How is AI steering the future of marketing strategy? With the convergence of AI and marketing tactics, Rex Briggs paints a compelling picture of what's possible: AI agents that revolutionize user interactions, and generative techniques that craft persuasive content. Drawing from his deep expertise in marketing measurement, Rex joins Tom Chavez and Vivek Vaidya to explore the cutting-edge of AI-driven marketing strategies. Listen for insights on harnessing AI's potential in modern marketing.
Marketing has always been about connection, but what happens when AI becomes the mediator? With vast strides in AI-driven campaigns, how do marketers ensure they remain true to their brand and their audience? Who ensures that campaigns are not just data-driven but also ethical and impactful? As we stand on the precipice of an AI revolution in advertising, how does one navigate the intricate balance between personalization and consumer control over their data?
Rex Briggs, a luminary in marketing measurement, envisions a future where AI-driven campaigns resonate deeper and more personally with audiences. With vast industry experience, Rex sheds light on emerging trends and their implications. Joining Tom Chavez and Vivek Vaidya, the trio explores the frontiers of AI in marketing, from its transformative capabilities in content creation to the nuanced ethical challenges it presents. They uncover strategies and insights essential for marketers in this AI-dominated age, emphasizing the synergy of technology and human intuition.
Find Rex Briggs on his LinkedIn (https://www.linkedin.com/in/rex-briggs-2811b3/) and X (https://twitter.com/rexbriggs).
PLUS bonus content: super{set} Spotlight on Headlamp Health co-founder Andrew Marshak and his experience so far working alongside Tom, Vivek, and the super{set} team at Headlamp Health. Hear about Andrew's maniacal commitment and find out how Tom Chavez and Vivek Vaidya show up as "true co-founders FOR REAL."
Learn more about Headlamp Health at www.headlamp.com
Find Andrew Marshak on his LinkedIn (https://www.linkedin.com/in/andrew-marshak/)
Read Andrew's latest blogpost on superset.com - "Why Headlamp Health is Bringing Precision to Mental Health"
Learn more about super{set} at superset.com
Find more episodes at www.theclosedsession.com
Welcome to The Closed Session, how to get paid in Silicon Valley with your host, Tom Chavez and Vivek Vaidya.
Tom Chavez: Welcome back to another episode of season four of The Closed Session podcast. I'm Tom Chavez.
Vivek Vaidya: And I'm Vivek Vaidya.
Tom Chavez: Today, I'm excited to have Rex Briggs with us. We've had a chance to work with Briggs... with Rex along the way. Rex is an author, public speaker, consultant, and an absolute expert in measuring marketing ROI, leveraging advancements in tech to reshape the industry, a place where you and I have dwelled very meaningfully over the last couple of decades. So you can run, but you can't hide from Rex Briggs. He's everywhere in AdTech, MarTech and modern developments in how AI has the opportunity to reshape legacy marketing. We're going to get into some of that here as we go. So Rex, so great to have you with us. Welcome!
Rex Briggs: Ah! It's good to be with you guys again.
Vivek Vaidya: Thrilled to have you with us here today, Rex. Let's dive right in.
Tom Chavez: Let's do it. So, Rex would love to have you just back it up for us a little bit here for our listeners. You've been at this a while and so, you know, how did you get into marketing ROI? Because, you know, you are kind of, I don't want to sound too, uh, breathless about it, but you're kind of the OG of marketing ROI. So back it up and tell us how you got into this in the first place. How did this interest take root?
Rex Briggs: You know, I started at Yankelovic partners before the internet and so we were doing segmentation and the goal was to create segments of one, which is pretty pure... purely theoretical at that point.
Tom Chavez: Right.
Rex Briggs: But the idea was, man, if you could talk to an individual and understand their hopes and their dreams and their fears and their aspirations, man, you could connect with them and you could serve what they really wanted from a business. And so...
Tom Chavez: One-to-one marketing, right? Way back in the nineties. Who was...
Rex Briggs: Yes, before Don Peppers and Martha Rogers, 1 to 1 Future.
Tom Chavez: Peppers and Rogers. That's right.
Rex Briggs: That's right. And so, uh, what was, what happened was I happened to be doing a study for Intel, uh, actually it was for IBM. It was about the Intel Pentium chip, which couldn't do math. You guys might remember that vaguely.
Tom Chavez: That's right. Arithmetic at problems in the chipset.
Rex Briggs: Yeah, and you know, Intel was saying, look, this is not a problem except for this tiny, tiny, tiny percentage of people that need that level of precision in our floating point math.
Tom Chavez: Right.
Rex Briggs: But it was still hurting sales for their chip. And so IBM was doing research trying to figure out what implications would it have. I was doing the research on their behalf. And midway through that point, Intel decided that they were going to reverse directions and they were going to replace that chip. And so in the research and the data, there wasn't a World Wide Web yet, but there was CompuServe, and that's where the IT tech people would go to get information on this Listserv. And we could see people who were part of that Listserv change their opinions within 48 hours of the announcement. We've never seen anything like that before. It used to take weeks and weeks. In fact, they were diffusion models would tell you how communication would filter through the population. And so when I saw that, I'm like, wow, I've seen prodigy before. My dad was a computer, you know, database guy and used it and so forth. But it made me realize it was going to change marketing. And so, you know, fast forward a little bit later, you know, yeah, I created the first study to... when the web was launched to figure out who these people were that were adopting this tool. Louis Rossetto, the founder of Wired, recruited me to be the first director of research. And all of a sudden I'm sitting on the other side, which is people are asking me, what is this ad worth? And, uh, is there anything besides a click-through that we get from this? And how do we know what the value was? And that's when I began to create the first lift study to create a control and expose group and figure out what impact that had. And then it was about, well, okay, after we did that study and prove that online advertising worked 1996 and 1997 with the IB study, the next logical question was, but should I reallocate money from television to digital from magazines to digital, which one gives me the better ROI? And answer that question I had to invent multi-touch attribution to try to connect to that information. So, I mean, that, that, that was really the process is you get confronted with a challenge to answer a question for business, to make things logically line up, to make a good decision. And you have to invent new tools and new capabilities to do that sometimes.
Vivek Vaidya: Wow! yeah. So you've obviously went go way back in terms of leveraging data and in advertising and marketing. So can you share a little bit about your perspective on the connection between marketing tech and data. What kinds of data were you using? How did you collect it? What did that process look like?
Rex Briggs: Well, some of it's quite shocking how long it takes for an industry to adopt a good idea. I mean, if you think about the... think about the internet, you're thinking, we actually knew that the advertising on the internet had value in 1996. And by 1997, we had expanded to, to measure, you know, 13 different industries. But it took another eight years before you really began to see the curves and the cycles taking off. And when we had behavioral targeting by 1996, in fact, we, I, the first study where I use neural networks to try to, to change the content that was being served on the Wired site was in 1996 as well. So you look at how long does it take them to have that adoption and oftentimes it's not a MarTech you know, challenge gap, it's an education gap and a business decision making gap. So, you know, it's remarkable how much data and information we had decades ago. And yet we're really only starting to grapple with how to effectively manage it today. It does remind me a little bit of the Industrial Revolution story of the first factories that were built with the steam engines were still organized the same way that you'd organize a factory along a river if you had a water wheel.
Vivek Vaidya: Oh, yeah.
Rex Briggs: But you weren't constrained to a water wheel anymore. So why were you still designing it that way? And we really have had that for the last decade to two decades in marketing, which is why are we still designing marketing as if we don't have the internet and one-to-one connections and communications. So I feel like we're finally turning the corner on that part. Uh, it's not that the data has changed, it's that some of the, though it has gotten better, it's some of the mentality has changed and the generation of people running marketing grew up with this data and this technology.
Tom Chavez: So I'll, I want to pick that up a little bit, Rex, because you have been here from the get. It's not, you know, you are the inventor of media mix modeling, multi-touch attribution, measurement of ad efficacy and, and its value, it's open market value. I mean, this is... and, and so you're pointing out that it takes a long time, it seems for organizations to sort of metabolize the possibility right at hand. Why, you know, on the one hand, yeah, it's encouraging that it happens. On the other hand, for people like us, you know, who've been excited about what we know the technology can enable today, that latency is damn frustrating, right? The time it takes for organizations to get off their duffs and actually do it. What's your theory? What? Why? Why so slow?
Rex Briggs: First, a quick correction. I didn't invent marketing mixed modeling that was Professor Little in 1972. And what's even more depressing is it took till the mid eighties until that was adopted. So it's not like this is the first time that only a decade was, I mean, I was the first person and I did have a patent for how do you put digital advertising into a marketing mix model. And, but again, even that took, that was in 2002 for P&G and, and J&J and a couple of other brands we worked with, and it's taken, it took another 10 years before that was broadly adopted. So you have these, this long lag. And so the question I've asked again and again is how do you shorten, shorten that cycle? And I do think this is where ROI becomes really important because if you can show the financial impact much faster then people adopt much more quickly. If they have to wait months to understand the return on investment or it isn't tangible, it isn't direct, then there's a lot of places for people to hide who don't want to change. But if you have black and white, here is how many more sales you will get here is how many more customers you will attract. It's a lot harder to hide from that and it gives the ammunition to the people who are the change agents that wanna move faster. So, you know, I, I, I think we've made a lot of progress. The cycles are getting shorter. I mean, I, I also did some of the original research on social media when we actually, when it was Myspace before Facebook was even open to the public. That model was repeated by Facebook and that helped them accelerate how much dollars moved in that space. We did the work with Greg Stuart at MMA global on mobile and that... shorten that cycle. And I think this cycle with AI working with arts AI. And, you know, I introduced and connected them with Clara tossed and here's a news flash that you'll read about tomorrow morning. They merged. So they will be bringing that technology to their customers. And I think that that if you have the enablement and you have the speed of this information with black and white ROI. I mean, they, they can now do like a money back guarantee, which is like, if you don't see this much lift or payback, you will get your, your budget back. So I mean, I think that that makes it adoption happen a lot faster.
Tom Chavez: So it's speeding up is what I'm hearing you say.
Rex Briggs: It should, but I think we should still ask the question, how do we go faster? And I don't fully know that answer. I know what we're trying to do. What I've seen cycle on cycle is that if you can get a consortium of marketers to do it together and publicly share that information. I mean, this is really what Greg Stuart's model is at MMA Global is bringing these consortiums together, getting them to share it, getting them to stand on stage and be public about it. Then it puts, it gives, it puts pressure on others to move faster. And it also gives, um, safety because you can look and say, Hey, it's not just our results that were good, but look at, you know, Kroger or look at ADT or look at monday.com. And the, they also had strong results and therefore we have more confidence to move faster. So I think that that's a big part of the formula. It's more human psychology than frankly, it is data and, uh, and reporting.
Vivek Vaidya: That makes sense. But. Let's come back to the data bit for just a minute, Rex, and go back to your cross media research, which was the first of its kind back in the day. And that's when you said you invented multi-touch attribution, which kind of combined different, uh, studied the impact of different forms of channels of advertising. What role did data play? What kinds of data did you collect? And what was the process that you followed to collect the data that went into this study? Uh, if you can recall.
Rex Briggs: The key, the key parts for the data that, that were really, you know, in my view, there was three parts that we needed to get together. We needed to understand the full funnel. So we needed to know attitudinally, uh, whether there was a shift in, uh, in favorability, brand perceptions, a brand I love, you know, these types of things. And we need to understand how that's correlated with behavior at the bottom of the funnel. Like, which of these things really explained why someone bought or didn't buy that product? Because within that, there's an interesting insight, which is that there are things that people overstate and they believe influence their brand, usually the physical attributes of the product. And then there's these emotional and social cues that they understate. But if you have both of that data, you can correlate how those understated, uh, you know, the brand that's growing more popular. I don't think that's important for how I buy, you know, maybe you buy that way, but I don't buy that way. But it turns out that when you correlate the data and see what you actually buy, you care a lot about more about social proof than you would believe. So I thought. The attitudinal part and then the behavioral part was critical. We had to be able to connect to offline sales.
Vivek Vaidya: Hmm.
Rex Briggs: It wasn't good enough to have purchase intent increase. We need to see actual, you know, we measured for Ford F-150. That was one of the a academic papers that was published in I think 2004. We need to know how many trucks were sold.
Vivek Vaidya: Yeah.
Rex Briggs: We could directly connect back to it online, in order to really know the incrementality, you had to have a control group so that you could see what would have happened if you didn't serve them an ad, but they were still on that same webpage. And so really when we think about the overall data measurement, we think of both the data that you need, the attitudinal and survey data, the behavioral data, the ability to correlate the two because people tend to understate the social impact of something, but they overstate the functional reasons. And so, but if you have both the questions about why did you buy and then, you know, whether or not they bought or not, how, and you ask them, how important are these variables about, you know, is it a brand that's going more popular or a brand your friends would drive or a brand that it gets, you know, good fuel economy or whatever people say, fuel economy is really important. Turns out not as important as you say it is, Mr buying the F-150.
Vivek Vaidya: Exactly.
Rex Briggs: So you can connect those data. And then the other part is profile data because if you see someone who's reached by the ads, and then they visit the web page, or they learn more information, But then they don't buy, why not? What's the difference in the profile between those people? And that's where you unlock the insights about how do you maybe change the, uh, change the message or change the audience. And frankly, that's really where the AI is getting really exciting. And we can maybe talk about that later on is that we can now, with all that data, automate that entire loop.
Tom Chavez: Well, in an earlier podcast, Rex, we were interviewing Seth Stephens-Davidowitz, who's the author of Everyone Lies, and so exactly to your last point about what F-150, Ford F 150 buyers say versus what they actually do, what they think they care about versus what they really, in fact, care about as evidenced by their buying decisions, right? What it reminds me of now is, okay, there are these pesky human beings on the other side of the screen buying and doing things, and as these strategies become ever more sophisticated, it feels like we're bumping up against a line of, wow, you can get very personal. It's one, is it one-to-one? I don't know if it's one to one, but we're getting deep into the nooks and crannies of what people actually think and do, which naturally calls to mind a set of, of privacy concerns, right? So how do you think about how much information is too much information? Do marketers actually care as the whole planet tilts towards broader regulations for privacy. Well, it's certainly in Europe and California where we live, many, many other places. Is it a tempest in a teapot? Should marketers or is it just a momentary little flash in the pan and it'll pass? How should we be thinking about these things? And do marketers actually care?
Rex Briggs: Yeah, I think that there's two layers that I'd like to talk about. One is on the first layer, it's important to recognize that I don't need to have data that's perfect about you specifically to connect with you. I need to signal, uh, and I can still have some noise. So when we were working on Alexis as a brand, the, you know, the data brokers, for example, gave us the target of people who were highly likely to buy a luxury car will turn out that their target audience of who they're giving us was twice the size of all the people who are in market to buy a luxury car. And so, you know, at 1st, I'm like, well, that data can't half of that data has to be wrong because I know how many are sold each year. And it's not that number, that's much, you know, it's a much smaller number. But as I looked deeper into it, what really mattered was as I looked at their data, I'm like, actually, the data is pretty good because this population, this audience has an index of 300%. They're three times more likely to buy a luxury car than if I just grab someone at random. So there was signal in that data, even though it wasn't perfect. So there's quite a lot that we can do with AI and analytics that doesn't require privacy or specific personal data. We can do it with cohort data, we can do it with, with information that, that's aggregated, that, that has signal to it. So that's the first layer I think to understand is that these things can get very, very good even without PII data. The second part is I have thought from the time I was Director of Research at Wired to today that consumers really should, uh, have more control over their data. And I've seen many attempts to try to get zero party data. Maybe this next wave, you know, we'll finally see that happen. I haven't seen the business model pull that off yet, but I am hoping I am cheering for a team, give consumers control of their own data, let them benefit from that data as well. I personally have I never got COVID 19, but I did get 19 pounds during COVID. So I want to lose that and if their marketers knew that and could help position and message to me things that would be healthier for my lifestyle or would help me achieve that goal, I would love that.
Tom Chavez: Yeah.
Rex Briggs: And so there are a lot of things where I think if we're connected with data and information, we can share that in a way that brings us the, you know, the type of outcomes that we want, that could be a really beneficial ecosystem.
Tom Chavez: Well, going back to the mid 1990s and to the point you just raised, Rex, I remember Bill Gates wrote a book called The Road Ahead in 1996-ish. Do you recall that book?
Rex Briggs: I do recall that.
Tom Chavez: Right. He was, he was foretelling the emergence of a data economy where an individual consumers could say, nope, I will, if you want to market to me, I'm going to control my data signature and you're going to pay me for these bits of data. So exactly to the point that you just made, you know, we, a lot of us including Bill Gates and many others have had this vision of data control for the consumer. It remains so elusive, right? I mean.
Rex Briggs: Yeah, it does. And, and, you know, they bought Firefly, which was the technology they used to try to create their passport and they, they can pull off and they had all the resources of Microsoft. It may have been too early. It may have been just the execution wasn't quite right. But I do think, I do think that we're about to enter a whole new phase with a whole new media. And that media, in my opinion, is going to be an agent, an AI agent that knows us and that interacts with us. And really look, you want to know where technology is going with, uh, internet, always look at porn first because that's where these things happen first. And right now there are people who are paying good money to chat with a virtual AI, flirt with them, have sexy talk, whatever you want to call it. And that is giving a tremendous amount of data about who you are and what you want.
Tom Chavez: Right.
Rex Briggs: In a very specific narrow sense. But if you imagine that there might also be one who's your buddy, who's watching the sports game and giving you tips for your fantasy football league, about which players you should go for or whatever, we already have agents that are helping us book travel and other things. And those agents will learn enough about us to where they can also have host read-ins pretty much advertising to us, right? So why not figure out in this next wave, a mechanism as we build these agents to where you have control over that data? I'm working with my son, Jared, who's in college, he's a junior. And this is an entrepreneurial idea that he's working with, which is, can I create agents? And could you have control over that data or have a different kind of model when you create that. I'm sure lots of other people will have something similar and we'll see, you know, which of those ideas when, but that's one way in which you might end up in a world where consumers do have control over their data because people build in the business model from day one, rather than what Bill Gates was doing back, you know, in the internet after it started taking off, trying to bolt it on after the fact. So I do think we have to catch it in a generational change and have it built in from the beginning for it to, for that idea of consumers and control their data to take off. And even then it might not work. It might be because it was just don't care. I think they care. I think they should care. I think us close to the data look at and say, you should care. But maybe they don't.
Vivek Vaidya: I think...
Tom Chavez: Yeah, sorry go ahead.
Vivek Vaidya: Sorry. The interesting thing over there is a lot of these ideas you talked about with agents, right. Like some of them are actually useful in that they save you time, like an agent that books travel for you. And some of them are vanity tools like, Oh, I have this agent that is telling me how to change my fantasy football team lineup or whatever in real time if I, if I can, right. So I think the interesting thing over there would be to see which agents... can you combine a set of agents into a business? Because individually each of them could be useful, but then is it worth building a business on? I think that's, that's where the key question is in my mind at least.
Rex Briggs: Absolutely. And if you think about it, what is a TV network? But a collection of different media programs and so forth. So why couldn't you have a business that is a collection of different agents and modalities and relationships? So I think that there's going to be a lot of exciting space. Eventually someone will come by and maybe consolidate the space and create a network of agents and data. And I think if you do it with consumer design in the beginning, where they are going to own their data and they're going to have the right to say who does or doesn't have permission to use this and market it. That could be really interesting. Uh, and you might even have an agent that's smart enough to help you advise on which things you should turn on or not.
Tom Chavez: We're looking at AI avatar projects inside super{set} right now and none of them are porn related, but you know, semi earnestly picking up on a joke that Chris Rock made where he said, listen, I don't want to go on any more, any more first dates, nobody should go on first dates, you should just send a representative. Well, in the realm of data, why not just send my avatar to your point Vivek, get some useful information, see if our avatars like each other. And is this actually worth, you know, $3.50 on a cup of coffee.
Rex Briggs: Yeah, exactly. And there are, there are agents right now that, well, first of all, in the darker side of dating, there are people who are already using AI to help them do the flirting part of it. And then they, they go in for the date later on. And, you know, that was a controversial story that was posted maybe a year or so ago about, uh, about a company doing that. And they weren't disclosing that it was AI behind. So I think the challenge as we enter this new phase is how do we have transparency and still have authenticity? And how does that balance and connection? I mean, Google got their hands slapped with duplex, which people thought was great when it first came out. Oh man, you've got this thing that's calling and making an appointment for my, my hair appointment. And it even pauses and says, um, uh, it sounds very natural, but it's a robot. Uh, so you know, then after all the applause died out and people said, ethically, is that such a good idea? You know, they, they put the brakes on and said, well, we'll figure out how to disclose. So I think we're in that weird space in society in the uncanny valley where we don't really know how to signal?
Tom Chavez: Right.
Rex Briggs: There was years ago, a robotic performer that had a, an actual bladder in that had air coming out and they blew a horn and it sounded incredibly analog and human like because it was designed to be that. And at the end of the performance, which was an amazing performance, people were like, do we clap? I mean, does a robot need applause? And so we're in that space where we don't really know how to interact yet.
Tom Chavez: Yeah, no, I mean, we're all figuring it out. The social norms are very strange. I don't know where we land, but I do look forward to the day, Rex, in about 10 to 15 years where I get to go to a party. And somebody's scolding me for something naughty or saucy that, that I did or said. And I'll say, listen, my avatar did that. I had nothing to do with it. And I, on behalf of my avatar, I want to apologize profusely. We're headed there.
Vivek Vaidya: Yeah, but yes, but take that to just a little, take that a little further. And your avatar does something or says something that results in something terrible happening. Are you liable?
Tom Chavez: I know. I mean, and that's, that's the world we're hurtling into, it's hard. I'm, I'm being a little cheeky and it's perilous. I agree with you.
Rex Briggs: Yeah, well it is. In, uh, in the other project I have with Jared's twin brother, Caleb. Caleb, uh, was the primary author in the book, The AI Conundrum. And, uh, and as we co wrote the very last chapter together, what I couldn't sleep at night because I'm thinking about the implications of anonymous, autonomous AI, these agents. And my fear is that if they're anonymous, uh, we don't have a responsible party attached to them.
Vivek Vaidya: Exactly.
Rex Briggs: Like, I think your point is that we actually do need to have an identity system for AI that is attached to an individual who has authenticated ID, who is accountable and responsible for the actions or a company that's accountable and responsible for the actions. Now, the challenges, if you go there, there and we've just removed a tremendous amount of privacy because if the AI has to have identity and authentic connection that we can trace back to what they do. So to humans, because you have to be able to track everything to be able to have that system work. So I think we're going to get to a place where we have to really have the debate about what is the risk and perils of giving up privacy for identity in this AI world. Or are we okay with things like chaos GPT, which is that, you know, the AI that was programmed to try to manipulate humanity and destroy humanity. And, you know, within, I mean, that, that came out less than a week after the first autonomous agents were being connected to GPT 3.5, I think in that, in that version, and here it was trying to get the nuclear bombs to destroy humanity. And when it couldn't do that, it went through the logic and said, what's the second most destructive thing I can do? It opened a Twitter account.
Vivek Vaidya: But actually, you don't even have to go that far and imagine crazy use cases like that, right? Let's go back to the travel agent, right? And if that travel agent is acting on your behalf, and if you haven't specified the right constraints, it ends up booking a ticket for you that costs $10,000 and suddenly you have to pay because it's, it's acting on your behalf. And now suddenly you're in this fight with the credit card company and the airline that, Hey, I didn't book the ticket, my avatar did, but it did it on your behalf. So who's liable, right?
Rex Briggs: That's right. So ultimately I think we have to have a model where you are responsible for your AI.
Vivek Vaidya: Yeah. And then if you do that, then you're right back to it needs human identity.
Rex Briggs: Yes. Yeah. Yeah. So I think that that is actually going to be the big issue that we have to contend with very relatively quickly. I mean, the other one is authentic content creation and, you know, we will have more and more AI content coming at us and the watermark ideas that the AI companies have said are, you know, help them, cover their ass, but actually doesn't do anything to protect humanity because any organization that wants to do something malicious with AI content simply won't use the watermarked version. So I think what we have to do is we have to flip the model and says that all authentic content has a trust mark on it, and if you don't see that trust mark, then you should question whether or not it's, uh, it's synthetic or created by AI. And if it is created by AI, but a scrupulous AI, it will have a trust mark and you can see that, you know, it's, it's different than the, the non AI generated, so...
Tom Chavez: which works just great
Rex Briggs: ...pretty soon before our election starts rolling.
Tom Chavez: Which will work just great right up until the moment until an unscrupulous AI learns how to fake the watermark.
Rex Briggs: Yeah. Yes. And that's a really interesting part, which is how do you do a zero trust system? And the good news though, is I think because of, uh, I'm not a big crypto currency person, but I do like some of the blockchain capabilities of doing zero trust, uh, authentication and tracing. And so, yeah, but, but you, yeah, you're right. I mean, the model then becomes, okay, if I see something that has been uploaded into Meta, uh, and I see the trust mark and I click on that, then I really have to trust that Meta did its validation of the content before it came up in that chain. That's not actually a hard technology that technology exists with blockchain right now with uh, so I think that that part is solvable. I think the harder part is getting people to adopt fast enough and flip, invert the idea that, that AI gets marked. And it's like, no, no, no authentic content gets marked because if it doesn't have it, then you question its reality.
Tom Chavez: Well, so speaking of AI generated content, Rex, and before we go, I was wondering if you would tell us a little bit about this more recent work you've done in the use of gen AI for the generation of content for advertising, right? And what happened as you saw the early results, I mean, just talk us through end-to-end because it is tectonic. And I hope it's okay. We can talk about it on the podcast. I know it's about to be on court and more, more broadly disseminated, but maybe our listeners can get an early view of, of the breakthrough results that you generated there.
Rex Briggs: Absolutely. I think the, so it started with some of the work around the pandemic and trying to figure out how do you talk to such a diverse country about vaccinations and trying to get authoritative information and facts and a company ArtsAI. I had volunteered to do some work with the ad council with me. And I had a sense that it should work better, but I was really blown away when I saw 43 percent lift. And then we rolled it out to more, more states. And ultimately it's, uh, with some of our work at Brown University, we calculated, we saved 3,500 lives and kept over 20,000 people out of the hospital because of that ability to connect more of the information that, that, that would resonate with someone and get them to want to learn more.
Tom Chavez: Amazing.
Rex Briggs: So we then brought that to other marketers through MMA Global's consortium for AI personalization. And it turns out that the vaccine in states like Mississippi and Missouri is a hard sell. And so that was the 43%. That was a low mark. And actually, when we did it in commercial businesses, the average lift has been 107%, so doubling of conversion rates. So, um, the next piece of that was connecting ArtsAI with Claritas because they have got all this history of this great data and inform information that isn't necessarily PII in some cases you can do it without it, but this cohort data that could make the advertising better and then connecting that with generative AI so that the generative AI could be fed profiles from Claritas and insights about, from ArtsAI about which ads work when and, uh, and with synthetic voices, you can create, uh, and OpenAI's GPT 4, you can create great scripts and synthetic music. All that comes together automatically and the information is feeding back and learning from itself. So we're doing the first couple of studies now. And, and, and as I said, you know, the ArtsAI, Claritas, uh, merger just happened. So, I had a vision of what they could do together and I'm really excited to see them come together. And I'm, I think we've, we haven't announced who the first marketer is just yet, but the market of the CMO has said, Hey, I need this to present in a board meeting in Q1. So let's go. So more will come soon.
Vivek Vaidya: Wow. Good luck with all that, Rex, this sounds fascinating as it kind of all comes together, right?
Rex Briggs: So it'd be fun. And I know you guys are working on some really, really great closed loop systems. And so I think that is the generation that's coming now is this ability to, I mean, it's a little bit frightening because it's a very, very powerful tool. And to your point, it may may make some products and brands that we might have questions about, you know, are they healthy? Are they good for us? You know, may, may help them accelerate. And so, I mean, that's, I think that's always the struggle, which is how do you make sure that we do things and use things in a way that that's responsible. And so, yeah, people have some thoughts about how do we help build it up. I mean, what the key thing that actually someone from Kroger and someone from PNG had told me is that they were showing the story that said, you know, we weren't trying to make a social statement or whatever when he was talking about when he worked at PNG, but when we did the Don ads, we wanted to make sure we weren't just showing women doing the dishes, even though the majority of people who buy Don are women, we wanted to show diverse representations that was important for, uh, so, cause we, maybe we helped create the society where....
Vivek Vaidya: ...exactly...
Rex Briggs: So how do we become positive and show more diverse imagery? And he turned to me and said, how do we do that with AI?
Tom Chavez: Mm-hmm.
Rex Briggs: It's automatically learning from its feedback loop.
Vivek Vaidya: It's a fascinating question.
Rex Briggs: And I do not know the answer to that question. Yeah. I do not know the answer. But if someone, some of your listeners do, I wanna make sure we solve that because otherwise we become regressive with our whole use of AI and I don't want that to happen to our society.
Vivek Vaidya: Yep. And what a great note to end on. Rex, thank you. It's been a fascinating, fascinating conversation. So much, so much interesting stuff. I hope our listeners enjoyed it too. Thank you for joining us today on The Closed session.
Tom Chavez: Really awesome, Rex. Thanks for joining. Thanks to our listeners for tuning in. Don't forget to sign up for the newsletter to stay up to date in our latest episodes and news at superset.com. Thanks all for listening and we'll see everybody next time.