Defending Truth in the Digital Age: Nick Loui and the Fight Against Misinformation — Founders In LA

Hey everybody, thank you for joining us and welcome to the Founders in LA podcast. I'm your host Ethan Cole, and this is an opportunity to shine the spotlight through a product lens on some of the exceptional startup founders we have as part of the LA community in an unedited, one-take, organic conversation. With us today is Nick Louie. Really excited to chat with you today, Nick.

Hey Ethan, thanks for having me. Super excited to be here.

Oh, this is, this is gonna be a great one. I love Peak Metrics. Um, but first, a word from our sponsors. Founders in LA is brought to you by Unita Club. Unita Club is a coworking space that sets itself apart with locations in El Segundo, Manhattan Beach, and Hermosa. Unita is where creativity flourishes. Unlike traditional offices, they provide an inspiring environment where ideas can thrive and businesses can grow. With UNITA, there's no hidden fees, flexible terms, options for dedicated offices, and unlimited access to conference rooms, a photo studio, a podcast studio, plus standard and 3D printers and 24/7 access at any of the locations. They're local champions who support neighborhood businesses, open their event spaces to nonprofits, and celebrate art, music, and culture. Join UNITA Club and experience coworking like never before. Learn more at www.unita.club. That's U-N-I-T-A dot C-L-U-B. We're also brought to you by the Product Manager Association Los Angeles, available at pma.la. They're the largest professional organization for product designers in Los Angeles with more than 4,000 members from over 500 companies. They host monthly meetups, organize the Product Leader Council, and have a mentorship program where they connect working product managers with students from underrepresented groups to build a better and more diverse next gen. They're also hosting the LA Tech Connect conference where practicality meets AI, dedicated to— with workshops and panels on how AI is being integrated into the workflows of today's software developers. So developers, engineers, designers, and product people, learn more at pma.la and latechconnect.com. Our guest today is Nick Lui. He's the co-founder and CEO of PeakMetrics. Nick, can you just give us a 30-second description of PeakMetrics?

Yeah, absolutely. So PeakMetrics uses machine learning to measure and predict how threatening online narratives spread online. So as part of that work, we look at different narratives that are emerging. We flag for risky patterns, such as if there's coordinated activity or toxic content or the use of deepfakes. So that our clients can get ahead of some of those issues. So some of the work that we've, that we've participated in includes helping folks like the Department of Defense track state-sponsored propaganda or election security issues, helping large financial healthcare institutions understand how different online narratives could be affecting their reputation or industry. So that's a little bit about us and what we do.

Wow. So in this world where it feels like the ability to create deepfakes has just exploded in the last year, I feel like there's a greater importance for folks like you to be able to identify what those are, who they're coming from, and how they're being spread. Are you seeing an uptick in that yet? Is it coming soon? Or is it, or the kind of business as usual, you've been tracking this forever and like, hey, it's been the same. Everyone else is now just becoming more aware of it.

It's definitely a particular moment in time. You know, I think as a society, we've known how, uh, the information environment and media, you know, affects a lot of our decisions and a lot of the way that the world evolves, uh, you know, particularly sort of post-2016 when we realized its impact on, on sort of political outcomes. What we've seen in the last 2 years is that it's the type of issue that not only affects, uh, public sector security, but it's affecting private sector as well. And generative AI is part of that, is really just this moment now where the ability for an adversary to create this content at scale is just, uh, much simpler in language that is much more, uh, English native, for example, uh, because of these large language models. And then of course to incorporate hyper-realistic image and video now, it is definitely a moment where technology is now rapidly accelerating the speed at which these types of narratives can spread.

Are you seeing with this ability for generative AI to be multi-language, is, has it become, is there more of a threat for it to be more of a global issue than, you know, just someone spreading something in English or people just attacking, you know, English accounts? Now that generative AI provides the ability to spit out all this information in whatever language you want within seconds, is, is that proliferating or is that not yet a thing?

It's definitely, it's definitely an issue we see. You know, if you look at, um, if you look at the world of misinformation, for example, which is part of the world that we, uh, analyze and, and try to understand as well, English content is, is of course the most well-researched, but what we are seeing is that, you know, particular minority communities are now exposed to this type of information as well. And I think to your point, with these generative models, uh, the sophistication is, is definitely, has definitely increased. There's another stat that came out a couple years ago now, but between, I think, 2017 and 2018, the number of disinformation for hire services, the number of countries that those services existed in, or basically like dark PR companies, it doubled. So that's a little scary. And that was, you know, that's data that's already a few years old. So it's, it's the type of thing that, you know, is only increasing in concern.

Okay, so I got to know what this is. I've never heard the term dark PR. What is it? Dark PR firm or company or agency?

So normally when you hire a PR firm, right, you're hiring to get exposure, uh, generally around, you know, positive messages, uh, for your, for your organization or your brand. Uh, there's an entire subculture of, um, quote, dark PR firms that you can, that companies can hire to do things like opposition research or, uh, creating narratives that are perhaps adversarial towards a competitor, uh, and they are not the most ethical, you know, type of organization, uh, but they, they certainly exist. And you can imagine how with these types of newer technologies, their ability to create real damage has just increased.

And is this typical to, you know, the United States, or these dark PR agencies? And I guess I think of, when I think of this Uh, it's more like the ability of foreign actors to, to really amplify their propaganda, right? So that's— I imagine that's been somewhat of a blocker in the past. Their resources are limited. They may not have people who speak multiple languages. Maybe they have someone who speaks English, maybe. Uh, but now it's like, hey, we can spread, you know, these you know, these dissenting opinions and maybe, you know, malicious opinions very easily across the world with at the speed of light, right?

Yeah, I think it's both scale. So, you know, the cost to create a few messages versus 1,000 messages, now that, that cost gap has been significantly decreased because of technologies that exist now. And then, uh, in terms of being able to create both convincing content as well as convincing accounts, right? So what I mean by that is not only are we seeing that you can create convincing, you know, say fake reviews or messages online that are native to the language that they're spreading in, but you can also create, of course, convincing imagery. And then on top of that, you can create convincing accounts, right? So I can create a convincing deepfake, you know, headshot. So the ability for us to actually detect, you know, when there's manipulation or coordinated activity is much more difficult than it was even just a few years ago.

Oh my gosh. What I'm thinking of now, what you got me going down is like using a Midjourney or a DALL·E to just create, you know, 500 images of my avatar in different scenarios, you know. And so like, You could create this very realistic profile. So no longer is it like some headshot from— they just grabbed off the web from some random person who you could really probably reverse image find. Instead, they could create a brand new person, a brand new image, and have that person within seconds be presented in a number of situations that you would make— would make you believe that that's a real person in a real scenario.

Yeah, that's exactly right. And, and, you know, on top of that, while there are guardrails that companies like OpenAI are putting into place to try and reduce, uh, reduce these activities from happening, what we're also finding are alternative models, almost like evil twin versions of ChatGPT, for example, that are pretty readily available on the dark web. You can purchase it for, what, $10 a month? $20 flat purchase, and all of those guardrails are removed. So you can actually use it to do everything from create a disinformation campaign to create a large-scale phishing campaign, for example, that is, you know, extremely targeted and very realistic. And, uh, so yeah, the, the need for sort of protecting the information environment and almost these human security or cognitive security threats, uh, has definitely increased.

Well, thankfully we got some people, the good guys, here on our side. So, uh, how did you come to create Peak Metrics? How did you find yourself in this world and trying to make a difference?

Yeah, definitely. So my background's a bit unique, or I suppose a bit weird, like many entrepreneurs, not necessarily a straight line. I've been an engineer basically all my life. I started coding when I was 11, 12, and Perl, OC, C++, all that type of thing. And, you know, in high school was heavily involved with computer science, got into trouble for hacking the school network a couple of times, sort of that person. And then wanted something different, wound up studying communications. My senior thesis in college was actually building a crisis communications tool on top of Twitter data. So this was in like 2010 or so, really early, trying to understand how to combine these sort of interests in, in the media environment and, and data together. And after that, spent most of my career working in a marketing function in, at technology companies. So started my career actually back in LA. Working weirdly enough for Ashton Kutcher at a company he created called Catalyst, where I was part of the team that got him to become the first person with a million followers on Twitter. We did a lot of really early work with partnering with large brands or NGOs to create what would eventually become influencer marketing campaigns. And after that, spent about 6 years at a tech startup, helping to grow it from the first hire outside the co-founders to around 80 employees. Eventually serving as its chief marketing officer. And consistently through all of those experiences, I was trying to figure out, okay, we are constantly as sort of marketers and communications professionals working to understand how this new sort of social or news ecosystem is developing. But there was still a lack of data to understand how you can plant a message in one place, like a subreddit, for example, and have that grow and grow and grow. And so I was really interested in this concept of how to actually put better numbers behind, uh, behind the information flowing online. And, and so that's a lot of our roots at PeakMetrics, is really trying to, one, put numbers behind how this information is flowing and spreading online. Uh, and then two, it was basically this Black Mirror image of everything I spent my entire career doing, which was building these sort of campaigns from a marketing perspective. Now it's trying to actually put a mirror to that and figure out how to trace them and understand why these campaigns are actually spreading. Because at the end of the day, the adversaries that we're looking at are using the exact same tactics that I basically spent, you know, my entire career living.

That's amazing. So it's like, and I think you mentioned there, you were part of the team that brought the first person to get a million view, a million followers on Twitter.

Yeah, it was my first, well, not my first, it was like, It was initially an internship, actually. I got it by, I think, tweeting at Ashton Kutcher, 'cause he was early on on Twitter. And yeah, we, at that point, Twitter was a really, really early platform. He was running a campaign against CNN to become the first person with a million followers on Twitter. So I was, you know, I got to be in the room as we were sort of executing some of those campaigns to race him to become the first person with a million followers on Twitter, which was a lot of fun.

Wow. So then you took these skills that you knew from building followers to then tracing how other people were trying to do it in, in, uh, kind of a negative light.

Yeah. I mean, I think if you, if you look at how these sort of narrative threats develop, they are using a marketer's playbook in a lot of, in a lot of ways. And, and. The more that we can put numbers behind, you know, why that information is actually spreading, try and develop sort of patterns for certain playbooks on how certain adversaries, you know, create these campaigns, the better. So yeah, it's sort of this, it's just sort of trying to trace and measure these types of campaigns.

So really curious to learn, how did you get your first customers? You know, this isn't a typical business where you kind of knock on someone's door, hey, do you need someone to help track, you know, dark PR firms for you? Like, how do you— how did you find those first couple customers, the first person who would say yes to you?

Yeah, so we started our work supporting different types of advocacy organizations, sort of, uh, the political sector. But we really escalated our work in 2020. Uh, we went through the Techstars program out of Boston, and we started to do some work with the Department of Defense. And the timing worked out really well, I suppose, because that year we of course had COVID emerging. And so a lot of our early work was actually trying to understand different types of narrative around COVID-19. So everything from public perceptions around COVID-19 through to helping vaccine manufacturers understand their audiences, and then even DOD work trying to understand how foreign adversaries were positioning like their vaccines versus say Johnson Johnson and Pfizer. So that was a lot of our early work. We actually got some awards in that space around some of our COVID-19 work. And then that funneled into the 2020 election and of course January 6th. And so all through the 2020 election and January 6th, we were applying our methodology towards understanding different types of election security issues. And of course, January 6th, understanding similarly how was the information environment lighting up, you know, as were unfolding. And, uh, that was, that was a lot of the basis. And then that then led to a lot of inbound from corporations trying to figure out what it meant for their strategy and, and continued work to support the public sector.

Wow. So was it through Techstars where you got the connection to DOD? Cause it sounded like that was really the, to use your words, the catalyst for it all?

That was definitely the catalyst. I mean, I think the playbook though was much the same, which, you know, was how many people can you talk to as quickly as possible to understand the different needs that different folks have? I think when we went through that program, we had to talk to like 60 people for just customer discovery conversations in the first few weeks. And it was a diversity of personas and, And that really at least helped us understand where we should be focusing. Um, and, and so as simple as it sounds, I think a lot of it was of course leveraging the resources of a startup accelerator to help us force that customer discovery process. But the actual process was the same that you'd run in, in any sort of early stage startup of just talking to as many customers as possible and, and listening.

That's fantastic. And, and that, that DoD connection, was that through Techstars connection or was that just through the discovery? Just curious to, to how, how you broke in through that, how you broke into that section.

Uh, so it was, it was through Techstars. The program that we went through was called Techstars Air Force. So as part of the application, we, uh, we, the hypothesis that this was a good market for us, and then, you know, it was validated through the application process where we were, you know, down-selected. And then as part of that, down-selection process, they had different public sector government stakeholders in the evaluation process to sort of say, okay, there's probably something here. And, and so by the time that we were in the program, there was at least some understanding that there was interest, and then it was a matter of us figuring out more specifically what that would be.

That's great. So it sounds like for folks who are like early stage or considering to create a company, you know, picking the right accelerator is important, uh, and can really lead you to that first, that first person who will give you a chance.

Yeah, I mean, we had some traction before that, right? So we had some revenue, we had some pilots and customers to at least prove that there was something there. But for us at least, there was definitely value in having an accelerator really, well, do exactly what it's supposed to do, right? Accelerate the business and speed up sort of that that, that process. And I think we were lucky in finding the right one because I think you're right, particularly with like Techstars where there's so many different programs that you can apply to, picking the right one with the right managing director does seem to make a big difference.

That's great. Well, thank you so much, Nick, and kind of grounding us back in Los Angeles and the LA community. We love to ask on this podcast, you know, what's your most LA thing that's happened to you? This can be tech-related or it doesn't have to be.

Such a good question. So I'm a born and raised native Angeleno. So I feel like maybe my perspective is a little different here, but I do distinctly remember in college, I was back from summer break. One of my friends at that point was dating an anonymous Disney star, and we were just hanging out in my parents' house in the backyard. And said anonymous Disney star came over to, you know, hang out with all of us. And I was telling my sister, who's 8 years younger than me, the next day. And she was extremely mad that this person was, you know, hanging out in my parents' yard and we didn't wake her up from, you know, her sleep. So that's probably an only in LA type of thing to happen.

Oh, absolutely. So you come back from college break and your friend is dating a Disney movie star. A Disney star, Disney TV star, and your sister's mad at you for not waking them up. That's fantastic.

Very LA.

Oh, that's super LA. I love it, man. That's great. Uh, well, thank you so much for, for stopping by and chatting with us. Uh, this is Nick Louie, the co-founder and CEO of Peak Metrics. Uh, well, thanks so much, Nick.

Yeah, thanks for having me.

I'd love to thank, uh, our sponsors again, PMALA and Unida. Like to thank you all for listening to us, and if you, if enjoying what you're hearing, please smash that subscribe button. Thanks again, thanks again for joining us, and we'll catch you next time on Founders in LA.