#43 - The AI Playbook for Booking & Marketing Shows

[MUSIC] >> All right, it's pleasure to have you back on. It's been about a year since you came on last. >> A year to this date, this is Matt Pat Day. June 22nd will always be known as Matt and Pat podcast day. >> The summer solstice. The longest day of the year pretty much. And this will be hopefully not the longest podcast in prism, live industry podcast history. So good to see you, man. >> You know I used to throw a music festival in the summer solstice, right? >> I did not know that. That would be the best. >> Yeah. >> It's day up all night. >> It was called the Solstice Festival pre-prism days. But it was a lot of fun. That's a tell you story and other time. >> It went so well you stopped doing it or what? >> Let's go let into prison when I was a promoter and I did this festival. And I honestly never intended it to become a thing. And then like four years into doing it, it was growing and picking up momentum. And it was a beautiful experience. And it was a key part of what got me into prism. But when prism was taken off, it was time to set down being a promoter. I'll leave that to the pros and. >> Part of me wishes. >> I think for a while I felt so, I felt like such an outsider in the industry. And because I never had that story, right? I think you know the story, but like, Minion started a ticketing company 14 years ago. Just because we like going to clubs. We were just like two kids in college who like going to show. So we started a ticketing company because that seemed like the fun thing to do. And it was fun. But we never really got that thing. Obviously never took off. It's hard to start a ticketing company. And now we're here. So, but I wish I would start off as a ticketing company. That's always a funny thing that I forget about y'all. >> Ticket. >> Ticketlabs.ca. We couldn't even get a dot com. We had a dot c a domain. That's how Canadian we were. You know? So, yeah. >> Was that like a long? Were you pridefully going with a dot c a? >> No, no, it was just cheap. We were broke. We couldn't afford the dot com. I think we ended up buying the dot com. It was a couple thousand dollars. The good ones are. But yeah, we had a dot c a domain for a while. So it turns out nobody really cares. >> Very classically before the podcast, we talked about what we're going to do. We're going to talk about it. And now here we are meandering in the first two minutes. >> Oh yeah. It's amazing. It's amazing. But you and I were at a, where this podcast came about is we were at Niva and got the opportunity to do a little fireside chat about AI and Prism's new product insights. And the whole thing was about 15 minutes long and both of us were like, then we could have talked for two hours. So let's do a follow on podcast to let that breathe. But it was an awesome conversation. I think the room received it really well. You and I have been following AI for a long time. Coincidentally, you were in the very early experience there during the start of high. I think someone at one of the frontier AI companies may not have pitched in one night when that was getting off the ground. We've both been following AI from the start. The mode has really shifted in the industry. But without further ado, why don't you give a really beautiful intro to that panel? Why don't you talk a little bit about what your perspective is right now on AI in the industry? >> Yeah, I think we had set it up by basically, first of all, just acknowledging and how I used the word "blast." You used it better word. You didn't like the word "blast." But how fortunate we are. You said privilege and I said, "Blast is a better word." >> Yeah. >> Blast, okay. Sorry. You said, "Blast." But how fortunate we are to get to work in such a fun space on something that we, and quite frankly, all of our customers are so passionate about. That is live events to be clear. You're on the wrong podcast if you're hoping we'd be talking about something else today. But we were kind of just like, man, how cool is it that we get to work for a server of people that are so passionate just like us. But yeah, we get to do it from the technology perspective, right? And we get to kind of like make those two worlds collide. So whether it's AI or fucking big data or whatever the buzzword of the day is, we were just kind of reflecting about how fortunate we were to be able to think about how we can take the new technology or the new cutting edge shit and apply it to the problems of the space that we work in, right? Figure out how we take all of the hundreds and thousands and thousands of customers, music venues, promoters, artists that we work with and figure out how we can take new technology that your mom has never even heard about before, but help people in life solve their problems. So I think we sort of, we opened it up like that. I think at least for me, you keep me honest, Matt. I think I also feel like we kind of have like a duty or a responsibility. And in some cases, like an opportunity or the right to also, we sometimes use the word like Sherpa along everyone into the AI age, right? And it's like you're a promoter, you're a music venue, you're a marketer, you're a booker, you're a buyer, whatever. Like, you shouldn't have to be keeping up with every single time and thrott big drops in your model, right? You shouldn't have to be trying out all of the 50 different, you know, AI transcription companies on your Zoom calls, like that technology is going to come and go. That doesn't, that in certain cases will probably help your business, but I think like we have a, we have a, we have the privilege, we have the, we are blessed to be able to take and say, okay, well, and thrott big just dropped a new model. What does that mean for, you know, the clients that we support and what would hypothetically the best AI product for live events look like? And I think like that was where you and me were coming from was like, we have a couple bets, we have a couple things that are working that like real promoters, real venues are using and wrapping their hands around and telling, you know, their friends or their, whatever, people they know about. And it's funny to kind of have that perspective. So I think we sort of like, I think that's how we sat down. We first of all, like, I think the room full of people. I think there was a lot of, there was a pretty, Neva's always a pretty friendly place for, for, you know, high and high and prism to visit. But I think it was, it was one of those cool moments where it's like, okay, very friendly place. Very friendly place. A lot of fun too. But it was, but we sort of sat down like, all right, listen, a lot has changed. Certainly, a lot has changed in the, in the years since we did this pod, right? I think we talked about, you know, the early kind of formings of AI in our products a year ago. And a lot has changed both in our products. Certainly in the market, the perception, the technology, what it's good at has changed. So I think like that sort of sets us up really well today to have, you know, a V2 of our conversation around AI. Yeah, I want to talk about what both of us are doing in AI. Something that you said here, which is really interesting is, you know, you kind of want to be a filter for the, the fire hose of what's happening in AI. And, and, and that, you know, music venue owners and promoters shouldn't have to be so as long as the prisms and the hives of their world kind of do that work of saying, all right, like what, what's usable in this, you know, in this renaissance and what's not usable. And, and how can we, you know, kind of take AI and put it into your existing workflows? I, I really like that sentiment. One thing that I love about you, like, I think at one point during one of our hangs, you like sat me down and like, look me in the eye and you're like, am I like too deep into this or something like that? And I was like, look, sometimes you tell me stuff that I'm not sure that you are fully bought into yet. But you're like, like, I'm a few inches back from the fire hose and you're, you're like, you're like rate, rate in it. And, and I love that about you. So like, how do you stay in the loop of what's happening and like, and what's signal and what's noise and you're, you're magnificent at doing it. And a lot of, I've learned a lot from you in, in AI just because it's like, oh, wow, like I didn't, you know, I heard about co-work from Ian for the first time and, yeah, harnesses and, you know, even when this was getting off the ground. So, so what is your approach for, you know, just tracking the signal through the noise and, and staying in the loop with all this? Because it's, it's chaos. Yeah, I think I'm, sometimes I feel just fucking manic, you know? So it is, it is, it's, I'm like a dog bite and, you know, trying to bite water coming out of a hose. That's, that's where my brain went. A few things, I think we are fortunate in that I have like some of my best friends, some of the people I worked with, you know, a decade ago, some of the folks that, you know, we started companies with around the same time. They still live in the Bay Area. They are like literally the ones working at these companies, right? They are the ones that are on their second and third and fourth AI started in the last five years. Those folks are the ones that are seeing the stuff and coming up with the words, you know, months before it hits TechCruncher. Those are the ones that are getting into Twitter fights with the funny folks on, on the internet. So, I'm, I try to maintain those relationships. Maybe that's a little bit selfish, but that is how, that's how, that's how we find out about that stuff, right? At the end of the day, like, you know, if you were going to go invest and start up to where, where, you know, lots of the great ones getting started, it's the Bay Area. So, you know, you keep your tabs in there and you visit those people enough and you exchange texts and you figure out what they're working on, what they're thinking about, what they are investing in right now. So, candidly, I think that's how we partially, you know, get a little bit of it. I love Reddit, dude. I'm a total Reddit nerd. I don't know about you. And I'm just like, scrolling Reddit all the time. It's surprising how early some of the conversations that happen there are if you know where to look, right? If you can get around all the other shit that's there, you know, Twitter acts whatever is another good spot for it. But a lot of it happens. No, no, no. I got two kids. I got no time. I wish I could be on Twitter. But I think part of it is just like, you know, having enough of a cadence to be able to catch the trays that they're coming by, but then I value like conversations with you or conversations with Ian, my co-founder, and being able to say, okay, this crazy thing happened. What do you think is the right take for us? What is the right take for our customer base, right? What is the right take on this new thing that's happening for our technology? Like, how does that, you know, how is that a good thing? How is it a bad thing, right? How should we mitigate against it? What does it mean for live events and live music that this new thing is coming down? So we have a lot of those conversations and sometimes it requires you like physically going somewhere and, you know, sitting next to a lake for a day or like sitting in a cabin for a weekend and like actually trying to think and be real about it. I think the other way that we stay on top of it is like, we prototype a lot of stuff. We try to build and build and build and build and build. So, you know, the AI stuff that's hitting for us right now was like the third or fourth prototype, right? Like, we have a real team of engineers and designers and marketers and managers like just circling around trying and building with all of the newest technology, right? So I think like you have to be able to read about it. You have to have a good filter on like what you should you actually invest in. And then when you see something, you got to call that fucking shot and go for it. And that's been, you know, more of a new muscle for us over the last couple years, but that's where all of our, you know, our new product comes from. It's from these things that were just like, you know, threads on the internet, turning into like conversations with the in and pat, turning into, okay, let's pull a few more people in, turning into like our 18 years, six months in a roadmap. Let's go try it out, you know? So it's tough, but like that's our job, right? And that's what we got here in the first place. So it's fun to get to do that against. I'm glad we're doing with AI, the whole crypto thing. Never really clicked for me, but like the AI stuff's been cool. And now importantly, we're starting to see it like actually hit product and actually be used by our customers, by our clients and they're selling tickets with AI, whether they, you know, whether they know it or not. So it's pretty cool. You weren't doing any like NFT emails. Email man, I had a, I had a job, you know? No, the crypto thing was like, I mean, I remember there was like a brief moment where I was like, oh man, maybe there, maybe Prism could have a it's own currency in the live music. We got to be careful, dude, this is going on the internet, man, omit at kid lives is going to come after us. Because we're talking crypto, you know, shout out to omits. I'm kidding. Yeah, I mean, there is, there is signal in that noise. I think the noise, the signal ratio for crypto is like much higher than AI. The trippy thing that you and I were talking about this like a year and a half ago, which is hilarious. Like the world is moving so fast these days. Like we basically first started sitting down and talking about this a year and a half ago at the Aspen Live Conference. It's crazy to think that that was just a year and a half ago because so much has happened since then. And, man, the signal to noise ratio in crypto, I think was extremely high. It's extremely high noise, very little signal, not zero signal, but very little signal. The weird thing about AI is the noise is extremely high, but the signal is also really high. You know, so really kind of discerning like what is sustainable for this? Like what's here to stay? What's not here to stay? It's been, I mean, I think this is why I've been talking about it so much on the podcast and with you and with other founders in the space. Like, you know, behind, behind closed, sometimes in open door settings. Just because it is a lot of, it's like, wow, you know, like this is a big change for the world. And look, in some ways the more things change, the more they stay the same. You know, like I think maybe we should get back, let's go on the conversation and like what is, what is it that you're actually doing with AI and how much? Because I think people are at the end of the day in terms of like the more things change, the more they say the same. They're still using hive, right? They're building emails and text messages in hive. But now there's this AI layer that you're starting to introduce. So like let's get in the weeds there. Like what is this AI product that you've been working on? Cool. Well, this is product, you know, three or four and lots of folks that are, will hopefully listen to this wonderful podcast, Matt. Probably have seen products 1, 2, 3, 4 because just like we're not afraid to try to build stuff, we also, you know, it's not clicking me throughout away. We focus on the next thing. One really interesting thing that we experimented with last year was like literally a chop. And this was like September, like this was a while ago. It feels like a while ago. You know, it's like chat with an AI in hive and it could do things for you, right? It was writing text messages at the time. It was like, oh, I have this event coming up. You know, recommend me some messages and it would suggest, oh, you could send a low ticket inventory warning message and hear how it looks. But it just felt a little too futuristic and and felt a little bit odd to that that was the only way that that you really interact with, you know, with a product like ours because there's so much stuff that that people are doing every single day in our product as, you know, they're looking at their data and they're building emails and SMSes and ads and segmenting and firing their data off to different services, etc. So one thing that was that that was very interesting was a rate around the time when we started adopting Cloud Co-Work, which is a really great tool that helps you use AI to, you know, do your daily workflows, whatever, do do research, drop documents, take action in other products. We're like, man, this is this is really cool. There's a version of this that I think our customers, you know, our clients, the venues and the promoters that we work with, like they would lose their minds if they had access to something like this on top of all of the interesting stuff that I was also in high. So on what if, you know, what if you had access to all of your ticketing data? What if you had access to, you know, everyone who's ever bought tickets, all of the events you've ever played out at your venue? What if an AI product was also able then to take the right pieces of that and draft your emails, draft your SMSes, build your ad campaigns for you? We had also started to feel an interesting shift inter customer base away from a year ago, which was true, which was like pretty AI hesitant towards, like definitely AI curious and like I started to hear it was mostly like upper end, you know, sports teams or like the really professional, professional teams that we work with, starting to actually try, they were trying to buy AI agents, right? They had been to like a Salesforce conference and they like were using the word agent. We were like, okay, maybe we started to kind of feel the ground shift a little bit and people were a little more open to actually trying to use AI tools in their workflows, which again was not the case, you know, 2024 or early 2025. So what we have been building, and this is all in, we're working with a couple key design, key design partners. So it's all in closed data, but we've been actively throwing more and more and more people into those beta. We call it marketing assistance, high marketing assistance. So they are AI agents, but because most of our customers don't know what an agent is or you know, or maybe a little hesitant or just sort of starting to become curious about AI, like we don't call them AI agents. They are marketing assistants and they live inside your high account. They operate on top of your data that's inside of high. They can do things just as you could do inside of your high account as if you were clicking all the buttons and navigating all the pages. But what's really cool is that you can, in your own words, describe what you want these things to do and they just work in the background. So for instance, you're able to say, "Okay, well, you know, a week before every single show, I need to send a one week out email. Here's, you know, the template I normally use. Here's how I like to target these things. Here's the type of content that I want you to put in it. And these things they run in the background on your behalf following those instructions. And then they present everything that they do to you for approval. So we built like a thing called the assistant inbox. New as a marketer get an email notification that's like, "Yo, Matt, you know, your one week out show for such and such, you know, a week from now is ready for your review. You log in as you give it feedback. It learns, it gets better over time. And you're able to kind of steer, steer how it works there. And now it's freeing up marketers to actually be able to go and start to do all of the work that they really should be doing, but they've been too slam because they don't have time. So we call those marketing assistants. The first few use cases were like the obvious things you want to do. My day of show, my number four you go is my thanks for coming. It's my just announced, right? But as you start to give these tools to people who are naturally creative and way more talented than, you know, me as a marketer, they've started to do some really crazy shit. These assistants, they don't just have access to, you know, your ticketing data, your event data, your customer data, all the past emails SMS you've ever sent. They also have access to the internet. So they're able to go out and do like research. So they know like what the weather's going to be this weekend. They know that people's going on right now. They know that, you know, they know that there's construction downtown Minneapolis. And you need to, you know, here's how it affects parking. Like they know all of these things or you can make them aware of all of these things. So it's been fascinating to see how people incorporate that into their marketing strategies. One of the coolest prototypes, the coolest thing that we've been, that we've seen working is for some of our clients that have, you know, 50, 80, 100 upcoming shows that have been announced, but not yet played out. These agents are going in the background and researching the artists that are playing at these shows. And they're able would say like, okay, you know, Diplo just was nominated for a Grammy. Still two months out. Normally we wouldn't do any marketing, right? We're in that maintenance window, but it's able to say, oh, Diplo is just nominated for a Grammy. He just dropped a new EP, just released a new boiler room set. It just, you know, was in the news for some other reason, and it's able to take those news early moments and package them up into marketing, which is really cool. Draft all your email, draft all your SMS. So it's been really fascinating because marketers are able just to kind of say, yeah, well, if I had infinite amounts of time, here's how I would do my marketing. And just from being able to explain that in words, these agents behave pretty well out of the box, and they're able to do all of that work on people's behalf. So it's been a really, it's been a really interesting thing to see where now the thing that is limiting the functionality and the product and the way it works is actually like the taste or the domain expertise of marketers as opposed to like, you know, our engineer's ability or something like that. You're literally able to take a great idea and express it in words. And then these agents are able to start to operate. So it's been cool. You know, our design partners, they went from having maybe, you know, 5% or 10% of their marketing volume every week. You drafted by AI to now we're like 40. I think we had someone hit like 55% last week, which is really cool. So we're starting to see that as you how do you measure that 55% drafted? Yeah. Like what does that mean? Yeah. It means like out of all of the messages that get sent in a week, half of them started off like the reason that they were, they were presented to the users because AI thought that it would be a good thing to do. Put it in front of them. They can tweak it, tweak it, tweak it, and then they hit, you know, schedule or approve. So it's been crazy to see like the first couple weeks like it was hot garbage. It was bad, right? But after a couple good feedback loops and be able to say, Oh, no, it didn't like how it did this or I wish it would have done this or I went ahead and made the edits myself and then our system, you know, can can can learn and retrain from that. It's been fascinating to see how good the results are across stuff that it wasn't very good at before, right? Copy generation it's killer at can write great subject lines that convert like crazy. It can pick really good event recommendations that should go in the footer of, you know, your, your know before you go as your thanks for coming. It's like it's even it's even starting to give future events. Yeah, that are very well tailored to the event that's playing out, right? Same thing with targeting, right? Targeting was something that we talked about. I think a year ago now, but we've been doing AI targeting on our just affinity segments and for a really long time. But for it to be able to say like, Okay, I know that these two people don't sound the same, but here's why it's a really good idea from a targeting perspective. Maybe, you know, they had some weird crossover or some collaboration in the past. Maybe their audiences have overlapped for this other like non-obvious reason. So targeting is another thing that these agents are getting really, really good at. And again, it's all a little talk about it'll promote five or six other emails or whatever in the footer. Sorry, five or six other, five or six other events exactly. Yeah. Yeah. I think I want to zoom off for a second and just say like an important context here in you and I, you and I have gotten this just for some kind of, you know, in depth conversations that you and I have had. I think your worldview on marketing is really kind of pertinent to understand like what's behind hive and it's that basically like every music venue has a community and you're going to say it better than me, but I'll start. But yeah, my understanding of what your worldview is that everyone, every music venue has a community that effectively could, you know, sell out every show. And look, and like maybe, you know, selling out a show has, depends a great deal on artist the man. It sells every ticket. If it's if it's properly engaged with it, could sell every single ticket that's meant to be sold. IE, like a lot of a lot of marketing is just simply, you know, awareness. Like you and I talk about this all the time. Like we miss shows that we, we would want to go to. It's very, it's very rare that a music venue convinces me to go to a show that I don't, I don't know the artist about, I mean, I don't know, ever, especially with three kids now or whatever, you know, collectively. What do you mean, or whatever? I think you get three. It's exactly three. Yeah. Maybe more, you know, we'll see we'll see what God has in store. But yeah, at this point, I go to shows that I'm like, you know, that I know the artist, but I just I miss the vast majority of shows that I would go to. So it's really a music venues job is to make sure that everyone that could buy a ticket has the opportunity to buy a ticket. And that's essentially accomplished through Hive's tool. It's like tax messages, email and yeah, yeah. Email and ads. Yeah. And whether you know, you're trying to acquire new customers who and new people to come to your venue that, you know, you've never had come before or you're trying to talk to people that have already purchased tickets in the past. The thing that we've seen is that the best marketing teams in the world are accomplishing their goals by doing what we call default good marketing or just like good marketing, right? And as data has changed over time, what's default good marketing? Good. Well, that's what I'm saying. As data has changed over time and data availability, availability has changed over time. That definition has shifted because new things have become possible. So like, you know, or five years ago, just getting, you know, your email newsletter out every single week, every single month was like the gold star, right? You do that. That's the best kind of email marketing you can do. It's cheaper to, you know, sell tickets over an email newsletter to people who have already bought them than it was. So, you know, buy them, acquire them off a Facebook with ads every single time, right? Maybe three years ago is two years ago is when we really started to see across the board, um, take it by your data, actually be able to be programmatically accessed from all of the ticketing companies, um, so that you could plug them into a tool like Hive or, you know, other tools that are out there to be able to say, okay, genre by genre. Now I'm going to start segmenting by genre, right? So you started seeing people do genre based lists or genre based newsletters. But now what we've seen is that the best teams in the world are actually taking a step, they're going even deeper and they're saying, okay, well like country music, country music, EDM, EDM, if I'm a club and all I do is electronic music, genre is, you know, it's not granular enough. So now what they're starting to do is show specific marketing, they're marketing individual shows over email over SMS over ads, high affinity ticket buyers based off of artist affinity. So it's much more nuanced than, you know, it's ever been in the past, it is unlocked by ticket buyer data, you know, at scale, like, like now we're able to access on behalf of promoters and venues. But, uh, but it comes with a huge increase in workload, right? So now if you have, you know, I always say like if you have five shows every weekend, you know, you have a month out, a week out, like a one day out, that's three, you have a, uh, no before you go to ticket buyers, so they know like what your bank policy is, you have a thanks for coming to cross-sell future shows. So that's five, five messages for five shows is 25 messages. That's just email, you need to layer on SMS, right? That's 50 messages and marketing teams are just underwater, right? So default good, when we say default good marketing is getting out the 20, 30, 50, 60 messages that you need to get out every single week, but doing it consistently with your brand and your voice, right? Making sure that you're doing the right amount of traffic control and you're not over-messaging your audience, right? And making good strategic decisions as it comes to that, mixing your channels really well, how much are you, you know, to playing on paid, how much are you playing over owned like SMS and email, but, but that, that like increase in workload is now the thing that I think is holding back all of the great marketers in the world from operating like the teams that I've been doing this for the last couple years because they are able just to like hire more people and click more buttons and analyze more data, right? So I think like one of the coolest opportunities that we see, especially for like that, you know, mid or independent market is like, um, democratizing access to those, those marketing techniques through the use of an AI-powered tool like like like we have with Hive. And that's been, that's been one of the coolest unlock. So right now it's like replacing half of your marketing with AI-generated stuff. Soon it is like roughly everything that you're producing with Hive is in one way or another touch by AI, optimized or drafted by AI, and then it's starting to get you to do even more than you've done before just because you haven't had, you know, the resources available. So, so as you unlock more data, you unlock more use, you know, more use cases and increases your workload, but you need, you know, you need a tool that can kind of keep up with it. So thankfully like, or I'm thankful that like it seems it is there. People are using it today. It seems to be working. But I I believe that that that's a really important piece of making sure that everybody that should come to a show, at least to know is about it and has the opportunity to buy. I would even challenge too. Like we've seen great marketers do like sounds like this type of really good sounds like this type of marketing. So even with some of their new acts, their audience is now getting trained that like when they make a recommendation for you to come to a show that you've never bought to before, they don't just say like DJ X Y Zed is performing at 930 Club, right? They are like, this is this person's first time. You've never heard them before, but here's why we think you would be so into it, right? Yeah, let me, let me, let me, yeah, let me qualify that. Like I'm, I'm not, you know, with three kids and at 36, like I'm, I'm aware that like I'm not the only profile of a, of a, um, a ticket buyer. So, you know, for people like me, it's about making sure that I'm aware of a show, but I'm sure there's a lot of people who just go and frequent venues and like and show up side unseen to a band. Um, and yeah, so I want to, I wonder, I wonder what that split is. We should, we should pull it. Yeah, it's, yeah, Yeah. 'Cause I think about myself too, it's just like, there are so many people, like I will travel to GoSea. There are so many great artists that I love so much, right? That I would do anything to go see them, and then I just fucking find out the day before, and I don't go. - Yeah. - It's like, oh, 'cause it's a two hour drive or whatever. So it's like, so anyways, we, does that answer your question, Mr. Ford? - It does, yeah, something. - Curious, yeah, for sure, for sure. So something curious came through, there where AI is oftentimes talked about as like, like maybe you can have a high quantity, a higher quantity, but like, it's gonna come at the cost of quality or something like that. And like if you think about, if there was a, you said something really interesting that like, okay, if I was in a vacuum, here's how I would do marketing, well, you're not in a vacuum. So here's the tools that could actually get you there. Like you can make a case that, like at first, like maybe there's, you know, like you could make a case that AI takes a hit with quantity, but you could, I think there's actually a stronger case that quality goes up because now you have systems in place that could send every kind of communication that you wanna send and you have the spaces of marketer to like put the quality on it that you want because it just gets through so much grunt work of the scheduling and the team up and the segmenting. And you know, I guess that would be the holy grail is like, you know, both the quantity and quality increase. But yeah, that was something that just came through as you were talking, I don't know if you spoke. - Yeah, and maybe like quality, maybe like nuance, right? Like there's so much nuance that has to go into the, offer the messaging, the channel selection, the timing, right, there's just so much nuance there. I think one thing that we've learned from speaking to so many of our customers about it, relentlessly is also like, it is really important to have humans in the loop, right? Like our stuff is, it doesn't just, right now, automatically send everything out, right? There is always an important checkpoint, whether you're doing a proof or whether you're getting artist approval or whether you're getting approval from your higher ups or what have you, right? There is always these review and approved moments that are baked throughout the product and the experience because it's not about automating everything so that it's AI slot, nuance AI slot. But like it is not about that, it is about, again, getting through all of the repetitive or the overly complex stuff or the things that require a ton of information, like synthesis down, like research and then targeting is a really good example of that, right? It is about getting through all of those things so that humans can make the critical decisions and steer it with taste and with approval flows. So that's been something that has always been core, I mean, for years now because it was clear that that's probably a good design practice, but especially in live events where there's creatives and maybe we don't see as much blind adoption as in other industries. So that's definitely been something that we believe a lot in and I think it's been a good call. - So in your best case example, there's clients now that 55% of the emails started with an AI agent. Do you have data that shows how many of those times did they, so did they edit the email? And for those that don't know, like the way that this product works, I've seen it demoed, it's amazing and potentially we're gonna send out a, a, a, a, a video showing it off. But if you don't see it now, you know, have a little eventually be, you know, pushing it out and putting, putting demos up online. But the way that it works is you show up in the morning and there's 20 like pre-drafted emails by, by hive. I, sorry, by the hive agents and marketing assistants, marketing assistants, yeah, by the marketing assistants. - Yeah, yeah, good, good. - Yeah, marketing assistants, which are little AI robots or whatever, nail in it. - Yeah, and in one, just some magical box, little kind of, you know, fairies in the, in the pixel. - That's how, that is how AI works, you're right. (laughing) - And, and you can, you don't just send the email, you can, you can send it or you can edit it. And like, and then if you edit it, you just pop right back into your, you know, beautiful hive kind of text that, when you know that you're used to and you can switch around the images and all that. So my question is, in those, in that situation where 55% are, you know, we're started by AI, what percentage of that 55% was kind of one-shotted by the AI and what, what percentage was edited and like how much was it edited and you don't have data on this, like that's cool. It's just a hundred percent of it, hundred percent of it gets edited. I think that's really important. - Oh wow, 100% gets edited. - 100%. Yeah, I mean, it's early, right? We're talking like design partners have had this for weeks, maybe, maybe two months, right? It's new, but, but importantly, like, it's not the AI's fault, right? It is stuff that like, it is limited in, for instance, limited information, right? You'd be surprised, like, just getting the right flyer or the right ad-mat for an event to put it into an email, like, it doesn't live in ticket master, it doesn't live in C tickets, right? That ad-mat, the right aspect ratio only lives in the drop box of, you know, the club's marketer computer, right? So, like, the AI, we haven't plugged them in yet to drop box as an example, right? But, like, just making sure that it has the right creative is an important thing that almost every single time, there's a tweak to some sort of copy. They need to upload some asset that, like, the system just doesn't have. So, in that case, like, the marketer's still doing the work that they've done before, you know, you remember it would save it, it uses it going forward, it gets better over time, right? But in almost every single case, it's important because there's always just, like, there's little bits, right? There's little bits and bobs and that's not something that, yet, I think the system fully takes advantage of yet, but, like, obviously, we're working on a Google Drive integration and the drop box integration so that it can just go and browse your entire asset library to find the right creative. But, from, like, we always say, like, the data that lives in your ticketing site, like, when your team does the build, is often very different information than you actually need for marketing. So, we, I mean, we have workflows and teams just built around trying to get the right marketing information for live events into it. - Yeah. - So, but, anyway, so that's nuanced, but you hear what I'm saying. - No, it's interesting and just, you know, it's funny because at the Niva invite, we only had, are the Niva, far I said, chat, we only had 15 minutes and now we only have an hour and a half or whatever, and I'm like, look at the clock and I'm like, man, there's a bunch of topics that I want to cover. And we have like 35 minutes or so, which sounds like a lot, but, you know, with the United, that's like three or four questions. Well, let's zoom out for a second and just talk about, like, what we've seen, like, chat GPT was basically launched three years ago, was it, it was in 2024. So, we're like in, was it, was it 2024? Or was it 2023? - I mean, they had early shitty versions of it forever. - It got great, yeah, it got really good in, - In 2024. - Yeah, yeah. - Yeah, it was in '23, okay, okay. And, so we're like two and a half years in now to like the launch of chat GPT, and you and I are like, you know, whatever, friends of people in Palo Alto and, you know, we're like really tapped into like the early days of it launching and knew some of the people that like, you know, started the whole darn thing. And, you know, the prophecies early on were just, you know, like, sales force is gonna topple over and fall. Like, their sales force is skyscraper is gonna become crumbling down or, you know, like Elon Musk is out there still to this day saying, like, you know, humans will not have to work. Like, don't, don't save up, don't save up money. You know, like, that's a ridiculous task now because there's gonna be like, suit, you know, super abundance because of computers and robots. And, you know, a lot of people were like, wow, like, is my job, is my, you know, purpose as a human, is it becoming extinct essentially? And, you know, you and I early on were like, and then the question then like came to software companies too, which is like, and especially as these machines got better and better at like writing software, like it became this really curious question. I don't know how many people are familiar in this podcast of the term like vibe code, but yeah, basically you can get on a clawed and like, and make a software application like very, very fast. Now, is it quality? That's a whole nother conversation. But anyway, we're like two and a half, three years in now to this being a part of the zeitgeist. And I'm curious to get your take. I mean, my take is that there's this like line of human value that AI is approaching, but like has not crossed. And it's like, I believe that it's asymptotically approaching at least with like this version of AI where humans are gonna be, you know, forever valuable. And even as AI in the last two and a half years has gotten better and better and better and better at things. Like what I see is these are just tools that humans are picking up ultimately. And I don't see any sign of like AI just kind of like running society without a very intelligent human using the tool and like an insightful way. So I don't know, am I preaching to the choir here with saying that and like, what's been your view of like how AI has, you know, developed over the years. And yeah, just go there for a second to podcast. That's some deep shit, Matt. (laughing) I don't know, man. I think candidly like I don't really think about it that, I don't try to plague seven dimensional chests like you did. I think that it's definitely accelerating. I think like, the pace at which releases are happening at frontier companies, the scale at which the data and the scale that they're using to train, you see it accelerating and accelerating. I don't know if the world's going to end or not, and we're all going to be subpoena clouters on the beach, but I like to think about it more as like, how can it save me time, how can it save my team time, how can we make sure that we're taking advantage of it to serve our customers really, really, really well, but like I have no fucking idea, dude. And it is curious why like some of the smartest, best engineers in the world are either like quitting and leaving or joining the frontier companies. Like there's something really interesting there, but it's hard to know if that's just quitting and leaving the frontier companies. They are quitting tech, moving to the beaches or bunkers, depending on your view, or they are like saying, I need to go work for these frontier companies, right? The anthropics of the world and I don't know, but also like we work in tech, like we have, I have never seen of any of our vendors that we've ever paid for in, you know, the 14, 15 years you've been around. I have never seen a spend ramp as hard as it did like anthropic now, right? And it's coming up the expense of other tools that we're using, but other than maybe AWS, like there has been-- >> Your spend ramp of internally. >> So let's talk about that. Like now companies like Hive and Prism beneath the scene are using AI to augment software development. And let's have a really, really conversation. How helpful is it and how much can you realistically say that like the product roadmap is like moving faster because of AI when you like net out the cost of quality? Because yeah, in long story short, for everyone that's like listening in software development, there's this, you know, quantity versus quality perspective. Like, yeah, you could like go into cloud today and like one shot, like a sales port. You could just promise, give me a sales force replacement. But then, and then you say, well, it needs to do this and it needs to do this and you have a thousand prompts later and then you have one a million of what sales force does. So anyway, so that's what I'm referencing about quantity versus quality. So yeah, how is it augmenting Hive? And then we can move to like, we went to how it's augmenting to customer about how is it augmenting your development. Yeah. We purposefully set up sandboxes. We purposefully separated out all of the different work streams that our team works on to call out areas where or themes of work where we need to run fast. We need to vibe code. We need to have the right amount of guardrails in place, but where we like favor velocity, right? Or experimentation and customer feedback and all of that stuff. And then there is stuff that we do like many other orgs that like absolutely that is not the case, right? We, we, we will use AI assisted tools to write code and to build up the infrastructure and to run all the important platforms and services that we need to support, but like it is not it is not a move fast and break things part of our part of our or so. Being able to separate out the two felt prudent. It was also the only thing we could do, right? To actually be able to start running early on. So there are certainly like there are pieces of product that customers using today that were fully vibe coded reviewed by humans, you know, good test cases, but but didn't go through a normal software development life cycle, right? They didn't have a necessarily designer, you know, stress over every single pixel or or the copy, you know, wasn't fully vetted by our by our product marketing team. So, so we think about it as like move fast, like fast thinking, slow thinking areas. But on places where like teams are given fully way to run as fast as they want, whether it's, it's more prototyping experimental type of stuff that we just want to get in front of customers to get feedback. Like, you know, we're we're shipping stuff that we would have, you know, would have otherwise never been on our roadmap. There's there then and then on the flip side, there's there's the exact opposite, right? Like we, you know, we support whatever 1600 event promoters and venues and those are real businesses, right? And if if their emails don't get sent, they don't sell tickets. People don't know where to park. They don't know what time to show up and like so they they they they rely so much on, you know, the services that we provide that we just we don't risk it there. So everything is moving roughly quicker, but and in some cases like light speed, but but it's not like we were able to just close our eyes and say, okay, everyone's allowed to vibe code everything because we, you know, we have a certain service part uphold. So we're constantly working on, you know, getting efficiencies out of the org, but but it wasn't just like I think anyone who's like, we're moving one e times faster. It's like, well, maybe in certain places, but, you know, if you have a real if you have real businesses and real people relying on your shit, you still have to move with the right amount of reliability and, you know, service quality. So that's that's my long answer. But separating it out, it's really important at least for us. Yeah, one thing and just kind of moving through a few different topics here, you know, I guess going from like science fiction and robots taken over and then go into like how how high is using it and now moving it into like, you know, just day to day like quality of life with AI. A topic that we, you know, right after you and I went there was another AI panel and it focused primarily on like distribution and like getting your events listed in in chat GPT and cursor, sorry, cloud and Gemini and you know, how how to do that effectively and how SEO is changing search and optimization is changing changing to like artificial intelligence optimization. It's technically a EO and I don't you know, I'm lost on the acronyms at this point and you're doing pretty good. Thank you. Thank you. You went to IVMC as well and there was another AI chat that was focused on that and I think that's a very important interesting topic but you and I both, you know, felt that there was a vacuum of how people are not just using AI to like get their events listed but are using AI to like improve their work streams and a lot of Niva and IAM, you know, like co you would say, hey, I'm using co-work and I said, what's that? And that was honestly shocking to hear because co-work is such a unbelievable innovation but you know, I think anyway, so I wanted to get your take on that and then you know, prisms also doing we launch our MCP server and the feedback that we're getting from clients is like really amazing and it's it's basically allowing AI to you, you can now, you know, connect your prism data in a secure way to a LLM of your choice whether it's GROC or CLAWD or cursor and so I keep saying cursor, GROC or CLAWD or chat GBT and you know, ask a question like, you know, run reports, you could be talking to it on the way to the airport and say, what's the show that we have coming up? Like the other day I was on the phone with the client and he was like, oh, you know, what's a, we're going to a conference and he was like, oh, you know, do a, do a happy hour out of venue and he's like, let me go into prism and you know, and check, I was like, oh, just pull out your phone and ask CLAWD and he did it on the phone, oh, the dates were available. Yeah, just a little like lightning enhancement. So, you know, I think, yeah, if you get out of the science fiction world, there is a tremendous amount of value to be had and adopting prisms AI, hives AI, CLAWD, CORE work and then there's the whole like data analysis thing, analysis thing, which like I would love to talk to you about and come over the mic to you as you becoming the host and you know, sharing some of the stuff that we're doing. Hey, insights AI, but anyway, there's a totally, totally, totally, words that I just threw at you. Yeah, it's funny, like so many of the massive, massive players in live events, this is my thesis, is that so many of the massive players, mostly taking companies, primary, secondary, whatever, their entire business is, is, relies, you know, pretty heavily on having the right amount of control over how consumers find out about events and then end up buying tickets, right? So naturally, those people are investing significant resources into partnerships, whether paid or organic, with with all of the frontier companies such that if, you know, everybody is chatting with chat GPT all the time and that becomes the place where concerts or events get, you know, bubbled up, like, that's a pretty important place for a few key people to, you know, have a lot of power, which is totally understandable and true. So naturally, I think like that's where a lot of the conversations have, have, you know, been circling around. My opinion is like for the average event promoter, there's not a whole lot that they can do to, um, influence that one way or another, right? If I own a music venue, it's like, how do I get my events to show up and chat GPT? My ticketing company is going to be the one that's going to install the shit on their website in order for it to happen, right? There, Banzentown is going to be the one that, that, you know, host that calendar and pushes it up just like to do to Spotify and whatever they've built amazing technology, right? Maybe my web developer or the agency that I work with who, who publishes my event calendar on my website, they're going to be the ones that marks up that information. In my opinion, like, those are to the obvious, but not that. super important conversations to be having as it relates to like how will AI change how live music venue operates right or a promoter operates. I think like we've seen it because we've had these like AI come to Jesus Moments over a weekend and then we come back on Monday and it's like everybody in the company we need to start using co-worker right now and everyone's like fuck Matt and Pat went on another vision trip and came up with another AI tool but we have to use but now like you know my self included so many team members like I exclusively work at a cloud co-work right I'm not drafting emails anymore it's just co-work and and I had said to you when we were prepping for this like maybe that is one of the things that we should have done at one of these awesome awesome conferences we were at is like let's sit down and start to install co-work on you know a talent buyers computer and like show them how to use it right let's figure out how we get it plugged into prism how we get it plugged into you know they're all of the tools that they use they're slack they're Google Drive their teams their Gmail right let's figure out how to get it access to you know they're 10 years of historical ticketing data and sales data whether it's through you know a great platform like like prism or whether we're you know just loading in massive massive lists of CSVs right and in my opinion like discovery is one thing very fan-centric right at the end of the day like you know you own a physical venue you're going to be booking talent that you know expires once the show's gone so people discovery buying tickets all of that will get solved by the people that care a lot about it because they're making a little bit of money along the way as a venue operator as a promoter like what are those tools that you should be investing in and training your team on in order to try to see some of the efficiencies of these tools that have now existed for like a year that so many people are using right who are going to be kind of the early adopters of that stuff so that's that's where I think like you know the the mcp server that you guys have built is so interesting because just like for you know for for years now it's been possible for you to dump in you know CSVs and CSVs and CSVs and ticketing data right and start to chat with it but takes a lot of time a lot of it's very expensive to do it that way it's very inefficient and now I mean and actually Matt I remember seeing how quickly you guys were able to spin this up on top of prism and offer the mcp server it is crazy the type of questions that people are able to get answers to now like what give me what do you think what do you think the craziest question is or what are some of the hot questions some of the most you like to use word curious what are some of the most curious questions that you've seen people ask about their data in prism well man the coolest mcp use case that I saw recently was like a contract checker where you know yeah again on on you know it's someone's using prism and they're doing like 500 or 600 events a year or whatever they they can they you know it's a lot of just okay you send the offer you get the contract back from the artist there's been an amazing amount of it and inefficiencies with that and you know some and you know sometimes it's like you have to like take the contract in and like look at prism and and like look and see okay did they get the deposit right did they get this or that right they can now feed I'm teaching my customers how to feed the contract to clawed it looks for the events in prism and it and it produces a spreadsheet of like here's 15 deal points in your in prism here's how the contract interpreted it did it get it right did it get wrong now similar to the high email like I tell my customers don't just take it at face value like double check it or whatever but it's it's an amazing value to kick off the work and like and like the red the obvious red flags are now uncovered oh they messed up the deposit they messed up that so like the the offer to contract process is amazingly inefficient and this is one way that like prism plus AI is helping kind of solve that problem like prism you have this like deterministic structured data and AI can like you know take in a contract immediately just boom you know cut cut a 30 minute task down to down to one minute um another super interesting use case and we're just getting in the process of launching this is like AI on top of on top of insights so and and if you're using whether you're using insights or are not are you whether you're using prism with insights you're just using insights as a standalone product um yeah for those that don't know insights are our data share that we launched where you know promoters and venues can subscribe to a data share and in exchange for putting all their data um and you know sharing all their data they can get um you know data back from the network and now we have built up um a larger data set than live nation has annually um it's about 80,000 events a year at this point in growing it'll be over a hundred thousand events um and and now and now you can kind of ask AI to kind of catch patterns and all that like what what genre you know should I be paying attention to in this market that I'm not paying attention to or even when you're you're about you're considering booking you know uh you know whatever I always use Mac Miller when you're considering booking Mac Miller and you're like um rest in peace and uh and you know the last you know 40 shows there's all this data inside of insights like you can see the gross ticket sales from the last 40 markets some of the markets are similar to you some of the markets are smaller to you smaller than you uh we have social media data that's in there now demographic data for the people that are attending and then you have an art then you have an agent who just said you know the the guarantee is 80,000 dollars in the ticket prices are 45 and you can ask the AI like hey here's the offer that I've built in prism here's all of the insights data um like you tell me what what like analyze it for me and like now all of our clients have like like a like a data now analyst that works for them and can produce a report well the Spotify data in Austin went up 30% and um and you know the Instagram fall like all the social media trends are heading the right direction and the hard ticket data you know like they're consistently selling out bigger venues and like all of that's living inside of an insights page but to have a plain English readout or maybe even like when you're walking down the street you ask clawed to like tell you it and you have this eloquent English man now tell well you know I'm not gonna fail to do an accent so those just get us they're all super interesting and I think very very very very early um and you know insights needs to get more and more data in there and and we need to train the models more and more on kind of interpreting the data more but um again it's it's split between yeah taking workflows that were once um really slow and tedious and speeding it up to like data analysis you know like what are the top 40 like look at all the insights data what are the top 50 bands that are no brainer for my venue that I'm not booking yet and you and I have talked about like throwing your high data into that mix um which would be even and I think that perked up a lot of people's ears at yeah let's let's repeat that is yeah yeah yeah because so it sounds like maybe just to summarize it sounds like you know there is now data available for um talent buyers to be able to um analyze make decisions maybe de-risk um better than ever before right and I think that was the big unlock with insights was um gathering up all of that data into one place um and making it accessible um and I think early and early results from that were crazy right you were starting to see people get burned on last show as you were starting to see them be able to be more competitive on offers because they knew that it was backed by you know the good good ability to sell through um and now with uh but it's still but it still took a lot of data analysis right and now with being able to take insights and connect it to AI like through quad or whatever um now you're able to it's it's so funny like it's just like simple questions that you know your users are users asked every single day right it's like how are my shows selling what shows should I be booking what shows shut out of these should I definitely not book right um but to go and answer those questions on your own feels like uh it takes a lot of work and it's really overwhelming when you even when you have data so I think it's sort of like um a pretty cool moment for the industry to be able to have access to whatever 80,000 100,000 shows worth of sales data um that traditionally wouldn't be available especially to independence right so I think there's something really cool and beautiful about that um yeah it's been a major it's been a major breakthrough for the folks that are using it which obviously like as you know the CEO of the company that provided the tool on like very feeling a lot of a lot of gratitude to the work sucks yeah you know yeah it's one of the things that I have a smile on my face about when I think about yeah what we've accomplished at prism and it's like it's not something that has like ever been done um or I shouldn't say ever you know Polestar did it 30 years ago um and god bless Polestar I love John I love Polestar um I mean they like you know in many ways they created a framework that like warmed up the whole industry to like hey like you know a rising tide raises all shipped that's like one of the biggest things that they accomplish and they and they had a business for 30 years that you know capitalized on on getting people warmed up to that idea um but it wasn't connected to real settlement reports it wasn't connected to ticketing integrations um and it was you know controlled by the way they got people to report was by um you know producing the lists of the top venues the top promoters the top agents so people say that on poster you you can see the potential of what something is but you don't you don't see the whole picture and like, you know, people ask us sometimes like, hey, can I just manually report to insights that were like, not right now, like, you know, like maybe at some point in the future, like we, we, we like let, you know, high integrity, situate, we can build systems around like manual report, but like our, our point is to build out, build out every single ticketing integration and like, just just automatically because we, we want to provide something different for the industry like we, like you, like bad data and bad data out with, um, with AI, right? Like, I think because insights is built on like you share everything, the good, the bad, the ugly, like we're not, it's an internal only tool for the industry. Like, there's not like journalists hanging out inside of insights. Like, like, you know, like, there isn't pull start, right? They're like just, you know, God bless journalists, but they're just waiting to be like, oh man, some show did terrible at x, y, z venue. That's not what insights is. It's like, it's a way to like actually predict it. Um, you know, where, where the lie, you know, how, how concerts are going to perform, um, at the end of the day, like that's the whole day because there's, there's people that are like making outsized bets every single day. Um, like, you know, you and I, we got, we got to talk about, um, the thing that got audible gasps in the room at, uh, at, at, at, at the Niva conference, which is, so if insights provides good market data, like multi-region, right, nationwide touring data, um, and that, that helps you get, and then, you know, Spotify provides a great sense of velocity and growth, one way or another, as it relates to listening, it streams, insert markets, you know, Instagram gives you, gives you some metric, whatever. Um, well, it's a good metrics to look at. One thing that, um, you and me have always chatted about, um, that we would love to do at some point is to be able to take, um, of any or promoters first party data, you know, all of their historical, historical ticket buyers, um, and be able to like underwrite or score a show based off of literally who in their, in their database is very likely to buy tickets, right? So, um, whether that's because they bought tickets to that artist in the past, or because they show like significantly high-fi affinity for some band that, that you hypothetically might book, right? So it's, um, so there's definitely some, some crazy opportunity there for folks that are using prism and using hive to be able to connect, you know, their individual and ticket buyer data to be able to say, okay, this is this Mac Miller show, it's pricier than I normally would want, except that I know out of, out of the 50,000 people in my database in hive, you know, this group of people is, you know, 80% likely to buy tickets. I've got a few folks that are on the fence, so it's significantly lets me, um, de-risk, de-risk for that buy, which is, uh, which is a pretty cool opportunity, and then, um, uh, buyers are able to, um, be able to like underwrite a show based off of, like a bottoms-up model, wouldn't that be so crazy? To be able to say like, okay, this isn't just like how it's selling based off of trends, but this is like literally, because I have a model of the people that I think are actually going to buy tickets, their name, you know, for some amount of the inventory that I'm going to sell. Yeah, and, and I think what I love about this, this future, and we're kind of teasing something that we're talking about doing at some point, what I love about this is it's like, okay, hey, you know, I, I, I, I, I, yeah, how many, what's all the important things we can feed, a brain about, you know, making a phenomenal decision on, on whether or not, a phenomenal prediction on whether or not a show would perform, like the inside network-wide data is vital, it's amazing, you know, the Spotify data, same thing, Instagram, whatever, all these different inputs are key, but, you know, you throw in the hive data, and then all of a sudden it's like, you, yeah, you, you can see that you have the people that are in your community also to support the buying decision, and the fact that people aren't utilizing that data to me, just it like, it puts a smile on my face, not because, you know, because of what's possible, you know. I think it, it, it reminds me a lot of how I felt when people were unable to do, we talked about it earlier on the podcast today, like, unable to do good genre targeting, right, or unable to do really get artist affinity targeting or marketing, because the data was there, ish, there just, there wasn't a good tool that allowed you to actually, you know, parse all of that data, streamline all of those workflows, right? And I think like what we're trying to circle around is like, there's a moment coming where that, there will be that unlock, right? There will be a bunch of unlocks, but it is really aided by the rise of, of, of these AI tools in that, like, they can parse lots and lots of data, they can make it really, really easy to connect products like hyphen prism, right? And I think, I think about it a lot the same. It's like, why hasn't anybody done this before? It's obviously a great idea. It sounds cool at least. It's like, because it's been fucking impossible, right? Because like, how many people, I can tell you, like, how many people literally have a clean database of all their ticket buyers for a decade, right? It's not that many people, right? And even if they did, then they'd have to connect it to something else, right? And it's like, that's, that's not the easiest thing to do, right? It's definitely possible, but it's not the easiest. So I think like, maybe to answer your, your good question, it's like, because it's really hard, and because like, if I'm a promoter, I'm not an IT, you know, I'm not an IT professional, I'm a promoter, dude. So like, you know, so I think like, it's a, it's fun that I think we can help connect those dots. You know, I think there's something really cool there. And then answering the question, too, when you combine like the hive and inside the data, which is again, like not, what one thing is, hey, this, this show has arrived on my desk. And should I do it or not? I mean, you know, that's, if only enough shows showed up, this is one thing that's kind of bothered me about the industry, is like, so many of these venues are relying on like the shows showing up on their desk, because, because they're planned by a tour and it makes a lot of sense, because, you know, the economics of a tour just makes sense for a band. So, but at the same time, like, there's just tons of opportunities for one-offs, regional, local, you know, somewhat regional, and like, you know, a good venue play, could like, cough a regional tour. And, and, and yeah, and if there was very good data that like, mitigated risk and, and, and, like, this could just be really game-changing and creating new concerts that wouldn't have happened. It wasn't for the data. And what venue couldn't use more shows, you know? Because I think the data is in there to say, hey, what's every single show that should be happening? That's, that's not, especially when you pair it with, but, you know, what, what, what's, what are all the interests of all the buyer data inside of my, my hive account, which is, you know, essentially the home of my venue community. It's very exciting. The prospects of it. I mean, it's super, it's super cool. So, what do you, so Matt, here's, here's a question for you. Sir, do you think robots are going to take over the world? No. So, so you're, you know, you know, you know, you're clients better than I do. All the customers that you get to serve, do you think that, you know, if you close your eyes in three years, is everybody going to be in co-work? Like, what, if you're, you know, you're a talent buyer, or you're on the other side, you're trying to route a whole tour. Like, what do you think? Where do you think those people are going to live? They're going to be just chatting with Claud, with the British Cloud as they walk down the street, or like, what's the, what is the future going to be there? I, I think that this, these are tools in the quiver and something like, like, is, it's co-work on a, like, replace, like, all of everything else? Is that, is that what your question is? Yeah. Are they going to use this? Maybe. I don't know. I think, yeah. I think, yeah. Is it, is it something that, that, everyone else is going to want to use as much as you and me do every single day? Should, should, we get this podcast sponsored by Anthropic and try to get some, some affiliate cut or what? The amount of people that I've trained at the topic, they, they should cut me in, but they're doing just fine without me. So that's, well, they haven't cut me in. But, and I'm partially joking there. But, yeah, you know, I think the way that AI works is, is it's, it's a, you know, probabilistic approach to fetching data, whereas, like, hive and prism with the exception of the AI tools is, like, you know, it's a deterministic, like, you know, if you're trying to sum all of your revenue across all of your shows in prison, like, we're not guessing on that number. Like, we run an algorithm that, like, adds up every, all the revenue on your shows, whereas, like, an LLM is just a language model is not meant to do that. And this is, and this is why, like, Claude can give you this unbelievable answer. Like, I was, Claude was helping me pick out, I have, I'm kind of a nerd about guitar pedals. I play electric cello and synthesizer. And it told me about these pedals that were not even on my radar. That was amazing. And in order to hook them up to one of my instruments, it, it made up products that weren't there. Like, and I was like, Claude, like, I'm not seeing this on the internet. And it's like, all right, I got to peel that back. Like, you know, I'm like, well, why? Like, so, you know, because there's no sign of that getting fixed. So I think like, what I say is you can trust AI. You just can't, you can't trust AI, but you can trust AI to be AI and you can learn how AI responds at things and you can trust that. So I, you know, I think AI is a, a tool in the quiver. You know, and, and it sits alongside of a lot of other things. And like, I'm using co-work more and more and more. And then I also find its limitations all the time. And like, you know, I, I, I've teed up these awesome automations. Like, when they got recommended clients, it's like, do a do a Monday morning report of all of your upcoming shows, grab the data from prison, push it out to Slack so your team just knows all the upcoming shows. We have our own version of that with the technology that's coming out. I just met with the person at my team who's owning that and it's down to a 25 minute task every day. Every week. I'm like, "Why isn't it zero?" It's like, "Well, it gets stuff wrong all the time and I want it to be right." It gets the date wrong and I've talked to about it getting the date wrong and I've drilled in the training. I think there's still a fundamental understanding of what these models are good for and what they're not and the answer is not zero and it's not everything. I think it will be like mobile or that's my prediction. Again, there's people out there that are predicted the end of the software engineer three years ago and I think it's still a pretty good time to be a software engineer. Still cranking. Maybe, "Oh, let me plug one more thing." I think because I'm picturing myself, maybe I'm a high-customer, a prison customer or maybe I'm curious or neither maybe we're one of the great partners that we work with. I think we always have tried to figure out, "Okay, where's the best place to point people? Where's the best place to go and learn?" You asked me this question earlier. One thing that we've spun up is a community. It's for event marketers, but welcome to the whole live event community, but it's for event marketers. It's called Backstage and you can check it out. We'll post a link maybe in the description here, but we have a library of clawed skills that you can go and use right now to help you plan out your marketing campaigns, to help you analyze your ticket sales data, to help you do that stuff. We have a Slack community, I think there's about 150 of the most talented marketers that I've ever gotten to work with. There they exchange who is using this new Facebook ads feature who has tried out this new version of clawed. Maybe that's something else that we can extend to the community mat is an open invite. We'll post a link to read a join. We don't shell our shit. If anyone tries to start selling their stuff, you included Mr. Ford. We'll kick you out of there, but it's for people in live events, whether you're on the talent side or you're buying a book in or you're an event marketer or you're managing a team of home. You can come in and ask questions, be curious, and it's everybody just trying to figure out this world world that we live in, especially as AI keeps shaping it. We can post a link to that, but that's been a really fun thing to see slowly evolve. We can get a few more clawed co-work users going in the next couple of weeks because it is such an unlock. Great plug there. I know you and I both got a wrap. What I want to end on is just a message of optimism that there's a future, a potential of live music industry future that I think is better because of all these tools. Especially when it's combined with the hard work of all the individual humans that are lifting this industry up and God willing, not only more shows are happening, but there can be more stability for families, people who have families in the space, and people who are doing the really, really hard work to make this whole thing an ongoing reality. It's not going to solve every problem. I don't want to leave on an unrealistic optimism message, but I do think it can solve a lot of problems to make more concerts happen. That's what we're here for, man. You definitely can't replace that feeling when you got the 100 or 100 thousand people waving their hands on the air screaming the same thing. I love that. I think it's a good thing for society if there's a really thriving live music ecosystem. It's a sign of an advanced society if you ask me and advance whatever peacefully evolved society. There you go. What a good, maybe that's a good note to end it on. What do you think, Mr. Ford? I think so. Yeah. Bad always a pleasure, man. Yeah. Good. We will do it again. Three hundred and sixty four days. Yes. Four days. And it'll all be different by then. Yeah. Yeah. Yeah. Yeah. Yeah.