Chapter 1: Introduction and Welcome 0:01 1 second All right, and we are live. 0:05 5 seconds So after 25 years in technology, I think we are entering one of the most exciting 0:13 13 seconds periods our industry has ever seen. And for decades, we built software to help 0:20 20 seconds attractions run their businesses. But now we’re moving towards something much bigger. technology that doesn’t just collect information, 0:29 29 seconds but now it understands what’s happening and now we know how it can help us make 0:37 37 seconds better decisions. And I’ve been saying this a lot lately. AI can build just about anything, but it does have to have 0:44 44 seconds context and it does have to actually understand and know what needs to be built. And that’s why I’m so excited 0:52 52 seconds about today’s conversation. And we’re talking about intelligent venues, hyperpersonalization, 0:58 58 seconds privacy, and how AI could completely change the way guests experience attractions. 1:09 1 minute, 9 seconds See you after the intro. [music] 1:19 1 minute, 19 seconds What the 1:23 1 minute, 23 seconds [music] 1:29 1 minute, 29 seconds hell? What the 1:32 1 minute, 32 seconds [music] 1:49 1 minute, 49 seconds All right, we are back. This is Live with Sandra. I want to welcome everybody who’s watching today. I’m Sandre Shannon, the CEO of Gatemaster Chapter 2: About Gatemaster Technology 1:56 1 minute, 56 seconds Technology. And before we jump in, a quick thank you to our sponsor for this show, Gatemaster. For more than 30 2:04 2 minutes, 4 seconds years, we’ve been building the technology for the attraction industry. 2:09 2 minutes, 9 seconds And today, our vision is so much bigger than just ticketing or point of sale. 2:13 2 minutes, 13 seconds We’re building towards becoming the operating system of the attractions, connecting ticketing, ecommerce, access 2:22 2 minutes, 22 seconds control, operations, and guest data into one platform that helps operators understand what’s happening across their businesses and so they can make better 2:31 2 minutes, 31 seconds decisions. I think that’s why today’s conversation is a perfect example of where the future is headed. Before we 2:39 2 minutes, 39 seconds get started, I want to say a quick hello to the Live with Sandre podcast listeners. And if you are watching on YouTube, make sure to like and 2:47 2 minutes, 47 seconds subscribe. Maybe even everybody share this episode because today’s guest is super [snorts] exciting and I want to Chapter 3: Meet Maris Ensing, Founder of MAD Systems 2:55 2 minutes, 55 seconds introduce you. So, let’s meet Mars. 3:01 3 minutes, 1 second Mars Ensing is the founder and chief architect of MAD Systems, an award-winning company 3:10 3 minutes, 10 seconds creating intelligent audiovisisual and interactive experiences for museums, themed entertainment, and visitor 3:18 3 minutes, 18 seconds destinations. He is an inventor, an author, and an industry innovator working at the intersection of AI, 3:26 3 minutes, 26 seconds personalizations, accessibility, and intelligent environments. 3:32 3 minutes, 32 seconds And when we asked what he couldn’t live without, I thought this was so sweet. He said his wife, who apparently does a 3:39 3 minutes, 39 seconds whole lot to keep him sane. And I’m very excited to welcome Maris to the show. 3:50 3 minutes, 50 seconds Good afternoon. Good afternoon. Welcome to the show. Thank you very much. 3:55 3 minutes, 55 seconds Hey, so you know, I wanted to start with the the the big question just so our audience audience can get to know you a 4:02 4 minutes, 2 seconds little bit better. And we we talked a little bit about this before the show, but tell us how you got started in this industry. Chapter 4: Maris’s Journey in Technology 4:12 4 minutes, 12 seconds That’s uh that’s a long time ago. Um at the time I was uh developing satellite 4:18 4 minutes, 18 seconds subsystems. Um I was at a facility in Germany where somebody let me into an 4:26 4 minutes, 26 seconds area where they said press the button on the side and I saw this amazing at the time tape slideshow that really got me interested in doing something similar. 4:37 4 minutes, 37 seconds Um and when an opportunity came up a couple of years later, I decided to jump into this industry um and started work 4:46 4 minutes, 46 seconds on it and that was uh that was a good move. I’ve enjoyed an awful lot of what I’ve done. 4:53 4 minutes, 53 seconds Yeah. And I think you said how long ago did you start your business? 4:57 4 minutes, 57 seconds I started I started a another business in 1991 and then I moved to the US in 5:04 5 minutes, 4 seconds round about 97 and started Matt Systems in on the 1st of January, 2nd of January I think of uh 1998. 5:13 5 minutes, 13 seconds Wow. 5:15 5 minutes, 15 seconds Amazing. I love that. I love these stories of how long First of all, congratulations because we know that 5:22 5 minutes, 22 seconds that is I mean to be in business for that amount of time. I mean I we started gate master started I was not there for 5:29 5 minutes, 29 seconds that the beginnings but started in 1994 and one of the things that I’ve learned over the years is how very difficult it 5:37 5 minutes, 37 seconds is for a business to stay in an industry and relevant and congratulations to you on doing that and I would love to really 5:45 5 minutes, 45 seconds talk about the big picture with you right jump right jump yeah so so staying relevant is a big deal and Chapter 5: The Evolution of Venue Technology 5:56 5 minutes, 56 seconds the good thing is that my R&D background and the background of my team um really has allowed us to do that. Um 6:04 6 minutes, 4 seconds they are all innovators, they are all um inventors and they’re tenacious. Once they get their teeth into something um 6:13 6 minutes, 13 seconds they want it to work. And so round about 2006, we decided when we were doing the Griffith Observatory AV systems here in 6:23 6 minutes, 23 seconds LA, we decided that we wanted to build a computer-based system. Um, and we did and we got the whole system up and 6:31 6 minutes, 31 seconds running as nothing but computers. The problem at the time was that it was still expensive for a lot of our 6:37 6 minutes, 37 seconds clients. And so what we did is that uh we had to go back for the intervening years to using different methods for a 6:46 6 minutes, 46 seconds lot of our clients. But round about 2017, some of our staff um decided to have another look at it. And we found 6:54 6 minutes, 54 seconds that computebased systems were viable in terms of costs. And so we started 7:01 7 minutes, 1 second developing um a system which is now called Quicksilver. Um and Quicksilver is uh all based on nonproprietary 7:10 7 minutes, 10 seconds hardware. So computers basically of all shapes and sizes. And it doesn’t matter 7:16 7 minutes, 16 seconds too much whether it’s a computer like this one um or bigger or smaller 7:23 7 minutes, 23 seconds provided it has enough compute to do the task. You can use compute in order to build a system. Nice thing with that is 7:30 7 minutes, 30 seconds that it’sworked um it’s intelligible to IT people. So a lot of venues no longer have AV staff to 7:38 7 minutes, 38 seconds be able to support these systems and it means that you have a computer-based system that isworked that is capable of 7:45 7 minutes, 45 seconds a whole lot more and that is what started the avalanche for us. Once that was up and running we realized that once 7:54 7 minutes, 54 seconds you have computers you can take on anything else video audio control interfaces to just about anything. Um 8:02 8 minutes, 2 seconds and then of course when uh AI started becoming available, there is no issue with starting to look at integrating AI 8:10 8 minutes, 10 seconds within these kind of systems because your architecture is already designed to take it. Chapter 6: Breaking the 10-15 Year Capital Cycle 8:16 8 minutes, 16 seconds Sure. Well, that leads me to one of my questions, you know, with regards to attraction operators and kind of breaking free from the traditional 10 to 8:24 8 minutes, 24 seconds 15year capital expenditure cycle with software and AI as it has evolved faster. 8:32 8 minutes, 32 seconds What is your thoughts on that? 8:34 8 minutes, 34 seconds So, one of the good things with using compute as your AV um system core um 8:41 8 minutes, 41 seconds your control core is that spares will be available for as long as Intel keeps manufacturing chips and computers um 8:50 8 minutes, 50 seconds which is going to be you know a a long time to come. So getting spares and being able to maintain and look after 8:58 8 minutes, 58 seconds these systems is going to be so much easier when you’re fully dependent on non-proprietary equipment. You can change screens, you can change 9:07 9 minutes, 7 seconds projectors, but if your core is compute, then things are only going to get better with time. And so we’re not only 9:15 9 minutes, 15 seconds benefiting from the fact that computers are getting faster and better, but anything that anybody comes up with that 9:22 9 minutes, 22 seconds computers can do becomes something you can integrate within these systems. And so it’s all about building the 9:29 9 minutes, 29 seconds foundation, right? Standard computing gives what we call AV++ the the 9:36 9 minutes, 36 seconds furtherance of AV into a new era capabilities um to run whatever becomes available 9:43 9 minutes, 43 seconds next. And that is what started happening when we started looking at recognition systems um looking at um you know all of 9:52 9 minutes, 52 seconds these different types of technologies that we’re now starting to see coming online. Um there’s a future for a system 9:59 9 minutes, 59 seconds that has capability to improve. And even as you’re going into time, um if 10:07 10 minutes, 7 seconds something fails, another computer is all that you’re going to require to make that to make that replacement. Monitors, projectors, anything like that will 10:16 10 minutes, 16 seconds always be available for a long time to come. And when the core is in a similar league, you no longer have that 10:24 10 minutes, 24 seconds traditional cycle. You will not be trying to find spares on eBay in 10 years time. 10:30 10 minutes, 30 seconds Sure. Yeah. That’s fascinating. Well, and I’m curious, just so the audience knows, you know, what does implementing? Chapter 7: Smart Venues vs. Intelligent Environments 10:36 10 minutes, 36 seconds So, what does this look like for a museum or an attraction or a family entertainment center? 10:43 10 minutes, 43 seconds Well, when you have standard computers running everything that it needs to do, it means that you have a practical path 10:49 10 minutes, 49 seconds to replacing computing and added capabilities. 10:54 10 minutes, 54 seconds and you really don’t need to do anything about your infrastructure because it’s standard computing architecture. Um, and 11:02 11 minutes, 2 seconds from a from an operator’s viewpoint, um, the fact that anyone with reasonable knowledge of it is capable of 11:11 11 minutes, 11 seconds understanding and maintaining these systems means that you don’t need anything really specialist, anyone with really specialist AV knowledge to be 11:20 11 minutes, 20 seconds able to deal with the bulk of what needs to happen with these systems there. So there’s a there’s a lot of benefits both in terms of longevity um maintenance 11:29 11 minutes, 29 seconds spares um and the people that need to look after it. 11:32 11 minutes, 32 seconds Sure. Now what’s the difference between a smart venue and a truly intelligent environment? 11:42 11 minutes, 42 seconds So, a a smart venue really talks about connected equipment and automated responses, but there’s a snag. 11:53 11 minutes, 53 seconds Sending a command does not mean that command was executed, right? So, a smart what is known as a smart system will automate and will make things happen. 12:05 12 minutes, 5 seconds But an intelligent system, an intelligent physical environment keeps the purpose active. So if we’re saying 12:13 12 minutes, 13 seconds in an intelligent venue that uh we’re designing it to have a speech intelligibility at this position of 6, 12:22 12 minutes, 22 seconds let’s say, or 7. Um the system, an intelligent system, doesn’t just put out 12:29 12 minutes, 29 seconds the audio or put out the video. It checks that the video is running and it checks that the audio is running and it 12:36 12 minutes, 36 seconds checks that speech in intelligibility is reached. And an intelligent system can also tune itself. So whether you have 12:44 12 minutes, 44 seconds one person in the auditorium or a thousand, the background audio level will go up, the background noise goes up 12:50 12 minutes, 50 seconds and and obviously um absorption of different frequencies will vary depending on how many people you have 12:59 12 minutes, 59 seconds and that kind of thing. So if you have a system that intelligently 13:06 13 minutes, 6 seconds listens and gets feedback, so you’re closing that loop, that system should be able to compensate for these additional 13:13 13 minutes, 13 seconds people, the additional noise and the uh changed absorption to make to make the 13:20 13 minutes, 20 seconds system work. So that is part of it. Um really from our perspective the real intelligence comes when a designer has 13:28 13 minutes, 28 seconds said this is what I want to achieve as the show and regardless of how that show changes with time equipment will change 13:38 13 minutes, 38 seconds speakers might change whatever that is the intent of the designer is maintained and if it cannot be maintained the 13:46 13 minutes, 46 seconds system tells you why not and where it’s going wrong. So that is what makes an intelligent system. Chapter 8: Personalizing Guest Experiences with AI 13:52 13 minutes, 52 seconds Yeah. Wow. That’s incredible. And I’m curious just so the audience can think about this, you know, why should these operators care about these intelligence 14:00 14 minutes systems? and and I mean obviously the adjustments that it’s making on its own and understanding what needs to improve 14:10 14 minutes, 10 seconds all of those things seem to make sense but you know from an operator’s perspective what do you think is the biggest thing that maybe they’re seeing 14:18 14 minutes, 18 seconds in terms of outcome or why they should care about this so there’s there’s a few things that we haven’t discussed yet um I believe that 14:26 14 minutes, 26 seconds one of them is to maintain show integrity and quality but the other big one is to actually tailor content to 14:33 14 minutes, 33 seconds individuals. Um, for too long, have we not supported different languages? Have we not supported different contents for 14:42 14 minutes, 42 seconds kids and adults? Um, have we not dealt with people who might be neurode divergent and need a different version 14:49 14 minutes, 49 seconds of the show? Um, and all of these kind of things are what AI, which can be built on a compute platform, is now 14:57 14 minutes, 57 seconds capable of supporting. So when you talk about intelligence from an operator’s perspective that really changes 15:06 15 minutes, 6 seconds um the way you start thinking about systems and obviously the other benefit is if you’re not always delivering the 15:13 15 minutes, 13 seconds same content but you can vary it. Uh think about return visits. One of the examples that I’m always giving to 15:20 15 minutes, 20 seconds people is that um I went around the Smithsonian in air in space with Dick Gordon, the commander of Apollo 12, and 15:28 15 minutes, 28 seconds I will never forget the time that he walked up to an Apollo door and he points at the Apollo door and he says, “Neil and I,” 15:38 15 minutes, 38 seconds I mean, that to me, you know, his buddy Neil Armstrong and him decided they didn’t like something and had it changed. And to me that really made it 15:47 15 minutes, 47 seconds clear that I wanted to give that kind of an experience to people who go to the air and space or go into a museum, 15:55 15 minutes, 55 seconds something personalized, something that they’re interested in. So imagine having an intelligent system that maybe uses 16:03 16 minutes, 3 seconds avatars at the moment. They might be on screens in the future. They might be on meta glasses, something similar. Sure. 16:10 16 minutes, 10 seconds Um, and the when you walk into the museum, you say, “Well, to today I want to go around with an astronaut.” And so 16:17 16 minutes, 17 seconds the astronaut takes you around and tells you the stories from an astronaut’s perspective. And it turns up wherever you are and speaks your language, knows 16:27 16 minutes, 27 seconds your age, so it speaks, it uses a vocabulary that’s suitable for your age, knows what you’re interested in. Are you interested in art or history or 16:35 16 minutes, 35 seconds engineering? And tailor the content to you. Now, next time you come a week later, a month later, you decide you 16:44 16 minutes, 44 seconds want to have a jet fighter pilot take you around and rather than talking about history, you want to talk about engineering, you’re going to get 16:53 16 minutes, 53 seconds completely different content without changing any of the hardware, without changing any of the core intelligence 16:59 16 minutes, 59 seconds that is within the system. But the the fact that people will come back because there’s a completely different 17:08 17 minutes, 8 seconds experience to be had is obviously another thing that should be of huge interest to operators. 17:14 17 minutes, 14 seconds Wow. This is amazing to me. And what’s amazing to me about what you’re saying is this is truly personalizing the experience, but also you know one of the 17:23 17 minutes, 23 seconds biggest things that I think so many operators have. being in ticketing. You know, this is a common question for us is how do we get the guests to return? 17:32 17 minutes, 32 seconds especially when dealing with cultural centers, museums, and um maybe places 17:39 17 minutes, 39 seconds that you might be in your in your backyard that you’re not you maybe don’t think you’re going to go there multiple times, you know, but what you’re talking 17:48 17 minutes, 48 seconds about is every time you show up to this attraction, you have the opportunity the opportunity to change the experience, 17:57 17 minutes, 57 seconds which is I mean that is mind-blowing to me that because this is the this is the time that we’re living in, right? And so 18:05 18 minutes, 5 seconds it really is. It It was mind-blowing to us when we started on it and what we’ve learned since means that we’re 18:13 18 minutes, 13 seconds completely comfortable implementing these kind of systems at this point. Uh we have one project going on at the moment where we have 20ome languages um 18:22 18 minutes, 22 seconds six age ranges. Um, none of this is really uh impossible when you have a compute system that you’re working with. 18:32 18 minutes, 32 seconds Um, and you’re not stuck with black boxes. It makes an enormous difference. Chapter 9: Accessibility Solutions: Language and Neurodivergence 18:37 18 minutes, 37 seconds Yes. And and and one of the audience members is saying that different content 18:45 18 minutes, 45 seconds without changing the hardware is brilliant and definitely the way of the future. And I agree. You know, when I 18:52 18 minutes, 52 seconds think, you know, you’ve given some great examples, but when I think about like the uh the Lincoln experience at Disney, 19:00 19 minutes it was the same experience every time, but something like this could change it. 19:05 19 minutes, 5 seconds I mean, granted, that was like my dad, one of my dad’s favorite things to always go and do. So, we went every single time, but I mean, after so many 19:13 19 minutes, 13 seconds visits, you have it memorized. And even as a kid, I used to kind of be like, I don’t want to do it again. But I can see where this can give give even those 19:22 19 minutes, 22 seconds kids. I mean you mentioned the accessibility as a uh aspect of what this can do for people like you said who 19:30 19 minutes, 30 seconds may be neurode divergent sensitive to sound fact that AI the AI that you’re using or the intelligence that you’re using can adjust for the the individual 19:40 19 minutes, 40 seconds is pretty fascinating to me and you know how could an attract you know and and and I just want to kind of bring this 19:47 19 minutes, 47 seconds idea home and that is like how the attraction with what you’re talking about can support not just languages and 19:55 19 minutes, 55 seconds interest but the accessibility needs if you don’t mind going into that a little bit more I think that’s absolutely absolutely so one of the things that we 20:04 20 minutes, 4 seconds have is called something Alice on board and other one is called Lori um these are tools whereby we found that there 20:13 20 minutes, 13 seconds for example when people have hearing aids um why not send the content to their phone so that they can listen to 20:21 20 minutes, 21 seconds it on their hearing aids it needs to be synchronized with the con with the main content. And we’ve cracked that part of 20:27 20 minutes, 27 seconds it. Um, and so once you’re delivering um content like audio to people’s to 20:34 20 minutes, 34 seconds people’s hearing aids, you realize, hey, I’ve got the screen. Why don’t I put sign language on there? So now we’re 20:41 20 minutes, 41 seconds delivering sign language. We’re putting the text on there. So subtitles on your phone. Why don’t we deliver it in their 20:48 20 minutes, 48 seconds language? No problem at all. We can now do that. We can now also deliver to braille readers and we can have somebody 20:56 20 minutes, 56 seconds that’s neurode divergent put on an ANR headset and basically say let’s stream a version that doesn’t have all the loud 21:05 21 minutes, 5 seconds noises and all the special effects but that has a more calm soundtrack so that they can still enjoy exactly the same performance as everybody else around 21:13 21 minutes, 13 seconds them but they have their own version that’s being delivered to them. And so we’ve spent um we’ve spent a lot of time 21:21 21 minutes, 21 seconds on developing these kind of um capabilities and we’re still in the middle. We have some new technology that 21:29 21 minutes, 29 seconds we’re currently um developing um both in my lab and in the lab next door funnily enough um where we’re looking at 21:37 21 minutes, 37 seconds multiple different approaches to take this even further. 21:41 21 minutes, 41 seconds It’s amazing. It’s amazing. I didn’t realize personally I didn’t realize that the technology was there already. Chapter 10: The Role of Context and Privacy in AI 21:49 21 minutes, 49 seconds Yeah. 21:50 21 minutes, 50 seconds And so this is fascinating. Now I want to kind of jump to uh you know about AI a little bit and its need for context. 21:57 21 minutes, 57 seconds And I know this is something that I’ve been thinking about a lot lately. In fact I pro if anybody’s watching they’ve probably seen some of the posts because I’m working these ideas out. In fact, 22:05 22 minutes, 5 seconds last night I was at a a Denver uh it it was a group of women uh CEOs who are 22:13 22 minutes, 13 seconds running technology companies and we we met to discuss AI and what’s 22:21 22 minutes, 21 seconds happening. So this is just something that I know so many people watching are thinking about not only can AI build 22:28 22 minutes, 28 seconds software and automate processes but you know understanding how it can 22:35 22 minutes, 35 seconds help attractions operate and I’m curious what your thoughts are on how how important is the real world operational 22:42 22 minutes, 42 seconds context when we’re building these intelligent environment very um so the first thing that we 22:52 22 minutes, 52 seconds learned was obviously that you need to be careful with AI because it can and a lot of people think it’s hallucinations. 23:00 23 minutes AI doesn’t hallucinate doesn’t hallucinate. It confabulates. Um I I won’t go into the detail right now, but 23:08 23 minutes, 8 seconds there’s a massive difference. And so what an AI will do when you push it into a corner, it will give you an answer 23:15 23 minutes, 15 seconds because you want an answer. And if it doesn’t know what answer to give you, it’ll make something up, right? Yes. 23:21 23 minutes, 21 seconds So you need to avoid getting it into that kind of a situation. So with Alice, which is our AI based delivery, media 23:30 23 minutes, 30 seconds delivery capability, we build what is called a body of knowledge. So we put everything in there, curated body of 23:37 23 minutes, 37 seconds knowledge. We put everything in there that either we or the client depending on who’s the work, who’s doing the work, um, wants the AI to know. and then you 23:46 23 minutes, 46 seconds put a fence around it and say anything that isn’t in there when you get asked a question you honestly answer that you do 23:54 23 minutes, 54 seconds not know the answer and maybe at the same time we’ll send a message to an attendant a dossent to say hey there’s 24:02 24 minutes, 2 seconds somebody here asking questions that I don’t have an answer to why don’t you go and have a chat with them so that opens a whole different set of capabilities 24:11 24 minutes, 11 seconds when you get further than that you realize that AI’s LLM that we’re normally using on a day-to-day basis are 24:21 24 minutes, 21 seconds book smart, but they’re not street smart. Yes. 24:24 24 minutes, 24 seconds And so this is what I read, what I’ve written two books about um some about something called the world model. And so 24:32 24 minutes, 32 seconds the world model is a way to use inputs, sensors from all over the place, cameras, audio, whatever that is, um so 24:42 24 minutes, 42 seconds [clears throat] that you know what is happening in a space. It’s kind of the next phase onwards from something that people might refer to as a digital twin. 24:52 24 minutes, 52 seconds Um, but it’s in real it’s in real time. 24:54 24 minutes, 54 seconds We need to know exactly what is happening. Hey, there’s a blind person crossing this square using our our um 25:01 25 minutes, 1 second blind guidance system. Do not let 500 people from the theater walk into that same square because they’re going to 25:07 25 minutes, 7 seconds have a problem. Um so street smart is very different from uh from books smart and once you have a system that is 25:16 25 minutes, 16 seconds street smart and you have guidelines so you govern the behavior of that system you now have something that can run 25:26 25 minutes, 26 seconds anything from a large museum to a theme park to a smart city because these capabilities are there and provided it 25:34 25 minutes, 34 seconds is governed intelligence in other words you separate out the I would like to do this, turn on a light, turn on the show, 25:42 25 minutes, 42 seconds whatever that is. And the intelligence saying this is the wrong time or yes, you’re allowed to do it. Or yes, you’re 25:49 25 minutes, 49 seconds allowed to do it, but you need to reduce the volume a bit because I’ve got something going on in the venue right next door. And we need to contain the 25:57 25 minutes, 57 seconds audio volume that you’re running in that space. Now, you have something that creates um an intelligent a truly 26:06 26 minutes, 6 seconds intelligent system. Yeah. Well, and I’m curious because this the the this space audiovisisual 26:13 26 minutes, 13 seconds is so foreign to me. I’m learning so much from you today and thank you for you know I I I I have to thank you for 26:21 26 minutes, 21 seconds being so willing to get up here and share the way that you are because it’s fascinating. But before this technology existed, you know, what did that look 26:30 26 minutes, 30 seconds like for people? Was it managed by people? Was it something that um and I I you know because I want to give a 26:38 26 minutes, 38 seconds picture for the audience of like where we came from. 26:43 26 minutes, 43 seconds So where we came from in the average location is that we have one set of content that has to be the same for 26:52 26 minutes, 52 seconds everybody. And so normally it’s designed for a 15year-old um in terms of vocabulary, in terms of 27:00 27 minutes content, in terms of everything that used to get delivered. The problem with that is that it isn’t suitable for younger kids because they don’t have the 27:09 27 minutes, 9 seconds vocabulary yet. So and it might be completely above their you it’ll just 27:16 27 minutes, 16 seconds fly beyond them. Um, and equally well, when you have an adult or and an expert especially, they’re going to walk in 27:24 27 minutes, 24 seconds there and learn exactly nothing. And so by by making it so that we can truly 27:30 27 minutes, 30 seconds tailor content for people and we can do that either by using a token like an RFID RFID device. We’re using phones a 27:39 27 minutes, 39 seconds lot where we just use the phone to identify yourself um using NFC or we’re using facial recognition [clears throat] 27:47 27 minutes, 47 seconds um in order to you walk up to a registration kiosk you say I want Spanish. Um, I’m eight years old. I like history rather than engineering or art. 27:59 27 minutes, 59 seconds Now, all the content when you go to one of the venues or an interactive or just 28:06 28 minutes, 6 seconds linear media can be tailored towards you or towards the group that you’re in, the school group that you’re in. And that 28:14 28 minutes, 14 seconds level of tailoring means that we just have a completely different world where by the time you’re 10, it is worth returning because you’re going to get 28:24 28 minutes, 24 seconds maybe similar content, but it’ll be at a different level and it always matches your level and it’s always designed to 28:31 28 minutes, 31 seconds be at your interest, what you’re interested in. And of course, when you have a school group and the teacher says they’re all eight and today we’re 28:39 28 minutes, 39 seconds learning about engineering, then you can fix that as as a as a fixed setup to do that. Um, you mentioned previously early Chapter 11: Real-Time Operational Decision Making 28:48 28 minutes, 48 seconds on. I just want to want to very briefly touch on that because we mentioned facial recognition, we mentioned 28:55 28 minutes, 55 seconds listening and we mentioned cameras. Um, one of the things that we have done to make sure that privacy is maintained. 29:03 29 minutes, 3 seconds The first thing is that on our facial recognition systems, we have no internet link. So we do not send it to Microsoft 29:10 29 minutes, 10 seconds and AWS and to everybody for them to do the recognition. We do it in-house. Um the average recognition time is about half a second. 29:20 29 minutes, 20 seconds cameras for other things like um when we have to track motion or something. We have cameras that are specifically 29:28 29 minutes, 28 seconds developed um so that they do not send um visual information across the network. 29:35 29 minutes, 35 seconds We call them a camera that cannot see. 29:38 29 minutes, 38 seconds And what it means is that we’re sending information like XYZ coordinates, but we don’t send pictures. And similarly, we have a microphone that cannot listen. 29:49 29 minutes, 49 seconds The data rate out of the device is so low that we cannot transmit audio, but we can transmit enough spectral data to 29:58 29 minutes, 58 seconds use it for masking and for other purposes that we want to be able to use it for. So privacy really matters in these kind of systems. 30:07 30 minutes, 7 seconds Yeah. No, I appreciate you you touching on on the privacy of of the guests that are experiencing this and how that works. And I and I actually want to kind 30:15 30 minutes, 15 seconds of go back to that world model that you mentioned. You know, I’m I’m curious, you know, how does 30:22 30 minutes, 22 seconds that technology or or this technology understand what’s happening in the physical environments? you touched a little bit about that with regards to 30:30 30 minutes, 30 seconds the privacy, but how does that help us make or how is it making better decisions if that’s something you could share? 30:38 30 minutes, 38 seconds Yeah. Yeah. Well, of course. Um, I can give you two examples. Imagine you’re a chaperone and you’re taking 10 kids into 30:46 30 minutes, 46 seconds a theme park. Um, and imagine you basically tell the system, I’m the chaperone, 10 kids. Here is where they 30:53 30 minutes, 53 seconds are. They’ve identified them all. Now, wouldn’t it be nice if one of the kids went astray, as they will, because we 31:01 31 minutes, 1 second were all kids and we were all probably um on that end of uh on that end of troublemakers. 31:08 31 minutes, 8 seconds Wouldn’t it be nice if you get the message on your phone that says, “Hey, one of your kids, Johnny, um is further 31:16 31 minutes, 16 seconds than 35 yards away from your group. I’m seeing him here.” and then gives you progressive wayfinding information to get you to pick that missing kid up. 31:27 31 minutes, 27 seconds Right? There’s one there’s one thing. 31:29 31 minutes, 29 seconds The other thing is supposing you go to um a theme park and you say um I want to see this attraction and that attraction 31:37 31 minutes, 37 seconds and that attraction. They’re top of my list. And the system says, “Well, I’ I’ve got long cues here and long cues here, but there it’s relatively quiet. 31:46 31 minutes, 46 seconds Why not steer people to the quiet part and then help them find those locations and once they’ve been there and done 31:54 31 minutes, 54 seconds that, get them to the next place and make sure that you do uh load balancing on all of your different rides by 32:01 32 minutes, 1 second personalizing the route that it suggests that you take. So now we can provide a level of 32:10 32 minutes, 10 seconds um of support for for visitors making sure that their Q lines are minimal um 32:17 32 minutes, 17 seconds as good as we can get them um and that kind of thing. Um equally well progressive way finding. So um mom and 32:26 32 minutes, 26 seconds the daughter want to go and see princesses, dad and the boy wants to go and do a ride somewhere. How did they find each other afterwards? Well, we 32:34 32 minutes, 34 seconds might be able to use phone, but with these systems, we can use progressive way finding and get them get them back together again. Those are just some 32:43 32 minutes, 43 seconds minor examples of things that we can do with these kind of systems in the future. And it helps enormously by um 32:51 32 minutes, 51 seconds improving the way things work. Now, take it outside of this context. Imagine you’re on a cruise ship. You walk into a 32:58 32 minutes, 58 seconds bar where there’s a bartender that has never met you who says, “You know, Sandre, your team is losing really badly. I’m sorry to tell you, but it’s 33:07 33 minutes, 7 seconds not going well. By the way, would you like your normal your usual drink?” And so, these are the kind of things that we 33:14 33 minutes, 14 seconds can now start doing by by using a mix of anonymous recognition mostly. I used 33:21 33 minutes, 21 seconds your name in this case, but I don’t need your name. I can just say, “Hey, whoever you are, your your your team is losing because you told me what your favorite 33:29 33 minutes, 29 seconds team is and you’ve told me what your favorite drink is.” Yeah. 33:33 33 minutes, 33 seconds So, th this kind of customer service is now within within grasp. It It’s all there. The systems are there to do it. 33:42 33 minutes, 42 seconds The capability is there to do it. 33:44 33 minutes, 44 seconds We have a lot of it running here in the lab in our Wonderland demo space. Chapter 12: Maintaining the Human in the Loop 33:49 33 minutes, 49 seconds Yeah. I mean, I think what’s going to happen here in the next few years is just I mean, we’re collapsing time in a at a 33:58 33 minutes, 58 seconds rate with AI that I I don’t I don’t think people are quite prepared for. 34:02 34 minutes, 2 seconds This is fascinating. And and and one of the things that I I I’m I have a question for you with regards to all of 34:09 34 minutes, 9 seconds this, you know, how do we keep a human in the loop in these conversations? What do you see that looking like? 34:16 34 minutes, 16 seconds So it really depends on to what extent an operator wants a human in the loop. I believe there should always be a human 34:25 34 minutes, 25 seconds in the loop. I gave you one example earlier on where you’re walking around in a museum. You ask a question, the AI 34:33 34 minutes, 33 seconds doesn’t know the answer to it, calls a human in to say, “Hey, you know, do you know you you have more information than 34:40 34 minutes, 40 seconds I have possibly um go and go and answer that question. go and and engage these people and work with them. That is one 34:48 34 minutes, 48 seconds aspect. The other aspect is obviously operational. Um do we what what happens when something doesn’t quite go right? 34:57 34 minutes, 57 seconds Well, rather than letting people guess what is going right and what is going wrong? If you have the system already 35:05 35 minutes, 5 seconds being able to tell the operator um when something happens and there is a difference between the expectations and 35:13 35 minutes, 13 seconds the delivered content, the delivered experience, whatever that is, uh you have a much better way of concentrating 35:22 35 minutes, 22 seconds um the human effort that needs to be part of this and balancing that human 35:29 35 minutes, 29 seconds versus AI um capability to make sure that you can provide that um that that interface, that assistance. And I think 35:38 35 minutes, 38 seconds also from a customer service um perspective, if you walk up to somebody because you’re angry about something and 35:47 35 minutes, 47 seconds you can say to the and and you walk up to a customer service representative who already knows that you’re annoyed about 35:54 35 minutes, 54 seconds something because something went wrong and you’ve already spoken to somebody about it earlier early on. And so when 36:01 36 minutes, 1 second you walk up to them, they’re, you know, I’m sorry this happened again. We really need to be able to help you out with this because they already know what 36:09 36 minutes, 9 seconds happened. You can diffuse situations so much easier and so much better. Think about when you’re walking on an airport 36:16 36 minutes, 16 seconds and your flight has been cancelled and you’re in a line of 30 people trying to get either a hotel on or another flight 36:23 36 minutes, 23 seconds and you walk up and the agent says, “Um, Mr. Shannon, I’ve got a solution for you. I caught a flight, you need to walk 36:31 36 minutes, 31 seconds over there, 15 minutes, you’re on your way. Um, you know, all this kind of direct interfacing which is now 36:38 36 minutes, 38 seconds possible. Um, and let’s face it, you walk onto an airport, they already know who you are with the touchless um, 36:46 36 minutes, 46 seconds capabilities that several airports now have. 36:50 36 minutes, 50 seconds This is not this is not an issue anymore in certain environment when it comes to privacy. 36:57 36 minutes, 57 seconds In some areas, it is. And that’s where we can maintain privacy, but we can always increase the and improve on 37:05 37 minutes, 5 seconds customer service and on delivering experiences to people. Chapter 13: Human Creativity and Technology’s Future 37:11 37 minutes, 11 seconds Yeah, you’ve blown my mind [laughter] with this conversation, but you know, I um because I I want to get into like the 37:19 37 minutes, 19 seconds part of AI and human creativity, but you just reminded me of something in this conversation because you’re talking about the guest experience and the 37:27 37 minutes, 27 seconds customer service aspect, which is always something that comes up in conversation even for us in in terms of you know, as 37:34 37 minutes, 34 seconds we are building out support using AI to build out support for the workforce and one of the things and it’s so 37:42 37 minutes, 42 seconds interesting because last night again we were I was in a room with aeros you know people who are designing drone technology and aerospace technology and 37:50 37 minutes, 50 seconds fascinating things and as we’re talking about how AI is being used in all of our various businesses 37:58 37 minutes, 58 seconds part of that conversation kept coming back to the fact that the human side of what we’re doing is becoming even more 38:05 38 minutes, 5 seconds important so you know you look at you know I I I’ll be the first to admit I have an agent that works alongside my 38:14 38 minutes, 14 seconds in-house counsel and myself and what I have found is that it has allowed us to have much more thoughtful conversations 38:22 38 minutes, 22 seconds about why we’re agreeing to the things that we’re agreeing to or why we don’t want to agree to that. And and and I 38:30 38 minutes, 30 seconds know and I’m not encouraging anybody to go out there and just use AI on contract negotiations or any or anything of that nature, but what I do believe is that it 38:39 38 minutes, 39 seconds brought a at least in my experience, it’s been bringing a lot of that humanity back into the conversations because the busy part of the work has done. 38:49 38 minutes, 49 seconds Yes. Right. 38:51 38 minutes, 51 seconds Yes. The part of the work that none of us like doing. Um to some extent I think that is true. I think the other thing 38:58 38 minutes, 58 seconds that in a on a day-to-day basis what I find is that expectations of things happening more or less instantaneously 39:07 39 minutes, 7 seconds um is obviously increasing pressure on humans um a lot because uh people expect an answer now 39:16 39 minutes, 16 seconds um you know I remember sending the first fax um when we were doing a project uh on a um a satellite called Rosat and we 39:25 39 minutes, 25 seconds managed to send the facts to Germany and then have a phone call where they already saw the sketch that we had just, 39:33 39 minutes, 33 seconds you know, put together to be able to talk about this interface problem that we were having. Um, now we’re at the point and and that was already a 39:42 39 minutes, 42 seconds reduction from being able to send a letter to people that that we had to wait for and then do the phone call. And 39:48 39 minutes, 48 seconds so we’ve got to the point now where expectations have obviously risen significantly and AI is part of that. Is 39:58 39 minutes, 58 seconds that a good thing? I don’t know. But it’s one of those things that is pretty well unavoidable. You know, Pandora’s box is open. 40:05 40 minutes, 5 seconds And I think that the important part of using AI, and that’s what I’m I’m really proud of what my team is doing. You can 40:13 40 minutes, 13 seconds use it for good and you can use it for bad. It’s like splitting the atom. um there are two ways to go. And so I 40:21 40 minutes, 21 seconds believe that we are working really hard to try and find positive uses that actually help people help people enjoy 40:30 40 minutes, 30 seconds things more, help people um not have to worry about certain aspects like losing kids maybe um about using their time 40:40 40 minutes, 40 seconds better. You know, I I’m thinking about people who save up for a lifetime, for a one-time visit to a theme park. Um, and 40:49 40 minutes, 49 seconds that’s all they’re ever going to be able to do. You know, they’ve saved for this for maybe 10 years, maybe 15 years. Why not make sure that we give them the best 40:57 40 minutes, 57 seconds possible experience that they can have, right? Yes. And if we can use AI to reduce lines for them, if we can help 41:05 41 minutes, 5 seconds them getting from one place to another without getting lost because they’re only there one time. So they haven’t got the benefit of some of us who who are in 41:13 41 minutes, 13 seconds and out of theme parks all the time and know exactly what is where. Um these are the kind of benefits that we can offer 41:20 41 minutes, 20 seconds people. Um, and I think the other thing is also that instead of, oh, we have we’re having to go to a museum tomorrow, 41:28 41 minutes, 28 seconds if we could get kids to say, I’m really looking forward to this because there’s there’s, you know, they’re going to talk 41:36 41 minutes, 36 seconds to me about things that I that I think are fun. And being able to do things like that, I truly believe is of huge 41:44 41 minutes, 44 seconds benefit when we’re starting to use AI in a very positive way. um and in a safe way by making sure that governance is 41:53 41 minutes, 53 seconds always the first thing that we’re looking at, control of the AI is always the first thing that we’re looking at and we don’t do things like pushing it 42:02 42 minutes, 2 seconds into a corner and get confabulation related problems um or wrong decisions. 42:07 42 minutes, 7 seconds Um there has to be a governance in place and this is what world model talks about extensively 42:14 42 minutes, 14 seconds where we have a constitution and we have a set of rules that might be just human rules. They might be the ten 42:22 42 minutes, 22 seconds commandments for starters, right? And then you work down from that and you add rules that that specific venue has. um 42:30 42 minutes, 30 seconds so that you never end up in a situation where an AI can take a decision that is detrimental detrimental to the humans 42:38 42 minutes, 38 seconds that it’s intended to serve and those are the kind of things that make me proud of the work that my team is doing here. 42:44 42 minutes, 44 seconds Yeah. No, I agree. You know, it was very interesting again last night I had the I we I was in this room. I just sometimes I can’t even I can’t even believe 42:53 42 minutes, 53 seconds sometimes I end up in these rooms I can’t even believe I’m in. Right. And I’m listening and one of the women in the room who deals with artificial 43:02 43 minutes, 2 seconds intelligence for the government, she says the only thing that can keep up with AI right now is AI, right? And to your 43:11 43 minutes, 11 seconds point, she also said we have to train enough of these AI to be good in order 43:18 43 minutes, 18 seconds to deal with the bad. And so I see where that pride comes from that you have because it’s it is important and it is 43:27 43 minutes, 27 seconds uh it’s something that I think the world needs to understand that there are people who are training this to do incredible things like make these 43:35 43 minutes, 35 seconds experiences so much better for people with accessibility needs and and and you know I I on LinkedIn I 43:44 43 minutes, 44 seconds often see conversations about how people are upset that jobs might be lost But at the same time, AI has now created 43:52 43 minutes, 52 seconds opportunities for people who maybe haven’t been able to work because they were missing limbs or the ability to 43:59 43 minutes, 59 seconds speak in ways and now they can communicate in ways they never have been able to before. And this has created 44:06 44 minutes, 6 seconds opportunity for those people. And that’s the good that you’re talking about. And this is the good that you’re building and designing for our attractions. And 44:13 44 minutes, 13 seconds just incredible. proud to know you right now. It’s it is it is exactly that, isn’t it? 44:21 44 minutes, 21 seconds And the fact that you can um that you can look forward and plan and look at, 44:29 44 minutes, 29 seconds you know, how can we use this in a positive manner. At the end of the day, as I’ve said earlier on, Pandora’s box is open. AI is here to stay. We’re not going to make it go away anytime soon. 44:41 44 minutes, 41 seconds And so let’s do the very best we can to get a positive result from it. At the end of the day, typewriters were going 44:50 44 minutes, 50 seconds to lose a lot of people who wrote their jobs. Um then computers came in and everybody was in a panic because 44:57 44 minutes, 57 seconds everybody was going to be no longer necessary. The reality is that there is a reconfiguration of working life, working society and there is a shift. 45:10 45 minutes, 10 seconds There’s no doubt of that at all. But none of these previous incredible changes that we’ve seen over the years 45:17 45 minutes, 17 seconds have resulted in people being um people no longer having work. The work changes. That is for sure. 45:25 45 minutes, 25 seconds Um but it doesn’t mean that that work goes away, that opportunities go away. 45:31 45 minutes, 31 seconds Um, and from our perspective being able to use it in the way that we’re using it, it just uh, you know, it it it kind 45:40 45 minutes, 40 seconds of supercharges the work that we enjoy most anyway, which is to provide good experiences for people. Yes. 45:47 45 minutes, 47 seconds And so from our perspective, um, it’s an incredible tool. Yeah, it is an incredible tool and thank you for being Chapter 14: How to Connect with MAD Systems 45:55 45 minutes, 55 seconds one of those pioneers that is doing great work and and bringing so much good into the world and into attractions um using this type of technology. And 46:03 46 minutes, 3 seconds there are so many more questions that I have for you, but we’re bumping up against time. So before I before we wrap this up though, I think it’s important 46:11 46 minutes, 11 seconds that our audience knows how do they connect with you and Mad Systems and learn more about world model and we 46:18 46 minutes, 18 seconds didn’t touch on Hatter. So, I think that’s something that we we we’re going to have to come back to on the next time because I I would love to invite you back um to continue this conversation. 46:28 46 minutes, 28 seconds But, you know, what is the best way for our audience to get in touch with you? 46:31 46 minutes, 31 seconds So, the easiest way is to look at our website madsystems.com. 46:36 46 minutes, 36 seconds Um if you want to know more about world model, look at worldmodel.global and you’ll find um all of the 46:44 46 minutes, 44 seconds information about world model that is that’s currently available. The books are available on Amazon. Um, one of them 46:51 46 minutes, 51 seconds is called The World Model and the other one is about hyperpersonalization. 46:55 46 minutes, 55 seconds Um, if you put my name into Amazon, you’ll find both. Um, and we’re easy to find. Um, we’re in the city of Orange. 47:04 47 minutes, 4 seconds Um, and, uh, mattsystems.com will get you our address and phone numbers and emails and everything else. So, 47:13 47 minutes, 13 seconds definitely get in touch if this is something you’re interested in. 47:16 47 minutes, 16 seconds Yeah. Thank you so much. No. Will you be at IAPA? 47:19 47 minutes, 19 seconds Most certainly. Yes, we’re on the we’re on the tea booth where we have one of the meeting spaces. Um and 47:28 47 minutes, 28 seconds mathsystems.com/book gives you an opportunity to book time with the team. 47:36 47 minutes, 36 seconds Very nice. Well, thank you again for being here and I really do love conversations like this. It’s it’s it’s 47:43 47 minutes, 43 seconds I I think I think we need to be having more conversations like this too so that people can understand where the technology is going and how much good is 47:52 47 minutes, 52 seconds happening from what’s being done and you know and and not just because I love software for software’s sake but I think 47:59 47 minutes, 59 seconds where we’re going with regards to technology is it’s an amazing time to be alive. 48:06 48 minutes, 6 seconds Absolutely it is. It’s all about compute and software and that really is where the where the world is uh is heading very rapidly. 48:15 48 minutes, 15 seconds Well, thank you so much for being here. 48:17 48 minutes, 17 seconds You’re very welcome. Thank you for having me. Yeah. And I’ll see you at IAPA. Most certainly will. Take care. 48:24 48 minutes, 24 seconds Bye-bye. Chapter 15: Closing Thoughts and Future Outlook 48:28 48 minutes, 28 seconds Well, like I said, I think it is a very exciting period for the attraction industry. It’s I mean I think about all 48:37 48 minutes, 37 seconds the years that I’ve been in this industry and I am most excited about this time. I’m most excited about this 48:45 48 minutes, 45 seconds era that we’re in right now and where we’re moving to and how it’s so much bigger. I mean connected intelligence that’s just blowing my mind. I mean 48:53 48 minutes, 53 seconds that’s a different part of the industry that I I knew that we were making advancements in but I didn’t understand in the full scope until Maris explained 49:01 49 minutes, 1 second this to us today. and and and again, thank you for being here and showing us what’s possible and how much good is 49:08 49 minutes, 8 seconds coming from these programs and systems that are being developed. So before I let you guys go, I just want to make 49:16 49 minutes, 16 seconds sure that you understand that this conversation is very much aligned with what we’re talking about at Gatemaster. 49:22 49 minutes, 22 seconds And our vision is to become the operating system of attractions. 49:25 49 minutes, 25 seconds Connecting the systems of operations, the data that operators depend on that they can make better decision about the 49:34 49 minutes, 34 seconds revenue parts of their businesses and how they’re making money and how people are um and how attractions are creating 49:42 49 minutes, 42 seconds better experiences because I don’t think the future belongs to whoever builds the most features. I think it’s starting to belong to companies that understand the problems that they’re trying to solve. 49:52 49 minutes, 52 seconds And I think today’s conversation really enforces that and it enforces something that I really believe deeply and that is that technology should expand human 50:01 50 minutes, 1 second creativity and not replace it and it should give designers more possibilities, operators more flexibility and guest more meaningful experiences. 50:11 50 minutes, 11 seconds I want to thank everyone who joined us today who is listening. Thank you to everybody who’s listening on the podcast. If you are a podcast listener, 50:19 50 minutes, 19 seconds we appreciate you taking the time to download this episode and to listen to us today. And if you are watching live and you enjoyed today’s conversation, 50:28 50 minutes, 28 seconds please share this with a friend. And if you are on YouTube, shout out to everybody on YouTube. 50:36 50 minutes, 36 seconds Please subscribe to our YouTube channel, share this episode, and uh let us know what you think in the comments. 50:45 50 minutes, 45 seconds I am Sandra and this is live. I’ll see you guys next time.