it was honestly a lot of luck to, uh, 0:03:45.160 --> 0:03:46.330 putting it all together. 0:03:47.460 --> 0:03:50.460 Yeah, I feel like it's really congealed over, like the 0:03:50.460 --> 0:03:53.520 last year or so, like where you kind of found 0:03:53.520 --> 0:03:56.940 that voice and you found the audience and you know 0:03:56.940 --> 0:04:01.740 what they are looking for. Yeah, really, really strong. So 0:04:01.740 --> 0:04:04.350 I guess I've got a number of questions for you. Um, 0:04:04.350 --> 0:04:06.750 but they all kind of have the same theme, which 0:04:06.750 --> 0:04:10.740 is what the hell has been going on? Um, like, 0:04:10.740 --> 0:04:14.190 so for like the last, I don't know, I can't 0:04:14.190 --> 0:04:16.710 even remember how long it's been. Is it like three years? 0:04:16.710 --> 0:04:22.170 Maybe three years, maybe a year and a half where. So, like, 0:04:22.170 --> 0:04:28.729 the economy is like. Or in cybersecurity specifically. We had 0:04:28.730 --> 0:04:31.040 a boom. I can't remember when the boom kind of, 0:04:31.190 --> 0:04:34.310 I guess the boom was post-Covid. It was it was 0:04:34.310 --> 0:04:38.440 like a post-Covid, like hiring boom. And then there was 0:04:38.440 --> 0:04:42.670 like all this excitement and sort of froth. And then 0:04:42.670 --> 0:04:46.930 at some point, um, there was like, just like the 0:04:46.930 --> 0:04:50.500 crash that came after that. Right. So how would you 0:04:50.500 --> 0:04:52.840 describe this? Like, it seems like you've got to be 0:04:52.839 --> 0:04:56.500 like the, the, uh, the expert here and one of 0:04:56.500 --> 0:04:59.169 the top experts in the world on this, and I'm 0:04:59.170 --> 0:05:02.200 trying to find out, of course, it's multivariate, but, like, 0:05:02.200 --> 0:05:08.800 what was this? Was this all the result of anti-Covid 0:05:08.800 --> 0:05:14.450 or over hiring from Covid? Or was it maybe pushed 0:05:14.450 --> 0:05:19.720 back on? Um. Uh, policies inside of companies being a 0:05:19.720 --> 0:05:23.140 little bit too pro employee. And companies are like, look, 0:05:23.140 --> 0:05:26.229 we only I've got like this theory of like, uh, 0:05:26.230 --> 0:05:30.280 the Alaskan fishing boat thing where it's like, I need 0:05:30.279 --> 0:05:34.719 absolute killers on this boat. Only the best. And like, 0:05:34.720 --> 0:05:36.820 if you say your ankle hurts, you can't come on 0:05:36.820 --> 0:05:41.170 the boat, right? Right. If if your priority is your family, 0:05:41.170 --> 0:05:44.380 you can't come on the boat. You want work life balance. 0:05:44.380 --> 0:05:47.740 You're not coming on my crew. And so I felt 0:05:47.740 --> 0:05:50.650 that happening at the same time. And I just I'm 0:05:50.650 --> 0:05:52.359 not as close to the data as you are, so 0:05:52.360 --> 0:05:57.170 I'm wondering what you think. Cause the boom caused the crash. 0:05:57.170 --> 0:05:59.240 And where are we now? Yeah. 0:05:59.570 --> 0:06:02.060 That's a great question too. And I think honestly, the 0:06:02.060 --> 0:06:06.110 the crash was building long before Covid happened. Um, if 0:06:06.110 --> 0:06:10.429 basically if you look after the Great Recession in 2008, uh, 0:06:10.430 --> 0:06:15.020 in the US, when the housing market at the first crash, uh, 0:06:15.020 --> 0:06:21.109 and it effectively recovered around 2012, 2013, that was a 0:06:21.110 --> 0:06:24.380 time at which interest rates had become very, very low 0:06:24.380 --> 0:06:28.909 to stimulate kind of growth in the economy that stimulated investments, um, 0:06:28.910 --> 0:06:33.050 outside of traditional banks. And so money was a effectively 0:06:33.050 --> 0:06:37.460 very cheap to borrow, which, yeah, venture capital loves startup 0:06:37.460 --> 0:06:41.330 founders love and that uh, that the old habits from 0:06:41.330 --> 0:06:44.120 early days of Silicon Valley and like the pre com 0:06:44.120 --> 0:06:47.810 bust era continued on. So it was uh, it was 0:06:47.810 --> 0:06:50.030 very easy to grow a company quickly. It was very 0:06:50.029 --> 0:06:52.820 easy to raise a lot of money quickly. It was 0:06:52.820 --> 0:06:56.870 very easy to scale. Um, it wasn't impossible to, to 0:06:56.870 --> 0:06:59.540 raise a couple hundred million dollars and then go IPO 0:06:59.540 --> 0:07:01.250 in this time frame, which is, you know, kind of 0:07:01.250 --> 0:07:04.280 the ultimate goal of, uh, you know, companies to raise 0:07:04.400 --> 0:07:08.539 venture capital money. Um, so all of those things played 0:07:08.540 --> 0:07:11.480 out across the entire industry, like, uh, not just cyber, 0:07:11.480 --> 0:07:15.739 but the entire set of industries. Uh, it just so happened, though, 0:07:15.740 --> 0:07:18.320 at the same time, tech had a, you know, a 0:07:18.320 --> 0:07:21.290 total run. Uh, they did really, really well as a 0:07:21.290 --> 0:07:25.580 general sector. And then cybersecurity did even better within the side, 0:07:25.580 --> 0:07:29.360 that small sector of tech. And it almost seemed recession, uh, 0:07:29.360 --> 0:07:33.380 recession proof in some ways. Um, and any time there 0:07:33.380 --> 0:07:36.470 was news funding went up or investments went up, any 0:07:36.470 --> 0:07:39.470 time there was a breach, investments went up. Anytime there's 0:07:39.470 --> 0:07:42.350 global conflict, investments went up. And so like, it kind 0:07:42.350 --> 0:07:45.860 of just kept happening in cybersecurity. Uh, and so there 0:07:45.860 --> 0:07:48.620 was really a good ten year or so period in 0:07:48.620 --> 0:07:51.590 cybersecurity where things were just as good as they could be. 0:07:51.590 --> 0:07:54.200 It could only go up into the right. Uh, and 0:07:54.200 --> 0:07:57.290 so a lot of people flocked to that. A lot 0:07:57.290 --> 0:08:00.530 of investments flocked to that, uh, a lot of people 0:08:00.530 --> 0:08:03.380 that would have put their money elsewhere, dumped it all 0:08:03.380 --> 0:08:07.160 in cybersecurity. Uh, some became specialists, some became just tourist 0:08:07.160 --> 0:08:12.530 investors who'd, you know, invest occasionally. Um, and then as 0:08:12.530 --> 0:08:16.430 you go up to Covid area, Covid era policies had 0:08:16.430 --> 0:08:19.610 to kind of reverse all of this low interest rate 0:08:19.610 --> 0:08:23.690 phenomenon that happened across the world. Uh, and each country 0:08:23.690 --> 0:08:26.210 took a different approach, but the US took a pretty 0:08:26.210 --> 0:08:29.150 hard approach. Uh, the US Federal Reserve, that is, took 0:08:29.150 --> 0:08:32.780 a hard approach on ratcheting interest rates way up to 0:08:32.780 --> 0:08:36.470 kind of slow down the possibility of having another recession 0:08:36.470 --> 0:08:40.640 or even a depression in some ways. Um, and since 0:08:40.640 --> 0:08:42.770 most of the money in the world invested in tech 0:08:42.770 --> 0:08:45.470 and cyber comes out of the US or originates in 0:08:45.470 --> 0:08:49.430 the US, uh, that hit the tech sector super hard 0:08:49.429 --> 0:08:54.410 and hit cybersecurity sectors very hard. Um, so imagine where 0:08:54.410 --> 0:08:57.110 you've raised, you know, a couple hundred million dollars over 0:08:57.110 --> 0:09:00.470 the past 3 or 4 years prior to about 2022 0:09:00.500 --> 0:09:04.189 is when, uh, early 2022, when this happened. And then 0:09:04.190 --> 0:09:06.800 you might be worth $1 billion. You might be worth, 0:09:06.800 --> 0:09:09.800 you know, a couple billion dollars. But then now, because 0:09:09.800 --> 0:09:12.350 you've spent all this money and time and effort into 0:09:12.350 --> 0:09:15.530 growing kind of at all cost, as opposed to like 0:09:15.530 --> 0:09:19.310 growing a product market fit and growing like products in 0:09:19.309 --> 0:09:21.440 a way that people actually want to consume, not just 0:09:21.440 --> 0:09:24.990 to compete or winner take all. People are now left 0:09:24.990 --> 0:09:28.170 with like, wait, sales don't match what my valuation says. 0:09:28.170 --> 0:09:30.990 Sales don't keep up with the money I raise. And 0:09:30.990 --> 0:09:34.980 so there was a super hard reckoning, um, from a 0:09:34.980 --> 0:09:38.280 lot of things like operational efficiency became the name of 0:09:38.280 --> 0:09:40.620 the game. Um, so it was no longer enough to 0:09:40.620 --> 0:09:43.200 raise money. You actually had to, like, make money as 0:09:43.200 --> 0:09:45.959 a company, which, uh, kind of cracked me up thinking 0:09:45.960 --> 0:09:46.890 about it, but. 0:09:46.890 --> 0:09:50.040 Yeah, let me let me jump in there real quick. So, um. 0:09:51.870 --> 0:09:54.300 I've heard. Uh, do you listen to Scott Galloway? 0:09:54.750 --> 0:09:55.350 I have. 0:09:55.590 --> 0:09:59.970 Yeah. Yeah, yeah. So he talks about this a lot. How? Um, 0:09:59.970 --> 0:10:02.310 I'm not sure he says it like this, but there's essentially, like, 0:10:02.309 --> 0:10:05.459 two phases or two different vibes that a company can 0:10:05.460 --> 0:10:09.530 be in, which is like this. Kind of like a 0:10:09.530 --> 0:10:15.170 fantasy phase where you disconnect fully from the underlying financial 0:10:15.170 --> 0:10:18.620 fundamentals and you're just kind of operating on, like you said, 0:10:18.620 --> 0:10:23.600 growth at any cost or just like this state of 0:10:23.600 --> 0:10:28.160 belief and hype, kind of like a crypto surge or something. 0:10:28.160 --> 0:10:31.939 And it has its own gravity and it takes on 0:10:31.940 --> 0:10:35.179 its own form. And he said the same thing you 0:10:35.179 --> 0:10:38.720 just did, which is at some point there is a reckoning. 0:10:39.580 --> 0:10:42.910 Eventually you're going to get back to somebody. Maybe it's 0:10:42.910 --> 0:10:46.989 like the replacement management team. They come in and they're like, hey, let's, um, 0:10:46.990 --> 0:10:49.360 let's do something crazy. Let's figure out how much money 0:10:49.390 --> 0:10:53.950 we're making and how much money we're spending. Right. Um, 0:10:54.160 --> 0:10:56.679 so so this is really interesting. So you have an 0:10:56.679 --> 0:11:01.209 underlying economic thing, which is kind of the biggest. Yep. 0:11:01.480 --> 0:11:05.800 Then you have maybe related to that, you have this, um, 0:11:05.800 --> 0:11:09.490 transition of a lot of these companies going for that 0:11:09.490 --> 0:11:13.090 other style of company where it's not super tied to 0:11:13.090 --> 0:11:16.720 the product itself. And the I think Scott calls it 0:11:16.720 --> 0:11:22.089 like the fundamentals, market fundamentals. Mhm. And then um, then 0:11:22.090 --> 0:11:27.240 you have Covid in addition, which is another disruptor. And 0:11:27.240 --> 0:11:31.390 it sounds like all those sort of combine and. To 0:11:31.390 --> 0:11:35.410 form the build up, but also to form the crash. Yep. 0:11:36.580 --> 0:11:38.410 I think that's a great way to think about it. 0:11:38.410 --> 0:11:41.439 And you know it. Also, you look at broader tech 0:11:41.440 --> 0:11:45.940 kind of, uh, plays that happened so largely, uh, tech 0:11:45.940 --> 0:11:48.849 sectors copy like people or companies in the tech sector 0:11:48.850 --> 0:11:53.530 copy each other. If, if a Facebook or a Google or, um, 0:11:53.530 --> 0:11:57.370 a Microsoft makes a move to reduce headcount, many of 0:11:57.370 --> 0:11:59.980 them follow suit. The other big companies and then companies 0:11:59.980 --> 0:12:04.750 start reevaluating. Maybe we should do that too. Uh, maybe 0:12:04.750 --> 0:12:07.479 the investors certainly come knocking and they say, hey, wait 0:12:07.480 --> 0:12:10.600 a second, shouldn't you be more profitable or more operationally efficient? 0:12:10.600 --> 0:12:13.420 Let's let's drive that cost burn down. Uh, and so 0:12:13.420 --> 0:12:15.700 that also added into the effect as well. 0:12:15.910 --> 0:12:21.550 Yeah. It, it feels like both of those are super contagious. The, 0:12:21.550 --> 0:12:26.870 the getting away from fundamentals is super contagious. In the hype, 0:12:26.870 --> 0:12:30.500 and then the returning of fundamentals is also equally contagious. 0:12:30.500 --> 0:12:34.790 Everyone's like, hey, what are we doing? Like, I'm talking 0:12:34.790 --> 0:12:37.969 to Chuck and Sally over here, and, uh, they're both 0:12:37.970 --> 0:12:39.860 saying we should get back to the books and we 0:12:39.860 --> 0:12:43.939 should absolutely do that. It's like fashion trends almost. 0:12:44.750 --> 0:12:46.550 Yeah, that's a good way to think about it, too. 0:12:46.550 --> 0:12:49.670 Like what's in vogue versus what's not. Um, and a 0:12:49.670 --> 0:12:51.709 lot of it, as much as people want to you 0:12:51.710 --> 0:12:54.080 believe that, like, you know, the market is is based 0:12:54.080 --> 0:12:57.290 on data and science and, um, hard proof. A lot 0:12:57.290 --> 0:12:59.390 of times, a lot of us just feeling like even 0:12:59.390 --> 0:13:01.850 investments are a lot of feeling, um, a lot of times, 0:13:01.850 --> 0:13:05.390 which is why, um, not all investments, most investments don't 0:13:05.390 --> 0:13:08.360 return any money. Uh, so, like, that's, you know, there's 0:13:08.360 --> 0:13:10.160 you can't predict some of these things. So you have 0:13:10.160 --> 0:13:11.930 to go with a bit of underlying feeling and gut. 0:13:11.929 --> 0:13:13.820 And so that is contagious as well. 0:13:14.510 --> 0:13:19.440 Um, yeah. Interesting. So what are some of the other 0:13:19.440 --> 0:13:21.750 trends that you're seeing that you're sort of tracking that 0:13:21.750 --> 0:13:22.890 you find interesting? 0:13:23.830 --> 0:13:25.270 So right now, I think we've had a little bit 0:13:25.270 --> 0:13:28.450 of a rebound effect because there is, uh, it's a 0:13:28.450 --> 0:13:32.170 kind of a macroeconomic phenomenon where people expect things to 0:13:32.170 --> 0:13:34.450 get better. And so if they expect it, it does 0:13:34.450 --> 0:13:38.800 get better. It's a strange, like, chicken and egg concept. But, uh, 0:13:38.920 --> 0:13:41.740 there was a bit of a rebound. Um, I've noticed 0:13:41.740 --> 0:13:46.390 that 2024, even in these first 12 weeks, is already, uh, 0:13:46.390 --> 0:13:50.080 almost as good as 2023. Um, and even though there's 0:13:50.080 --> 0:13:52.870 been less transactions, there's almost the same exact amount of money, 0:13:52.870 --> 0:13:56.710 so there's fewer transactions happening, but there are more, uh, 0:13:56.710 --> 0:14:00.280 profitable each time or more, more revenue is, uh, or more, uh, 0:14:00.280 --> 0:14:05.020 cash is exchanged hands each time. Um, and like the 0:14:05.020 --> 0:14:08.080 M&A has also kind of been one of those, uh, 0:14:08.080 --> 0:14:11.440 kind of second or third order effect knock ons that 0:14:11.440 --> 0:14:16.630 have come from the fallout of 2022. So companies who, uh, 0:14:16.630 --> 0:14:19.810 started in that time frame probably had a really tough time. 0:14:19.810 --> 0:14:23.500 They either sink or swim very quickly. Um, and I'm 0:14:23.500 --> 0:14:26.950 seeing companies who started early. Now they're getting acquired quicker, 0:14:26.950 --> 0:14:29.260 which is, you know, it might be, you know, a three, 0:14:29.260 --> 0:14:32.140 four year turnaround, uh, for an acquisition now as opposed 0:14:32.140 --> 0:14:34.989 to like 7 to 10 in some cases. So I 0:14:34.990 --> 0:14:38.680 think that is interesting. Uh, because, uh, the, the underlying 0:14:38.680 --> 0:14:42.190 premise there is like companies either need a lifeline because 0:14:42.190 --> 0:14:44.680 they can't survive and actually make sales themselves, and they 0:14:44.680 --> 0:14:48.850 need like to be absorbed or acquired into a larger company. Um, 0:14:49.480 --> 0:14:51.670 or they just have something that's so good that larger 0:14:51.670 --> 0:14:55.239 companies like can't iterate or can't innovate without it. So 0:14:55.240 --> 0:14:56.890 they buy it up real quick to try to get 0:14:56.890 --> 0:14:58.960 it off the market and try to stay relevant with 0:14:58.960 --> 0:15:03.910 the other large companies. Um, so I think that's that's 0:15:03.910 --> 0:15:08.680 pretty interesting. Um, another concept I would say is, uh, 0:15:08.680 --> 0:15:12.160 you know, obviously application security has exploded with AI. I 0:15:12.160 --> 0:15:15.190 think that's probably like one of the quickest and, uh, 0:15:15.190 --> 0:15:17.080 probably one of the better use cases to kind of 0:15:17.080 --> 0:15:20.590 help support that function. Um, and it feels a lot 0:15:20.590 --> 0:15:24.250 more tangible than, you know, an autonomous stock agent or 0:15:24.250 --> 0:15:27.640 like a, a replace your SoC. Analysts like levels one 0:15:27.640 --> 0:15:30.280 and two with a, you know, an all knowing agent. 0:15:30.430 --> 0:15:32.860 It may come their that way eventually, but uh, I 0:15:32.860 --> 0:15:36.280 think um, the, the first the shine has worn off 0:15:36.280 --> 0:15:38.740 of the first level of AI security or security for 0:15:38.740 --> 0:15:41.620 AI companies. Uh, and people are starting to see that 0:15:41.620 --> 0:15:45.820 there's like real tangible benefit and like the application security 0:15:45.820 --> 0:15:47.260 and the data security side. 0:15:47.260 --> 0:15:51.250 When you say the first layer, you mean pre this 0:15:51.250 --> 0:15:54.070 new gen of AI, you mean like the old AI 0:15:54.100 --> 0:15:58.510 security companies who were like. Yeah, automate the sock. But 0:15:58.510 --> 0:16:01.540 that was before Gen I. You're saying that the Gen 0:16:01.540 --> 0:16:06.070 I stuff is like injecting some real, real, um, energy? 0:16:06.070 --> 0:16:09.729 I definitely agree with that. But how? We're using it manifest. 0:16:09.730 --> 0:16:12.100 Yeah, I'm seeing some energy. So I think there were 0:16:12.100 --> 0:16:15.550 some last year, uh, you know, after after OpenAI and 0:16:15.550 --> 0:16:17.800 everyone came out or like, it's 18 months ago. So 0:16:17.800 --> 0:16:20.080 now there's a lot of early, early companies that were 0:16:20.080 --> 0:16:24.910 trying to do something with security as like as new startups. Um, 0:16:24.910 --> 0:16:28.030 most of them were just like, uh, proxies or casbs 0:16:28.030 --> 0:16:30.370 or just like basically a firewall to kind of limit 0:16:30.370 --> 0:16:33.520 the exposure that a company would have. Um, those had 0:16:33.520 --> 0:16:35.650 a quick pop and then they've kind of like trailed off. 0:16:35.650 --> 0:16:39.700 And then you saw the next iteration even after like, uh, 0:16:39.700 --> 0:16:42.820 OpenAI came out and the others, uh, of like replace 0:16:42.820 --> 0:16:45.190 the sock, but like in a different way than how 0:16:45.190 --> 0:16:48.700 they used to say, replace the sock. Um, now more 0:16:48.700 --> 0:16:51.460 of like, just like a the co-pilots, if you will, 0:16:51.460 --> 0:16:54.070 like the the everyone, every product now has an AI 0:16:54.100 --> 0:16:57.370 copilot and an AI chatbot to go with it. Yeah. Uh, 0:16:57.370 --> 0:17:00.010 even that's starting to lose its its shine now, and 0:17:00.010 --> 0:17:02.380 they're trying to get a lot more around help me 0:17:02.380 --> 0:17:05.230 automate the fix. So I think that that's a pretty neat, uh, 0:17:05.230 --> 0:17:09.400 iteration of it. Now, it's no longer just purely kneejerk, uh, 0:17:09.400 --> 0:17:12.760 you know, cover my legal basis. It's let me actually 0:17:12.760 --> 0:17:16.450 make appsec or things like that, uh, faster and more, uh, 0:17:16.570 --> 0:17:20.830 more concrete for all the parties involved. Uh, so I think, uh, 0:17:20.830 --> 0:17:25.659 that one is really interesting. Um. And then kind of 0:17:25.660 --> 0:17:28.359 at the same, the same time, uh, a kind of 0:17:28.359 --> 0:17:30.790 a smaller space that people don't really think about as 0:17:30.790 --> 0:17:34.060 much that I find really interesting now is the explosion 0:17:34.060 --> 0:17:36.969 of trust and safety, because, like, it's so much more 0:17:36.970 --> 0:17:40.480 important now to even understand which videos, which content, which 0:17:40.480 --> 0:17:43.210 anything you're consuming is real or not. And it's it's 0:17:43.210 --> 0:17:46.480 more than just a social media problem. It's becoming a 0:17:46.480 --> 0:17:48.669 company problem as well. Like it's a brand. It's a 0:17:48.670 --> 0:17:52.959 reputational problem. Um, that extends beyond just somebody making a 0:17:52.960 --> 0:17:56.020 deepfake video. So I think that's there's a lot of 0:17:56.020 --> 0:17:59.170 going on that. And then you add on geopolitical contexts 0:17:59.170 --> 0:18:01.030 and like wars on top of that stuff. And it's 0:18:01.030 --> 0:18:04.420 it's very interesting, um, like how you, how you sort 0:18:04.420 --> 0:18:06.370 all that out. And I think this is a place 0:18:06.369 --> 0:18:07.810 where I can really help. 0:18:08.869 --> 0:18:13.760 Yeah. Interesting one section I think's going to absolutely blow up. 0:18:13.760 --> 0:18:16.790 And I don't think um, I don't think there are 0:18:16.790 --> 0:18:19.910 many people playing here yet, but actually there's a decent 0:18:19.910 --> 0:18:28.879 number like, um, validation. Of ground. Truth is a space 0:18:28.880 --> 0:18:33.560 that I think needs to, well, is just going to, uh, 0:18:33.560 --> 0:18:38.159 blow up. So, so I, I see this as. Um, 0:18:38.160 --> 0:18:40.949 me and you having some sort of, like device. It 0:18:40.950 --> 0:18:42.540 might be on our phones. It probably will be on 0:18:42.540 --> 0:18:45.150 our phones because it's expensive to carry an extra device. 0:18:45.150 --> 0:18:49.080 So basically, you and I have a pre-existing, like, uh, 0:18:49.080 --> 0:18:53.190 signal connection. Mhm. So we go to get on this, 0:18:53.190 --> 0:18:56.670 this call right here. And um, like I've seen you 0:18:56.670 --> 0:18:59.340 move your hand when you move your hand also with me, 0:18:59.340 --> 0:19:03.149 you see, see the green glitching. Yeah. The green screen. 0:19:03.150 --> 0:19:10.170 Same happens with you. Um, and when, um, we get 0:19:10.170 --> 0:19:13.419 ready to do a call. I think we're going to 0:19:13.420 --> 0:19:18.760 do something like. Okay. Initiate, uh, pre-shared key connection or something. 0:19:18.760 --> 0:19:23.440 And you put up actually on your, on your screen 0:19:23.440 --> 0:19:27.160 and show a code which is corresponds to the code 0:19:27.160 --> 0:19:29.920 I have, which is part of the Pre-shared. Uh, now 0:19:29.920 --> 0:19:33.580 it's part of the video. As well. And so now 0:19:33.580 --> 0:19:38.050 we've validated that this video feed. With a combination of 0:19:38.050 --> 0:19:42.220 live and a combination of pre-shared. We have an actual 0:19:42.220 --> 0:19:46.960 connection and you're not a deepfake. Um, that SpaceX is 0:19:46.960 --> 0:19:49.030 going to absolutely heat up. The other one is going 0:19:49.030 --> 0:19:55.260 to get crazy is, um, AI agents acting as humans? Um, 0:19:55.440 --> 0:19:59.940 and calling APIs and everything, and basically IAM type companies 0:19:59.940 --> 0:20:03.690 or identity type companies basically getting a handle on who's 0:20:03.690 --> 0:20:06.540 human versus who's emulating human. Mhm. 0:20:07.630 --> 0:20:10.209 Interesting. Yeah, I think that all makes a lot of 0:20:10.210 --> 0:20:12.730 sense too. And it kind of plays into one of 0:20:12.730 --> 0:20:15.730 the earlier posts you've written about that. I don't think 0:20:15.730 --> 0:20:18.730 a lot of people have, uh, really, uh, done too 0:20:18.730 --> 0:20:21.070 much with or like looked too far into. But it's 0:20:21.070 --> 0:20:24.790 really you've got AI security, you've got security of AI, 0:20:24.820 --> 0:20:27.490 you've got security for AI, but you don't have anybody 0:20:27.490 --> 0:20:31.300 really talking about security from AI other than deep fakes. 0:20:31.300 --> 0:20:33.460 But like, that's just like the tip of the iceberg 0:20:33.460 --> 0:20:35.590 where this could go and what, you know, could be 0:20:35.590 --> 0:20:39.250 launched against you or as a person or as a company. Um, 0:20:39.250 --> 0:20:43.000 so I think that space, obviously it'll include misinformation and 0:20:43.000 --> 0:20:46.120 disinformation kind of signals, but there's got to be more 0:20:46.119 --> 0:20:49.030 than that. Like what else is being undertaken against me 0:20:49.030 --> 0:20:53.350 from a personal like an OpSec standpoint, almost. Um, so 0:20:53.350 --> 0:20:55.209 I think that space is going to just going to 0:20:55.420 --> 0:20:57.490 as people get more, uh, attuned to this and they 0:20:57.490 --> 0:21:00.670 get more paranoid, I think that's going to become, uh, um, 0:21:00.670 --> 0:21:04.330 have a greater impact on the world of privacy than 0:21:04.330 --> 0:21:08.020 privacy or regulation ever will, if that makes sense. 0:21:08.020 --> 0:21:12.129 Yeah. Yeah, I think that's right. Yeah. One way I 0:21:12.130 --> 0:21:15.190 like to think about that is like a, uh, and 0:21:15.190 --> 0:21:18.820 this is a little bit in the future according to like, this, this, uh, 0:21:19.690 --> 0:21:25.210 plan or idea that I have, but it's like a continuous, uh, defender. 0:21:25.480 --> 0:21:28.630 So you essentially have like a digital assistant, uh, which 0:21:28.630 --> 0:21:30.820 is what I call, like the permanent AI that you 0:21:30.820 --> 0:21:35.080 have with you. And it's a continuous filter for everything 0:21:35.080 --> 0:21:38.770 coming in to see you. Um, um, and it could 0:21:38.770 --> 0:21:42.879 actually detect when you're being, like, overly flattered for the 0:21:42.880 --> 0:21:46.360 purpose of manipulation because they're in sales. Yeah. Um, so 0:21:46.359 --> 0:21:48.550 one's like, oh, yeah. You know, that outfit looks really 0:21:48.550 --> 0:21:50.290 great on you. And by the way, I love that 0:21:50.290 --> 0:21:53.649 talk that you gave. I especially love slide number seven 0:21:53.890 --> 0:21:57.970 from your talk in Berlin. And, uh, and suddenly you're like, 0:21:57.970 --> 0:22:00.189 oh wow. I just feel really cool. And then they 0:22:00.220 --> 0:22:02.860 hit you with the sales pitch, right? Right. And so 0:22:02.859 --> 0:22:05.350 you could have something pop up in your glasses or whatever. 0:22:05.350 --> 0:22:11.260 It's like NLP detected manipulation, uh, expect inbound sales pitch 0:22:11.260 --> 0:22:14.380 or something like that, right? Or if, um, you're getting like, 0:22:14.380 --> 0:22:18.250 this really threatening language that's coming at you, it might 0:22:18.250 --> 0:22:22.449 actually just rewrite that for you. Um, so, uh, one example, 0:22:22.450 --> 0:22:26.830 I was talking with a friend named Joel the other night, and, um, 0:22:27.640 --> 0:22:31.060 an example of this would be like, say, someone's been 0:22:31.060 --> 0:22:34.810 traumatized by a partner and the partner is constantly emailing 0:22:34.810 --> 0:22:38.290 them or texting them or something. And the the texts 0:22:38.290 --> 0:22:42.399 are like they're life disrupting. It's like a paragraph of, like, 0:22:42.400 --> 0:22:47.739 nastiness and narcissism and like gaslighting and whatever, under the 0:22:47.740 --> 0:22:50.739 pretense that they're actually agreeing when to pick up the 0:22:50.740 --> 0:22:56.740 kid with shared custody, your AI just rewrites it. Pick 0:22:56.740 --> 0:23:02.050 up Jimmy at for. The whole paragraph is gone. It 0:23:02.050 --> 0:23:06.040 just leaves the content so it doesn't. It strips out 0:23:06.040 --> 0:23:08.740 all the triggers and everything. So I think that I 0:23:08.740 --> 0:23:11.710 filtering thing is really powerful. Yeah. 0:23:11.830 --> 0:23:13.990 Well that makes a lot of sense. And it kind 0:23:13.990 --> 0:23:16.900 of plays into the whole like, uh, personal digital sovereignty 0:23:16.900 --> 0:23:19.000 and like being able to control that aspect of you 0:23:19.000 --> 0:23:22.300 online and offline. Uh, that makes a lot of sense. 0:23:22.660 --> 0:23:25.690 Yeah. Yeah. And the other one that I think is 0:23:25.690 --> 0:23:30.320 really interesting is like, um. We're going to have to 0:23:30.320 --> 0:23:35.480 validate true content. This one I've not seen a space 0:23:35.480 --> 0:23:38.630 really heat up yet with either, and maybe because it's 0:23:38.630 --> 0:23:44.960 just so hard, but, um, essentially watermarking and validation like 0:23:44.960 --> 0:23:49.790 trusted feeds so that um, for, for example, I love 0:23:49.790 --> 0:23:53.120 UFC stuff. I'm all into fighting stuff. So when I 0:23:53.119 --> 0:23:56.780 watch fight videos now, I'm having this really weird experience 0:23:56.780 --> 0:24:03.950 where I'm having fight videos narrated by Joe Rogan. And 0:24:04.280 --> 0:24:07.640 I'm like watching the fight video and I'm enjoying it 0:24:07.640 --> 0:24:11.930 and I'm like three, four minutes in before I finally 0:24:11.930 --> 0:24:16.979 catch some weird word. And then I realized for the 0:24:16.980 --> 0:24:23.690 entire four minutes. Joe Rogan is a multi-millionaire. I am 0:24:23.690 --> 0:24:27.530 on a very small YouTube channel. Um, he didn't come 0:24:27.530 --> 0:24:31.700 to this channel and narrate this. Interesting. This person went 0:24:31.700 --> 0:24:35.810 and collected all the clips and he just got the voice. 0:24:35.810 --> 0:24:39.950 He wrote the script. Um, or maybe even I wrote 0:24:39.950 --> 0:24:43.909 the script, but it's all being set in Joe Rogan's voice. Um, 0:24:43.910 --> 0:24:47.000 which is fairly harmless. I mean, it makes the videos 0:24:47.000 --> 0:24:51.710 quite nice, uh, quite enjoyable, but, like, what happens when 0:24:51.710 --> 0:24:55.700 that's actually news instead, right? What happens when that's something 0:24:55.700 --> 0:24:59.720 really important? Um, or if it's like a Joe probably 0:24:59.720 --> 0:25:02.810 doesn't care. Maybe he does. I doubt he does. But 0:25:02.810 --> 0:25:06.709 a lot of, uh, Hollywood celebrities actually would, and definitely 0:25:06.710 --> 0:25:11.510 government officials really would. So, um, that thing should either 0:25:11.510 --> 0:25:17.300 either have a fake symbol or lack an authentic symbol. Right? 0:25:17.300 --> 0:25:21.710 Whereas the official symbol coming out of, uh, Rogan's feeds. 0:25:23.170 --> 0:25:26.619 Should have the have the symbol right or lack the 0:25:26.619 --> 0:25:29.530 fake symbol. That is an entire industry that's going to 0:25:29.530 --> 0:25:33.700 be a multi-billion dollar industry. Um, yeah. And I'm sort 0:25:33.700 --> 0:25:35.260 of just waiting for that to happen. 0:25:35.890 --> 0:25:38.410 I think that's a really good concept too. And I've, 0:25:38.440 --> 0:25:40.600 you know, I've talked to actually a few startups who 0:25:40.600 --> 0:25:44.110 are doing that. Exactly. But they're they're trying to basically 0:25:44.109 --> 0:25:50.260 create truth out of multiple sources for news reports. And 0:25:50.260 --> 0:25:52.630 the goal of this, you give it to corporate communications 0:25:52.630 --> 0:25:56.170 teams to let them decipher what's real and what's not. 0:25:56.170 --> 0:25:58.630 Like who actually said this thing. And it really comes 0:25:58.630 --> 0:26:02.170 down to using AI to go find, uh, the original 0:26:02.170 --> 0:26:04.659 sources or the original, and the original shares the original 0:26:04.660 --> 0:26:07.000 posts and then tie it back, kind of like to 0:26:07.000 --> 0:26:09.939 information actors on the back end to say, oh, this, 0:26:09.940 --> 0:26:12.910 this post that was widely shared by this person on 0:26:12.910 --> 0:26:17.859 Facebook originated from a Russian disinformation mill. This person is 0:26:17.859 --> 0:26:19.869 also associated with these 12 other things. It has a 0:26:19.869 --> 0:26:22.390 high degree of chance that this is all fake. And 0:26:22.390 --> 0:26:23.979 so then it gives you like kind of risk scores. 0:26:23.980 --> 0:26:27.460 Then you can say corporate cops can say, we know 0:26:27.460 --> 0:26:29.830 this is fake. Uh, next one, you know, so they 0:26:29.830 --> 0:26:32.230 can they can move on and they can then have their, 0:26:32.230 --> 0:26:35.379 you know, just kind of situational awareness per, per company 0:26:35.380 --> 0:26:37.600 around that. But then that could that could easily extend 0:26:37.600 --> 0:26:41.290 to like video contexts as well. Um, but it's it's 0:26:41.290 --> 0:26:44.230 fascinating a time to be in this space just because it's, 0:26:44.619 --> 0:26:48.820 I think the, the, uh, technology hype cycle and the 0:26:48.820 --> 0:26:52.600 curve has it occurred it occurred like every other, you know, 0:26:52.600 --> 0:26:55.360 platform we've seen with like, mobile and cloud, which might 0:26:55.359 --> 0:26:57.280 have taken ten years. But this a lot of this 0:26:57.280 --> 0:26:59.500 has happened in like 18 months. And it's gone through 0:26:59.500 --> 0:27:02.560 a couple iterations of that loop. Um, so it's just 0:27:02.560 --> 0:27:04.510 so hard to keep up with it all, uh, and 0:27:04.510 --> 0:27:06.070 know what you're actually looking at. 0:27:06.550 --> 0:27:11.109 Yeah. I don't think it's even close to where it's going. 0:27:11.109 --> 0:27:14.649 It's it's not even close. I mean, I'm very much 0:27:14.650 --> 0:27:19.810 I'm a self-described, uh, very pro AI, but I feel 0:27:19.810 --> 0:27:24.459 like I see this really clearly, and we're barely getting started. Um, 0:27:24.460 --> 0:27:28.330 so when I hear about AI hype and there, you know, 0:27:28.330 --> 0:27:33.600 there's some places where, uh. People got over their skis 0:27:33.600 --> 0:27:36.360 and they got overly excited about a very small thing, 0:27:36.359 --> 0:27:41.379 like a giant company that generates. I don't know. Uh, 0:27:41.380 --> 0:27:44.770 Midjourney images and sells them or something. That would be 0:27:44.770 --> 0:27:49.090 more crypto like, or more NFT like, where it's way 0:27:49.090 --> 0:27:53.920 over inflated for what it is, right? Right. Whereas most 0:27:53.920 --> 0:27:56.980 of I, especially the stuff we've just been talking about, 0:27:56.980 --> 0:28:00.670 hasn't even started yet. And it's not even possible without AI. 0:28:01.240 --> 0:28:04.030 The hype cycle is not even. It's not that it 0:28:04.030 --> 0:28:06.670 hasn't got to the top. It hasn't even started yet. 0:28:07.000 --> 0:28:10.480 And that's not counting for like the real, the real 0:28:10.480 --> 0:28:15.820 core stuff where all of your entire corporate data is 0:28:15.820 --> 0:28:20.460 inside of a data lake. That I can see. And 0:28:20.460 --> 0:28:22.890 now you could basically ask any question and take any 0:28:22.890 --> 0:28:26.040 action as a result of that, which is kind of 0:28:26.040 --> 0:28:29.100 the long term thing. And people haven't even really started 0:28:29.100 --> 0:28:33.179 doing that because of the limitations of context. Windows and Rag. 0:28:33.180 --> 0:28:36.750 It's like that's all still building up. There's there's a 0:28:36.750 --> 0:28:41.160 few hype phases that are kind of really small and inconsequential. 0:28:41.160 --> 0:28:45.690 But the ultimate S-curve of AI, I think, is just starting. 0:28:46.620 --> 0:28:50.070 Yeah. No, I totally agree. And I think, um, you know, 0:28:50.070 --> 0:28:52.080 we it's one of those things you can't even imagine 0:28:52.080 --> 0:28:55.560 it because we didn't even imagine generally available. I on 0:28:55.560 --> 0:28:58.230 your phone like 18 months ago, at least the average 0:28:58.230 --> 0:29:02.280 person didn't. Yeah. Um, and now the average tech person, like, 0:29:02.280 --> 0:29:05.489 uses it multiple times a week. Um, and even the 0:29:05.490 --> 0:29:08.640 average non tech person is starting to at least know 0:29:08.640 --> 0:29:11.250 what it is in general. Um, which is, which is 0:29:11.250 --> 0:29:14.220 pretty impressive. Um, but yeah, we're just kind of scratching 0:29:14.220 --> 0:29:17.820 the surface on what's, what's possible there, which I think 0:29:18.240 --> 0:29:20.610 it's going to take kind of like a step function to, 0:29:20.640 --> 0:29:24.420 to get that next level of AI security kind of realm. Um, 0:29:24.420 --> 0:29:27.030 you know, every investor I talked to has that as 0:29:27.030 --> 0:29:28.890 part of their core thesis. Now they they have to 0:29:28.890 --> 0:29:33.300 do AI, some variation of security because nobody knows where 0:29:33.300 --> 0:29:35.100 the rocket ships going, but they know it's a rocket ship. 0:29:35.100 --> 0:29:38.220 So they, they want to latch on in some way. Yeah. Um, 0:29:38.700 --> 0:29:41.010 and uh, it's, it is going to make some really 0:29:41.010 --> 0:29:44.400 cool stuff possible that otherwise wouldn't be. And I think, uh, 0:29:44.400 --> 0:29:46.770 I also would not be surprised in terms of trends 0:29:46.770 --> 0:29:49.020 to see like the, you know, the 1 to 2 0:29:49.020 --> 0:29:53.700 person security unicorn company. Uh, yes. Like they oh, you 0:29:53.700 --> 0:29:56.580 can just build out your own complete attack surface management 0:29:56.580 --> 0:30:02.100 platform using AI and self-hosting it or, um, and as 0:30:02.100 --> 0:30:04.560 long as you can, uh, then learn the sales motions 0:30:04.560 --> 0:30:06.570 and get that going, you can get bootstrapped, you can 0:30:06.570 --> 0:30:09.510 get going, you can get funding, um, or may not 0:30:09.510 --> 0:30:11.910 even need funding. So I think like it's going to be, um, 0:30:12.540 --> 0:30:14.940 a pretty open world. Uh, which would be better, I 0:30:14.940 --> 0:30:17.760 think for like, you know, competition would be better for 0:30:17.760 --> 0:30:21.900 practitioners eventually. Um, because all that stuff is constantly expanding 0:30:21.900 --> 0:30:24.900 and collapsing in the market, like, because security is not 0:30:24.900 --> 0:30:28.080 a winner take all market at all. Like from a 0:30:28.080 --> 0:30:32.790 product standpoint, it's a many, many people can be winners 0:30:32.790 --> 0:30:36.030 in this space as well. Um, even if some of 0:30:36.030 --> 0:30:37.890 the players are trying to act in a zero sum way, 0:30:37.890 --> 0:30:42.600 like there's there's not a singular like dominant player across 0:30:42.600 --> 0:30:47.700 all fields. There's like pockets of dominance across multiple fields. Um, 0:30:47.700 --> 0:30:49.890 so I think it's just going to make that bubble 0:30:49.890 --> 0:30:53.160 even bigger and maybe, maybe more challenging for practitioners for 0:30:53.160 --> 0:30:55.440 a while. I hope I can help with that with security, 0:30:55.440 --> 0:30:57.630 like make a little more sense of that. But, um, 0:30:57.960 --> 0:30:59.220 it's going to be interesting. 0:31:00.410 --> 0:31:03.140 Yeah, I think you're absolutely right. 0:31:03.170 --> 0:31:05.840 There's the there's the big trend of, uh, you know, 0:31:05.840 --> 0:31:10.430 personal digital sovereignty, like that kind of concept of being 0:31:10.430 --> 0:31:13.340 security secure from AI and all the kind of attacks 0:31:13.340 --> 0:31:16.550 on you as well. Um, there's going to have to 0:31:16.550 --> 0:31:21.290 be some more iterations around, uh, improvements in AI security 0:31:21.290 --> 0:31:23.300 in general for it to be like the next iteration 0:31:23.300 --> 0:31:26.420 for like for to be beyond just like I think 0:31:26.420 --> 0:31:29.000 a lot of what we've seen right now is just been, uh, 0:31:29.000 --> 0:31:33.530 adding AI to existing capabilities as opposed to AI creating 0:31:33.530 --> 0:31:37.070 the new capability. Uh, so we're not quite there around 0:31:37.070 --> 0:31:39.890 that curve yet. And then once whoever makes that next 0:31:39.890 --> 0:31:44.570 step is going to, like, do really well. Um, but 0:31:44.570 --> 0:31:47.480 it's obviously there's, uh, there's a context and there's a 0:31:47.480 --> 0:31:50.660 data problem that has just got just expanded now with the, 0:31:50.660 --> 0:31:56.120 with AI. Um, because more, more AI means more cloud workloads, 0:31:56.120 --> 0:31:59.180 which means more tech service means more configuration mismanagement. All 0:31:59.180 --> 0:32:04.070 these things are just have have expanded. Um, like you said, I, uh, 0:32:04.070 --> 0:32:09.890 exaggerates the thing or highlights the thing. It isn't the thing. Yeah. 0:32:11.540 --> 0:32:15.620 Yeah, I like what you said about a single person 0:32:15.620 --> 0:32:20.120 company launching and getting really big because you mentioned like, oh, 0:32:20.240 --> 0:32:22.820 as long as they can learn the sales motion well, 0:32:22.820 --> 0:32:24.680 what if they don't have to because they just hire 0:32:24.680 --> 0:32:28.580 an AI company? Yeah, that's automated sales. Right. And then 0:32:28.580 --> 0:32:32.180 you hire a company that's automated support and then you've 0:32:32.180 --> 0:32:36.650 got automated like uh, EA's. So you've got like your 0:32:36.650 --> 0:32:40.520 personal assistants and basically all of your company infrastructure is 0:32:40.520 --> 0:32:46.380 kind of mostly outsourced. Service companies that are AI based. 0:32:46.380 --> 0:32:50.190 And then you basically have, like yourself, a co-founder and 0:32:50.190 --> 0:32:53.790 maybe 1 or 2 people. Yeah. And maybe it's just you. 0:32:53.790 --> 0:32:56.490 And like one person I would say is probably one 0:32:56.490 --> 0:33:00.180 of the smallest versions. I'm sure there will be actual 0:33:00.180 --> 0:33:04.950 single founders who make pretty big companies as well. Um, 0:33:04.950 --> 0:33:08.700 Altman's been talking about that as well as the first unicorn. 0:33:08.790 --> 0:33:13.770 One person. Yeah. So I think that's very possible. Um. 0:33:16.790 --> 0:33:23.870 What about the balance of private equity VC versus startups like. 0:33:25.240 --> 0:33:29.830 I feel like this last down period. Oh, by the way, 0:33:29.830 --> 0:33:33.850 what are the years for these? These periods? 21 was 0:33:33.850 --> 0:33:35.710 that the start of the down? 0:33:36.100 --> 0:33:40.810 So 22 was like the start of the down basically. Okay. Yeah. 0:33:41.110 --> 0:33:45.310 22 we started the down 23 was down and 24 0:33:45.340 --> 0:33:46.930 were starting to see the uptick. 0:33:47.020 --> 0:33:50.350 But yeah a bit of a rebound. Um, and um, 0:33:50.530 --> 0:33:54.250 you know 2021 late 21 had a few signs of distress, 0:33:54.250 --> 0:33:56.410 but not like no one thought the fed would raise 0:33:56.410 --> 0:33:59.590 the rates so high. Yeah. Or keep on raising it. 0:33:59.590 --> 0:34:02.739 And like the so that it just kind of had 0:34:02.740 --> 0:34:05.290 like a an iterative effect there and just like kind 0:34:05.290 --> 0:34:06.730 of kept making things worse. 0:34:07.300 --> 0:34:12.910 Right. So 22 and 23, uh, private equity did really 0:34:12.910 --> 0:34:15.940 well because they're swooping in, grabbing things. Right? 0:34:16.210 --> 0:34:16.810 Yep. 0:34:16.810 --> 0:34:20.110 And VCs did not do so well. Startups didn't do 0:34:20.110 --> 0:34:23.799 so well. And now there's a whole bunch of companies 0:34:23.800 --> 0:34:27.070 that are kind of sitting out there. They've been growing 0:34:27.070 --> 0:34:31.150 slowly and they're kind of large and bloated. Mhm. I'm 0:34:31.150 --> 0:34:33.640 sure you've seen a bunch of these and they're just 0:34:33.640 --> 0:34:36.400 kind of like stagnating. And every time they have to 0:34:36.400 --> 0:34:39.250 go to investors, investors are like tired of hearing from them. 0:34:39.250 --> 0:34:42.580 And they're like yep you got to do something like 0:34:42.580 --> 0:34:45.580 I've heard the same thing. You're not growing. It's not 0:34:45.580 --> 0:34:50.920 super exciting. Um, what does this look like as things 0:34:50.920 --> 0:34:55.210 start to boom now? And I think it's going to accelerate, 0:34:55.210 --> 0:34:58.900 but who knows. But as it starts to boom, who 0:34:58.900 --> 0:35:02.890 starts to benefit? Uh, I guess the private equity benefits, 0:35:02.890 --> 0:35:06.160 if the stuff they bought cheap starts to be worth more, 0:35:06.190 --> 0:35:10.239 that's pretty obvious. But I guess VC and startups are 0:35:10.239 --> 0:35:13.510 the main benefits if things start going up. Is that right? 0:35:14.030 --> 0:35:16.790 Yeah. I mean that's that's exactly right. So like, if, 0:35:16.790 --> 0:35:18.680 you know, if the bets the P made at the 0:35:18.680 --> 0:35:21.200 acquisitions and like the takeovers they made and they and 0:35:21.200 --> 0:35:23.569 they and by the way, P has been very good 0:35:23.570 --> 0:35:27.350 at buying companies and making them super operationally efficient for 0:35:27.350 --> 0:35:29.090 a couple of years, getting a return on their investment 0:35:29.090 --> 0:35:31.969 and selling it. Uh, they just had a much better 0:35:31.969 --> 0:35:33.980 time doing that, you know, the past couple of years 0:35:33.980 --> 0:35:37.339 than they had before. Um, you know, VCs, a lot 0:35:37.340 --> 0:35:39.980 of when it comes to, uh, venture investing, you know, 0:35:39.980 --> 0:35:43.310 most investments don't make money at all. Uh, and typically, 0:35:43.310 --> 0:35:45.530 only a very small portion of what you invest in 0:35:45.530 --> 0:35:48.860 actually makes up 80 or 90% of the the value 0:35:48.860 --> 0:35:53.180 that you get back. Um, and so, uh, some did 0:35:53.180 --> 0:35:55.730 really well, uh, at the time and some and, but 0:35:55.730 --> 0:35:57.980 a lot of them did not. So like they, they 0:35:57.980 --> 0:36:01.940 struggled their companies shut down, uh, funds shut down. Uh, 0:36:01.940 --> 0:36:05.690 they didn't return any money to their investor base. Um, 0:36:05.690 --> 0:36:07.489 so that's all bad. And that makes it harder to 0:36:07.489 --> 0:36:09.470 raise more money next time, or get people to commit 0:36:09.469 --> 0:36:12.380 to give you money to go invest next time. Um, 0:36:12.710 --> 0:36:17.210 you know, there's a couple carry overs that will be constant. Um, 0:36:17.210 --> 0:36:21.049 we're we're too close to when operational efficiency was the 0:36:21.050 --> 0:36:24.050 king to go away from it now. So it's like, yeah, 0:36:24.050 --> 0:36:25.670 we might be bouncing back, but you got to be 0:36:25.670 --> 0:36:29.120 a lean bounce back. You cannot be overinflated. I agree you, 0:36:29.360 --> 0:36:33.080 you never hire at 200, you know, percent or like bro, 0:36:33.110 --> 0:36:35.090 you know 1,000%. Like the memory. 0:36:35.090 --> 0:36:36.140 Is too fresh. 0:36:36.350 --> 0:36:37.969 Yeah, it's too fresh. Like, I think we have to 0:36:37.969 --> 0:36:40.250 have another 4 or 5 years of, like, a bull 0:36:40.250 --> 0:36:42.800 run for them to say, yeah, forget that. Let's just 0:36:42.800 --> 0:36:45.560 go big again and like, uh, but you also have 0:36:45.560 --> 0:36:48.650 to remember at the same time, the goal posts have 0:36:48.650 --> 0:36:51.860 been significantly moved on both ends of the spectrum. So 0:36:51.860 --> 0:36:55.940 now you have to have less people be operationally efficient 0:36:55.940 --> 0:36:58.670 and burn less cash. So that's bottom in the spectrum. 0:36:58.670 --> 0:37:00.920 At the same end, if you want to go public, 0:37:00.920 --> 0:37:03.109 it's a lot more of a higher barrier now. So 0:37:03.110 --> 0:37:05.850 it's not you can't just be a $1, a 1 0:37:05.850 --> 0:37:10.700 billion uh, valuation. You need to be three, $4 billion valuation. 0:37:10.700 --> 0:37:12.950 You can't just be a $100 million in RR, which 0:37:12.950 --> 0:37:15.980 used to be the magic number to make companies go public, 0:37:15.980 --> 0:37:18.469 or at least start the roadshow. Now you need to 0:37:18.469 --> 0:37:21.469 be 200 or 300 million. And so like the pole 0:37:21.469 --> 0:37:23.930 or like the, the the narrowing of the pyramid is 0:37:23.930 --> 0:37:26.630 like gotten like even tighter now. So you have a 0:37:26.630 --> 0:37:31.190 lot more competition, which means sink or swim is like 0:37:31.190 --> 0:37:34.310 a lot more aggressive. Um, and so that's why everyone 0:37:34.310 --> 0:37:37.160 has kind of been decrying, oh, platform ization is finally 0:37:37.160 --> 0:37:40.129 happening in cybersecurity. Like we're finally going to have less vendors, 0:37:40.130 --> 0:37:42.859 which is will never, ever be true. Um, it'll just 0:37:42.860 --> 0:37:47.569 go through pockets of of consolidation. Always like to explain 0:37:47.570 --> 0:37:50.300 it as, uh, it's more of an accordion like parts 0:37:50.300 --> 0:37:52.880 of it are open and like expanding and parts of 0:37:52.880 --> 0:37:55.370 it are closed and contracting, but you're still like you're 0:37:55.370 --> 0:37:59.330 making music, but you just like but it's not going 0:37:59.330 --> 0:38:02.150 to look like, you know, one company, Palo Alto, is 0:38:02.150 --> 0:38:03.770 not going to rule them all, even if you want 0:38:03.770 --> 0:38:07.880 them to, um, tomorrow and like so this they will 0:38:07.880 --> 0:38:11.090 never be that one. True. Uh, so it's now become 0:38:11.090 --> 0:38:15.859 harder to, to do that. Um, and then all the while, like, 0:38:15.860 --> 0:38:17.239 private equity is still going to have a pretty good 0:38:17.239 --> 0:38:18.620 run for the next 3 or 4 years, because a 0:38:18.620 --> 0:38:21.610 lot of those companies who raise money in 2020 or 0:38:21.610 --> 0:38:25.010 2019 or even 2021, some of them are still just 0:38:25.010 --> 0:38:27.350 limping on, and they may have another a few more 0:38:27.350 --> 0:38:30.530 years of runway before they need to go private. Like 0:38:30.530 --> 0:38:32.900 they may have raised too much money, like some of 0:38:32.900 --> 0:38:36.020 these companies raised close or more than $1 billion in 0:38:36.020 --> 0:38:39.620 funding may not have anywhere else to go. Uh, and 0:38:39.620 --> 0:38:41.660 less like a Cisco or like a, you know, a 0:38:41.660 --> 0:38:45.020 Microsoft buys them, uh, so either they'll die a slow 0:38:45.020 --> 0:38:47.780 death or they'll go take, you know, they'll private equity 0:38:47.780 --> 0:38:50.690 or go shape them up and, you know, possibly dismantle them, 0:38:50.690 --> 0:38:54.860 possibly resell them. Um, so there's a lot of more 0:38:54.860 --> 0:38:57.680 complexity now, I think, than there used to be, uh, 0:38:57.680 --> 0:38:59.779 in terms of, like, making a go of it. 0:39:01.120 --> 0:39:06.879 Yeah. Interesting. I really like your accordion thing. I think 0:39:06.880 --> 0:39:11.140 that makes a lot of sense. I do think platform 0:39:11.140 --> 0:39:16.880 migration is absolutely a thing, I think. That's like the 0:39:16.880 --> 0:39:19.160 permanent home for anything that I used to have this 0:39:19.160 --> 0:39:27.920 thing where it was like, idea. Um, then, um, application, then, um, 0:39:28.550 --> 0:39:32.719 operating system. So it's like things move through that phase 0:39:32.719 --> 0:39:35.630 where it's like, uh, it's like a web app or 0:39:35.630 --> 0:39:38.089 a little utility or something, and then it gets brought 0:39:38.090 --> 0:39:41.390 into the OS and it's kind of absorbed the the 0:39:41.390 --> 0:39:45.380 knowledge is now so known and understood that it just 0:39:45.380 --> 0:39:46.940 becomes part of the furniture. 0:39:47.450 --> 0:39:48.290 Right? 0:39:48.440 --> 0:39:51.890 Um, but especially with AI. So I think one of 0:39:51.890 --> 0:39:55.250 the biggest things happening with AI is the fact that 0:39:55.250 --> 0:39:58.370 it's so much easier to start a company and actually 0:39:58.370 --> 0:40:02.299 be effective, kind of going back to the whole, um, 0:40:02.300 --> 0:40:05.719 unicorn thing. But forget about the unicorns. Like I'm about 0:40:05.719 --> 0:40:10.010 to launch a SaaS myself and it's just me. And 0:40:10.010 --> 0:40:11.390 like a dev. 0:40:11.860 --> 0:40:12.610 Nice. 0:40:12.610 --> 0:40:15.340 And it's a full size, right? It's a full size. 0:40:15.340 --> 0:40:17.870 It's like end to end like it. And there's a 0:40:17.870 --> 0:40:21.100 million people like me who are doing this because of AI. 0:40:21.219 --> 0:40:24.910 So that is the accordion, but the bottom of the 0:40:24.910 --> 0:40:29.920 accordion or whatever. The new ideas, the new companies, the 0:40:29.920 --> 0:40:33.430 new sauces that are out there for the platforms to 0:40:33.430 --> 0:40:37.180 pick and choose which features they want via acqui hire. 0:40:38.050 --> 0:40:38.500 Is. 0:40:38.500 --> 0:40:42.130 Going to be much larger because I think the barrier 0:40:42.130 --> 0:40:45.219 to get in to be a startup person five years 0:40:45.219 --> 0:40:47.200 ago is way higher than it's going to be in 0:40:47.200 --> 0:40:48.100 two years from now. 0:40:49.190 --> 0:40:52.339 Yeah, I think that's true. And I think, um, I 0:40:52.340 --> 0:40:54.590 agree with everything you said too. And I think, you know, 0:40:54.590 --> 0:40:57.200 the concept of bootstrapping or just doing it on your own, 0:40:57.200 --> 0:41:00.890 with your with your friends and family support, uh, is 0:41:00.890 --> 0:41:03.980 going to become a lot more realistic, um, than it 0:41:03.980 --> 0:41:05.690 used to be. Like, you see this in other industries, 0:41:05.690 --> 0:41:08.000 but you don't really see it much in cybersecurity. Like 0:41:08.000 --> 0:41:09.920 maybe like think of, like canaries, like kind of the 0:41:09.920 --> 0:41:12.620 only one I could think of that's, uh, truly like, 0:41:12.620 --> 0:41:16.109 kind of bootstrapped, but like most. Uh, you know, it's 0:41:16.110 --> 0:41:19.950 hard to get behind that model. Um, and I think, 0:41:19.950 --> 0:41:21.719 you know, a lot of things that I've been trying 0:41:21.719 --> 0:41:24.569 to also help people kind of realize the return on 0:41:24.570 --> 0:41:27.600 security is that, like, you know, understanding, like who the 0:41:27.600 --> 0:41:31.890 investments or investors are behind a given, you know, company 0:41:31.890 --> 0:41:34.410 you can kind of see, like repeatability and like track 0:41:34.410 --> 0:41:37.710 records of success or like playbooks they use often. And 0:41:37.710 --> 0:41:39.900 you kind of say like, okay, I know that it's 0:41:39.900 --> 0:41:42.570 going to, you know, backed by this company, their goal 0:41:42.570 --> 0:41:46.020 is to be acquired by a bigger company. That means 0:41:46.020 --> 0:41:48.630 if I invest in this now, it may not be 0:41:48.630 --> 0:41:51.030 here in four years, three years or whatever, whatever contract 0:41:51.030 --> 0:41:52.590 they're trying to get me to sign. So like, I 0:41:52.590 --> 0:41:55.290 kind of need to think about that, um, based on 0:41:55.290 --> 0:41:59.440 kind of where it's come from. Um. But it's. I 0:41:59.440 --> 0:42:01.570 agree that it'll be easier to get in a startup, 0:42:01.570 --> 0:42:03.340 it'll be easier to get a lot more traction and 0:42:03.340 --> 0:42:06.550 actually get paying customers, especially if you then go back 0:42:06.550 --> 0:42:08.859 to like using AI to enhance what you know or like, 0:42:08.860 --> 0:42:11.080 fill in the gaps of what you can't, you know, 0:42:11.140 --> 0:42:14.049 do where you couldn't have done before is easily like 0:42:14.050 --> 0:42:16.240 it's yeah, it's it's going to it's going to skyrocket. 0:42:17.690 --> 0:42:23.089 Yeah, I'm really excited for it. Well, um, this has 0:42:23.090 --> 0:42:26.540 been fantastic. So tell us, um, in addition to your 0:42:26.540 --> 0:42:31.279 return on security, which is the newsletter and the overall brand, uh, 0:42:31.280 --> 0:42:33.259 what are what other stuff do you have brewing? Do 0:42:33.260 --> 0:42:37.009 you have like, uh, reports coming out, new research, like products, 0:42:37.010 --> 0:42:39.049 anything we should be aware of? 0:42:39.590 --> 0:42:40.250 Yeah, I've been. 0:42:40.250 --> 0:42:42.350 Trying to, uh, write a bit more on the blog 0:42:42.350 --> 0:42:43.880 this year. Try to just try to get more of 0:42:43.880 --> 0:42:46.700 my thoughts out there as possible. Um, I've kind of 0:42:46.700 --> 0:42:48.980 gone back to doing kind of monthly, and I'm going 0:42:48.980 --> 0:42:51.260 to be releasing a quarterly roundups as well, because I 0:42:51.260 --> 0:42:54.200 want to be able to track these things over time. Um, like, 0:42:54.200 --> 0:42:57.140 I'm big on kind of showing like look backs and saying, like, 0:42:57.140 --> 0:42:58.640 this is where we were last year this time, or 0:42:58.640 --> 0:43:00.920 this is where we were last month this time to 0:43:00.920 --> 0:43:04.190 kind of help understand trends. Um, so I'm just kind 0:43:04.190 --> 0:43:06.890 of really increasing the output of because I want to 0:43:06.890 --> 0:43:09.650 be a true resource. I want to be like the 538, uh, 0:43:09.650 --> 0:43:11.089 I want people to be able to use it and 0:43:11.090 --> 0:43:16.280 actually consume it. And, um, I also plan to do 0:43:16.280 --> 0:43:18.739 I'd like to open source the data and have somebody, 0:43:18.739 --> 0:43:20.930 some data science people say, like, all right, tell me 0:43:20.930 --> 0:43:22.430 what you can come up with. Like what? What is 0:43:22.430 --> 0:43:25.700 interesting like either, you know, run some models on it, 0:43:25.700 --> 0:43:28.580 run something and let's let's see what, what ideas you 0:43:28.580 --> 0:43:32.270 come up with, um, and what kind of trends you 0:43:32.270 --> 0:43:34.940 can uncover from a data science perspective. Um, so I 0:43:34.940 --> 0:43:36.320 think that's pretty fun. 0:43:36.320 --> 0:43:36.710 Yeah. 0:43:36.710 --> 0:43:39.620 Um, yeah. Yeah, I want to mention that that that's 0:43:39.620 --> 0:43:43.250 something that's super awesome about what you do. You have 0:43:43.250 --> 0:43:46.190 the data. You've been collecting this data in a very 0:43:46.190 --> 0:43:53.580 sort of manicured and, um. Detailed way in rigorous way 0:43:53.580 --> 0:43:55.500 this whole time. So you have this data set and 0:43:55.500 --> 0:43:59.250 you put out different products around it. Um, I love 0:43:59.250 --> 0:44:03.050 the idea of when I go to the site. Um, 0:44:03.050 --> 0:44:05.810 I would love like a dashboard that is very much 0:44:05.810 --> 0:44:10.010 like FiveThirtyEight, which FiveThirtyEight really only spins up for, like 0:44:10.010 --> 0:44:13.160 big elections. I'm sure it's getting big right now, but like, 0:44:13.160 --> 0:44:18.070 I would just like to. Get quick narratives like, remember 0:44:18.070 --> 0:44:21.100 the thing we talked about, okay, what was 2021? What 0:44:21.100 --> 0:44:25.600 was 2022? So what I would really love is like, 0:44:25.690 --> 0:44:29.890 just explain the whole thing. Starting five years before Covid, 0:44:29.890 --> 0:44:32.710 the thing that you did, you said, look, we have 0:44:32.710 --> 0:44:35.890 this build up of general economics. Then we had Covid 0:44:35.890 --> 0:44:38.080 happen or whatever. If I could get that in a 0:44:38.080 --> 0:44:42.910 paragraph or one sentence or a full page, imagine dragging 0:44:42.910 --> 0:44:44.860 this kind of related to the thing I'm getting ready 0:44:44.860 --> 0:44:50.080 to release. Imagine dragging your explanation from one sentence of 0:44:50.080 --> 0:44:53.560 15 words all the way to a full page of analysis, 0:44:53.560 --> 0:44:57.219 but it's exactly the same thing that you described, right? 0:44:57.219 --> 0:45:02.770 So it's like, give me the the TLDR of the 0:45:02.770 --> 0:45:07.000 entire security space. Yeah. Um, and then you have different 0:45:07.000 --> 0:45:09.879 graphs there, right? It's like, show me VC performance. Show 0:45:09.880 --> 0:45:12.520 me the number of startups, show me like whatever. And 0:45:12.520 --> 0:45:14.470 I feel like that's what I'm getting from you in 0:45:14.469 --> 0:45:17.710 your newsletters. So I feel like that could be like. 0:45:18.430 --> 0:45:21.670 Definitely like a future direction for you with like, this 0:45:21.670 --> 0:45:25.570 really rich front page dashboard. And it's just like it's 0:45:25.570 --> 0:45:29.930 like a. It's like a universal dashboard to just state 0:45:29.930 --> 0:45:30.830 of the Union. 0:45:31.250 --> 0:45:32.600 Yeah, I love that idea. 0:45:34.110 --> 0:45:36.149 I love that idea too, because I kind of like, 0:45:36.150 --> 0:45:37.530 I think about what I do is like, almost like 0:45:37.530 --> 0:45:40.110 a public utility. I'm like, I want you to use it. 0:45:40.110 --> 0:45:41.669 Like it is. It absolutely is. 0:45:41.670 --> 0:45:43.680 Already I love it, I love it. 0:45:43.680 --> 0:45:45.960 That's it's I love that idea too. I'm definitely gonna 0:45:45.960 --> 0:45:47.520 noodle on that one and try to figure out how 0:45:47.520 --> 0:45:51.540 to how to do it. Because I want to share more. I, 0:45:51.540 --> 0:45:54.120 I've made so many like great like connections, and I've 0:45:54.120 --> 0:45:55.710 learned so much just by doing this. And so I 0:45:55.710 --> 0:45:56.759 want to be able to like kind of share it 0:45:56.760 --> 0:45:58.049 back as much as possible. 0:45:58.380 --> 0:46:01.590 Yeah. Well, I encourage everyone to go check you out. 0:46:01.590 --> 0:46:06.270 It's our return on security. Absolutely. The the best and 0:46:06.270 --> 0:46:09.870 really the only newsletter in that space. Just absolutely crushing it. 0:46:09.870 --> 0:46:12.390 Can't wait to see what you do next. And thanks 0:46:12.390 --> 0:46:13.140 for coming on. 0:46:13.530 --> 0:46:15.390 Awesome. Thanks so much for having me. It was great talk. 0:46:15.390 --> 0:46:15.870 All right. 0:46:15.870 --> 0:46:17.310 Take care. Thanks.