Why Most B2B Data Is Useless and What To Do Instead

00:00:07 Mark Smith
Welcome to AI Unfiltered, the show that cuts through the hype and brings you the authentic side of artificial intelligence. I'm your host, Mark Smith, and in each episode, I sit down one-on-one with AI innovators and industry leaders from around the world. Together, we explore real-world AI applications, share practical insights, and discuss how businesses are implementing responsible, ethical, and trustworthy AI. Let's dive into the conversation and see how AI can transform your business today. Welcome back to the Microsoft Innovation Podcast. Today, we're heading to the greater Phoenix area in the US to meet someone who's quietly reshaping how businesses think about data. Please welcome Brian, Chief Strategy Officer and founder of Five by Five. If you've ever wondered how to turn fragmented, unreliable data into something that actually drives product innovation and sales precision, our guest has built a career doing just that. From democratizing access to high-quality data through member-driven ecosystems to balancing automation with human sincerity in AI-powered sales, this conversation couldn't be more timely. Full links are in the show notes for the episode, as always. Welcome to the show.

00:01:31 Brian Perks
Thanks, Mark.

00:01:32 Mark Smith
Brian, it's good to have you on. I always love to get to know my guests outside of, you know, the tech discussion that we're going to have. So let's start with food, family, and fun. What do they mean to you?

00:01:43 Brian Perks
I love that. I, for me, food is tacos, begins and ends with tacos, carne asada, del pastor, you're talking my language. On the family side, my beautiful wife, Renee, and I have a household of five kids, which is crazy. It's a crazy household. And when it comes to fun, I've been a pilot for about 26 years now. And so we like to travel and that gives us the ability to wheels up and head across the country. We go all over the place, whether we're going to basketball games for the kids, lots of basketball, lots of sports related travel these days, but it is definitely fun being the private pilot and chauffeur for the family.

00:02:28 Mark Smith
I love that. I love that. I helped a mate of mine become a pilot, gave him his first kind of introductory course. He's got on, he now owns his own small aircraft, a twin engine job there that he's just, it's his passion, right? He's just, it was amazing to see how far he's taken it. Do you own your own plane or do you just hire one each time you want to use one?

00:02:51 Brian Perks
I do. Right now I've got a 172 and I have just like with boats, it's that two feet-itis. Bigger, faster. And so I'm in the market. So anybody that has a Cessna Silver Eagle listening, I'm in the market.

00:03:06 Mark Smith
I love it. I love it. Tell me about your company and kind of what are you really focusing on in 2025? What's top of mind for you? know, AI is, if you like, become the center of my universe in my career. What are you seeing? What are you doing? And Yeah. What's your view in mid-2025?

00:03:31 Brian Perks
Man, that is a loaded question. So we at Five by Five are essentially a data utility. So we operate in a space that has been dominated by traditional brokerage models. Aggregators are going to compilers, assembling data sets that they then monetize. The traditional monetization has been some sort of a cost per thousand or a credit or a token, but a unit economic in which the data is being consumed. And we all consume data. We consume data as we're listening to our podcasts, as we're on the internet. And as we all have found, we like to consume data in an unlimited way. We don't like to be paying AOL by the minute. You know, that didn't last very long. Our cell phone minute plans are all gone. I took my kids to Chuck E. Cheese. They're now offering an unlimited pass for Chuck E. Cheese. which reminds me of Blockbuster and the final twilights of Blockbuster and their model as well. Our approach in the space is that we are making the data available to the platform in its entirety. We are aggregating as an original compiler, the transactional interaction data that comes out of an email service provider. They send millions, hundreds of millions of emails on behalf of their constituents. And the outcome of that data is engagement data. It fuels linkages and identity graph. And we're working across fraud and identity. We're working across sales and marketing. We're working across ad tech, publishing, where we're helping to provide identity, measuring traffic, providing attribution, helping publishers monetize. And so for us, at the core of this is coming to the market with a licensing model that is favorable and allows the data to essentially be made available freely to the platform's customers. So whether you're SalesLoft, you're ZoomInfo, everyone has these platforms that are consuming data. And in the consumption of that data, that's where the value was created, right? The data in my database doesn't really do anything except for let me, you know, if I'm a marketer, sleep at night. And so what I see in the change in the market is now that I can have all of the data, it eliminates this kind of scarcity that I have as a consumer of this data and allows me really to focus on who do I really care about? Who's likely to be my customer?

00:06:03 Brian Perks
I don't need nor do I want the liability of housing 5th grade teachers and nuclear scientists in my data set. I only want the sales, the marketers, the people that are going to buy my products, the founders, the product owners, the VPs of engineering that we work with. And I think that has illuminated the realities and the limitations on really what data can I use right now that I can have all the data, now that you can surf the whole internet, how much of it is really relevant to what you're doing? And so we're seeing a big shift in the market as we're making this data available in the same way that once a restaurant or your favorite watering hole has Wi-Fi, why do they do that? Because you stay longer. Companies that are using 5x5 data and making that data available to their customers have customers that essentially stay for forever. They don't have to go out and source the data. The data is endemic to the platform. It simplifies the onboarding process, privacy. You're working against a unified standardized data asset. So those are the big changes that we're seeing. You know, our partnership with ZoomInfo that was announced. has been kind of a big public move, but in the background, we have quietly been providing this data, to in excess of 1000 offerings in the market.

00:07:21 Mark Smith
Wow. So it's interesting because, I'm in a position in one of my companies that's head office out of London, where we're a new startup, we're going after a very, we've created a very specific ICP of our ideal customer. And Traditionally, a bit over a year ago, I was at IBM, and we would use LinkedIn Sales Navigator and stuff to drill down. And what I have noticed is tools like, well, particularly LinkedIn Sales Navigator, they are removing features. They are becoming less and less and less ability to actually reach your potential target audiences with those tools. And so I see gaps in how that offering is benefiting me anymore. The price seems to go up and the feature set seems to be dropping and my ability to use it effectively because they want everything controlled in their walled garden, right? Of, and nothing sits in a silo, right? As in, yes, I have the messenger feature in there, but I'm going to need email. I'm going to need to have multiple touch points Because a client is not, just because I'm wanting to sell today, they might not be ready to buy for another six months, 12 months, whatever. But I don't want to say, hey, because you're not ready today, see you later. So tell me, where does your product finder fit that mix of, and how would I use it per se in that scenario I just gave you?

00:08:56 Brian Perks
Yeah. One of the analogies that we love is that the data that we provide is like when you make paper, there's tiny fibers that are all smashed together, and then we print on top of those fibers. Not every fiber in that piece of paper, in that map, if you will, is relevant to everyone. right? And not necessarily relevant to every journey, every destination. And so what we're doing is we're providing kind of this canvas, this blank slate that covers the entire universe of the data that allows for that ICP to be optimized. Because anything beyond mark your optimal ICP is the deals that you can and should win, plus deals you probably will not win and will cost you money, resources, staffing. And so people say, Brian, you've been in this space for a decade, or 25 years, sorry. I wish it was just a decade, but I'm old with that. You've been in this space for 25 years. What data would you buy? I would always want to buy just the list of people who will become my customer because I don't have to overstaff, there's no waste there. And so with what you guys are doing with that very specific ICP, it is difficult, as I work with founders, I can sell, this product has universal applicability. I mean, how many times have we heard that from our brother and sister founders in the market? And I just, I pound to be specific, stay focused and be disciplined on the deals that you can win easily. Don't go out and show a force, you know, rainmake because you can. Go build that repeatable sales process that scales beyond you as the founder. And that is always the advice that I give in the segment.

00:10:42 Mark Smith
Yeah, so, important. Now, with your 25 years experience in the space, you've now seen since 2022, the generative AI coming to the market and I suppose changing a lot of the way this is done. But I assume that because you've got these solid foundations, you're able to stand on those shoulders and leverage up. So where does AI come into the mix in what you're doing now?

00:11:06 Brian Perks
So there are things that AI cannot generate. It cannot spontaneously generate my email address or your mobile telephone number or your LinkedIn profile. This kind of data is the byproduct of interactions. And we, as you know, what has traditionally been an original compiler of data and on a massive scale, we're just aggregating, you know, all of this content together and making it available because these models can figure out what to say to a director of product, but they aren't going to be able to produce the phone numbers, the activation, the cookies, the maids that are going to go into this. So you're always going to have this activation layer, and that's the layer in which 5x5 sits as a combination of identity verification, this is big in card not present transactions. This is big in, identification, FCRA related validation. Are you, Mark Smith? Am I Brian Perks or am I a Korean, North Korean operative that is now trying to infiltrate, right? We're seeing a lot of this. And so our purpose is we're going to take and enable the activation, the execution layer across sales intelligence, sales activation, telephone outreach, the email component. And then we're also doing some things that are very interesting, specifically in the publishing that are built on top of that audience activation layer. We work with publishers to help them understand who's on their site, then to help them monetize that either first party, selling to their first party audiences, renting lists, lead generation, all those things that publishers used to do before they kind of lost their data strategy, which they're coming back to.

00:12:49 Mark Smith
 Yeah. Now, It's interesting because you've talked about identity here a few times. And then in my mind, my security side ****** up and thinks about fraud, identity theft. I think I heard a stat from the chief security officer at Microsoft to the effect that identity theft has certain frequency. It's now gone up to something like 30 times a second identity theft is happening. that if cybercrime was a nation state, it would be the third largest based on GDP in the world. So America, China, then it would be cybercrime before Germany, which is currently the third unit for GDP. So there's a lot of money in the space of fraud and identity. What are you uncovering or what are you seeing from a security perspective?

00:13:41 Brian Perks
I think that the advantage that Five by Five has in our, essentially cohort of member users of the data, most of these fraud analytics products have been relegated to what we call the data keyhole. They're looking through the keyhole in the door and their view is limited. You don't really know what is in that room, but you have enough of a visible or enough of a line of sight to have some idea of what's going on in there. The advantage that we have, as we're firing our tag billions of times a day, 10s of billions of times a day, we understand the footprint in which we're all using the internet. Has someone compromised your email and is now using what they call account takeover or ATO to now take over and say, I'm going to reset your banking passwords, your brokerage passwords, things of that nature. So we act as the barometer to kind of let you into the room and see what does the digital universe look like? What do the 4.6 billion global radical IVD4 addresses look like? Where are people opening their email? There's 164 million postal addresses in the United States. That's business and consumer, not counting things like suites and floors within that universe. But every router, every IP address is either in one of those buildings or operates virtually as like a VPN or similar. And so when you start to understand the footprint for people, you can now use this data to protect the interest of those data subjects. Someone is in a location mark that you're not normally in on an IP and device that we don't recognize. You probably should hit the two-factor button before you let not Mark reset Mark's banking password.

00:15:25 Mark Smith
Yeah, It's interesting. I wonder how many people see a two-factor authentication notification. And even not doing anything relating to it, hits yes. And let something go ahead, right? I'm sure it happens. I'm sure it happens.

00:15:42 Brian Perks
My mom definitely has done that, Mark.

00:15:45 Mark Smith
Yeah. Tell me about LinkedIn. What's your, particularly, you know, I've done advertising on LinkedIn. I found it very ineffective in getting any cut through. Is fraud happening on LinkedIn when it does come to advertising? What's working, what's not? How are you having the conversations around how LinkedIn plays? Because, you know, I feel that since Elon purchased Twitter now X, that a lot of the business discussions that used to be over there evaporated. And I've had for some years, maybe 13,000 odd followers over on LinkedIn, and it's like a ghost town, right? Like they used to be there, they're not there anymore. I mean, they're still connected, but nobody's home, right? But I saw this massive migration to LinkedIn. LinkedIn seemed to accelerate away for really that business or the B2B particularly discussion. What are you seeing with LinkedIn?

00:16:45 Brian Perks
I think that as a publishing entity and as a publisher, and I like to loop Meta, all of, Facebook, Meta, Twitter, X, all in this kind of universe as being publishers. They're content providers, and it is user-generated content, but that is still publishing. And I actually think the issue is not a LinkedIn problem, but it is a lack of understanding on our side, what is LinkedIn? LinkedIn has been your professional resume. It is not Facebook for business. It is not X. It is not trying to be. And because there are no other business social networks of any real gravity, we have been in this situation where we're expecting things from LinkedIn that LinkedIn never said they were going to give us. On the other side of this, one of the biggest challenges you mentioned, the sales navigator, features, functions. You really have to understand, like any other publisher, what is the audience? So you've got a big job hunting audience that kind of rotates through cyclically as they're looking for a job. You're engineers, you're non-sales, marketing, and HR recruiting people as they are there in transition. The rest of us that want to live there as sales and marketing guys, and you know, I make a post and a billion people see it and I feel good. The reality is that as a publishing platform and as a marketing function, when they purchased Xandr, a lot of that emphasis was in support of what they call audience extension. It's a normal publishing function where unless you are specifically turning audience extension off, your ads are not all running on LinkedIn, right? They are using traditional DSP activation on a trusted network of, you know, quote unquote professional media. But if you don't understand that 100% of your audience without making that change is not all happening on LinkedIn, then you are going to be disappointed, right? Because how much content are we all consuming? How much time are we spending on there? How much reach does the platform has? And then also understanding that in any of these user-based platforms, you have what it's called. called an MOU. And that is a monthly active user. That's not someone who's hitting the website. They are getting an email rollup and digest or notification at least once a month. And so understanding the difference between the daily active users, the MOUs, and this audience extension is really kind of the foundational education function for anyone that's going to do any digital activation, especially in B2B. It's been convenient. It has been effective. But those CPMs, it's very difficult to backend into the CPM, but they're like $65, $70 CPMs on LinkedIn, which is a lot of money for audience extension run of network.

00:19:42 Mark Smith
Yeah, So interesting and just that concept that you had that LinkedIn was never designed to give us what we think we're wanting from it. And I think the other thing that kind of shocked me is that I'm in the tech space in what I do in my career. And there's a bunch of technologists that are in that space, which visit LinkedIn maybe once a day. But if you're not in technology, I find there's, and I did this piece of work a couple of years ago where I was targeting business decision makers in hospitals across the US. And I was like, mate, this will be sweet. LinkedIn here, I can find them all and everything. Crickets. I'd reach out, DM them, using my, nothing. Because, and it wasn't that they were ignoring me. They weren't even getting my, they weren't even, they weren't logged in, right? There's nobody home. There's nobody, nobody there to initiate or correct or have that conversation. Because, you know, sometimes I feel more and more we get into our own echo chambers and feel like everybody's just like me. where you'll find out nobody's like you, as in everybody is elsewhere and they're not where you are, and you think they're following the same patterns you are. Tell me, how do you jump beyond vanity metrics to stuff that moves the dial? And something you said there earlier on was that If you want to engage a list of people that you want to engage with that ultimately will drive a sale, you don't care about all the fluff. You don't want 100,000 leads of which 990,000 are not your target audience because they might sound, wow, look at this massive lead list, but they're not your target market.

00:21:26 Brian Perks
That's correct. I mean, most companies don't run the kind of analytic to really understand their soft costs. as a hard cost. A number of years ago, IBM was very open with it, cost them about $46 to initiate a phone call. And that was the same whether or not the record and the phone number was good or bad. It still costs us software, people, overhead, layers of management, marketing as a $46 execution. And so when you start to think about the cost, it's even if the data is free, You have real costs in your organization in every initiative, even if the initiative is hitting the send button on 50,000 emails. Yeah. It is then the time and the effort and the bad habits that your sales team develops with a bad ICP. The things that I have seen salespeople do over the last 20 years in pursuit of a bad fit prospect is frightening. Yeah. and almost as entertaining as it is frightening. But as a gentleman who has normally been running revenue teams, I don't like to see that because I know how expensive that is. And it is really concerning. And so I really look back at, even how the market has shifted. So, I used to work at Bombora and I was involved in their data acquisition publisher partnerships. And what has happened in that market is No one wants IP-based audience and IP-based intent. IP addresses are places, right? The Empire State Building, all of the tenants in the Empire State Building, they're not paying for a dedicated IP. They are all on the same IP. And so when you start to look at it, you're standing out in front of the building, screaming whatever your product is. You are not getting to the buying group and the decision makers. And so What we're trying to do is help bring in the scope and help salespeople and the platforms to be more concise. Not everybody, the core group that you need within that organization, if you are buying IP-based ABM marketing right now, you are literally any device on that IP, any company in that, you're hitting them all. And you're paying for every one of them.

00:23:52 Mark Smith
Some years ago, working as a partner to Microsoft, I would say, hey, I'd be talking to, let's say, their internal account manager of a particular target accounts that I would be interested in as well. And they had this amazing thing is that they could bring up that company entity and then they would get a feed of, let's call it buying signal, that came in from all their digital estate. So for example, somebody looked at this web page about this product, that somebody did it two days ago. They stayed for X amount of time. And then due to other logged in credentials, browser, right? They obviously, in the case of owning Edge Browser or a Bing account or any other number of accounts, might be Xbox account, etc. there was the ability to identify who that individual was in that organization and light that up. And I've always thought, is that the realm of only the mega companies that can afford that level of infrastructure, or is it available for startups?

00:24:58 Brian Perks
It is. So 5 by 5 powers, individual intent. One of our partner companies, Intensify, won the Forrester Wave with their person level intent built on our infrastructure, which allows for that individual identity-based targeting and attribution. And in the same way that electricity, power, natural gas, all the things that you would think of as utilities as essential to run your business, those are not just Fortune 500 resources. The bodega on the corner, the latte stand, and IT and Chevron are all consuming the same utilities. And it really has leveled the playing field to where these small companies and founders no longer have to be in the data business first to build data-enabled products, which is big. That was not the way it was when I started.

00:25:59 Mark Smith
Yeah, So interesting. I'm definitely going to take a further look. What are your thoughts on AI as a whole? Like what's your position on it? You know, I just had a WhatsApp message go off to me to say today that Microsoft is once again laid off a whole bunch of staff. We're hearing the rhetoric and media more and more that I think 70,000 tech roles have been lost in the US alone this year. Things are changing, and I wouldn't say that it's very easy for us to go, hey, it's AI that's generating that. I don't think that is absolutely the case, but it's a good headline. Where do you think the world's going from a employability of people to the impact that AI is going to have really on every person's career, I suppose, at this point?

00:26:51 Brian Perks
I think the big tenets that I tell, especially kids, Learn it. Learn to use it. Don't have it do your homework, but have it check your math, right? You need to learn how to use this utility because this is not going away. This is here to stay. The fundamentals that I see as issues in the industry, if you and I go prospecting for gold and we find a giant gold nugget, we're rich. However, Stanley shovels and picks that we purchased, Stanley makes no residual from that transaction. And so the problem that AI has now is getting fair representation in the value that is created when AI does its thing. So I think that we are going to see, you know, people say, Hey, Brian, you're a data guy. I'm not really a data guy. I am a license. guy. And the underpinnings of the licensing model for AI is problematic, right? It's not just problematic in the fair use functions for publishers, which I think we'll talk about here in a bit, but it is also problematic because it's not perfect. again, find that we ask of AI, but don't really understand the tokenization process, like how big is that query on the back end? And so a case in point, I was working on a document around privacy and California's CCPA, and the AI made a mistake. And I was not surprised by the mistake it made, right? It is a language model. It is assembling things as it is, you know, believes is the natural order of things. But you have to remember that In any of those interactions, that query is limited. It is guardrailed. It is not covering all of the corpus of information. It is pulling the most relevant information that it believes based on your query. And so it will make mistakes. And so as we're seeing roles and people are leaving, there is still going to be the need for shepherds of this AI, people to tend to this virtual employee flock to make sure that these answers and that the content that is being is created is correct. Because no business wants the liability, especially in customer interactions, contracts, the liability that can be created for an entity when your chatbot goes rogue, on the website and starts giving free subscriptions to all of the visitors, right? These are the things that we hear about. And so I think we're going to see more of almost data and information czars, people that are going to put the rubber stamp, you know, as a pilot, an A&P will do the work, but an aviation inspector stamps it. I think that we're going to see great output. lots of unlimited access to these virtual AMPs, but we're still going to need those AIs to come along and say, you know what? As a human being, I looked at it and made sure that this is not just a hallucination.

00:30:19 Mark Smith
Yeah, Universal basic income and IP.

00:30:25 Brian Perks
Exactly.

00:30:25 Mark Smith
What are your thoughts? What are your thoughts?

00:30:29 Brian Perks
So the CEO of CloudFlare made a very interesting presentation at Conn's Lions. And I'm going to butcher the stats, but directionally, three years ago, if a publisher who created the content had LinkedIn or ZoomIn or had Google scrape their page or Bing scrape their page, under fair use, they were getting 2 visitors to the website. In the last 12 months, that number has gone from 2 visitors to one scrape to tiny fractions, now thousands of a user. Something like 95% of Google queries are now being answered right on google.com as the AI function. Do not pass go, do not visit any of the content. I got what I needed and I left. And this is a problem for all of us. And the reason is, number one, most of these AI models, perplexity is one of the better at citing its work. But how do you cite the amalgamation of thousands of sources in these functions? It's difficult. But in the same way that we require any any professional, any kind of institutional document, you need to have the citation, you need to know where you can go and verify and build that trust. I think we need to see more of that in the AI function. And that is going to hand in hand eliminate the fact that these models are still dependent on human journalistic research.

00:32:13 Brian Perks
The AI cannot do that. The AI cannot head out into the field and do the interviews to collect that kind of data first party. And so I think the challenge is that if we don't start to protect the publishers who are actually generating the content, we're going to see a decline because the traffic and the revenue monetization for publishers, which has been in trouble. The stat is that in the last six months, it's now about 1000 scrapes. per visitor. This is all being done, perpetrated maybe, under fair use. And if you look at the four tenets of fair use, the 4th one is, there fair compensation or interruption in the monetization for the publisher? And I think that the 1000 to 1 stat I mean, CloudFlare is a content delivery network, sees all of this interaction, sees this transaction, is perfectly positioned to tell us exactly what's happening. And if they're telling us that there is an imbalance, we got to get this figured out. And Sam Altman's offering to pay for the content. We've got to figure out what we're going to do to make sure that the publishers are still getting paid, whether it's revision in the robot text, license models to where that license, that data can be licensed. But the publisher and the publisher model, we all like free media, we like free data, we like free Wi-Fi, and we like free content. And so, you know, it's an interesting function. I do think, though, that with all of these, you know, the AI is not immune to this. The AI is feeling the pain of, I just used your AI to make a multimillion-dollar company and you got nothing. right? Or you got $19 a month. That's where they feel the inequality and the publisher is in the same function. It's going to come from a change with residuals on what you create using AI. They're going to get an equity package just like any other employee. So it's going to become that employee equivalency model. And the publishers are going to have to be compensated in a way that allows them to be at parity of what they've been from a monetization standpoint, because it is expensive to do journalistic research.

00:34:31 Mark Smith
Yeah. It's going to be interesting to see whether regulation policy, et cetera, catches up in a timely manner, you know, before it eats itself, so to speak, that people stop publishing.

00:34:44 Brian Perks
I think it's dangerous, though, for us to think that the regulators have any idea of what is best. This is a clear case of the industry is going to need to find its own because no one else is going to save us. I mean, if you watched Zuckerberg and his commentary with, you know, our elected officials, you will very quickly figure out that they are not licensing experts either.

00:35:11 Mark Smith
Yeah, totally, totally. And I would say a lot of cowboy type nature happening. based on recent books I've read about meta and stuff, flying by the seat of the pants. Interesting. This has been such a great discussion. Before I let you go, tell me about, you know, what homework do you set for the listener? And that could be top books that you recommend, movies to go watch, sites to go thing, knowledge to acquire. Give me 3 things that pop to mind with this question.

00:35:46 Brian Perks
I am a big believer that one of the things that I have found is I am investing in some entrepreneurs that I work with. I am learning more from guys like Josh Perk at Vector that are out there doing really interesting things. And you think as an entrepreneur with some gray hair that you are going to teach them a lot. And I am amazed at what I continue to learn from these guys. Daniel Weiner from Autobound is another guy that is a great listener and then spits out these really insightful things that help us with our business at Five by Five. So that is a huge function is investing in and having a core group of entrepreneurs that you can bounce ideas off of. The other function that I am finding here is I am always looking for anyone who is taking and building specific models that are unique to what I'm trying to do. It amazes me what's coming out, the applications, the products. And I'm finding that general, these general AI models are great. You know, jack of all trades, master of none. But as soon as you start to get into, you know, Claude from a marketing standpoint, you're really amazed at, you know, what that additional reference data can really do to improve the AI model. And then the third thing that I tell people to do is get out, go for a walk, get moving, get your brain in gear. It is amazing how that just is a reset. And I even find myself taking my own advice there regularly. But those have been the things that I have been leaning in on as of late.

00:37:33 Mark Smith
You've been listening to AI Unfiltered with me, Mark Smith. If you enjoyed this episode and want to share a little kindness, please leave a review. To learn more or connect with today's guest, check out the show notes. Thank you for tuning in. I'll see you next time, where we'll continue to uncover AI's true potential one conversation at a time.