AI in Tourism: How It's Changing Travelers and How It's Changing Our Work

So, who here has been to a conference where you're talking about AI? Right? The trick. With being somebody who presents at conferences about AI these days, is that everybody is like it is a whole spectrum of familiarity with AI, and it's hard to know where to pitch the the conversation because oh you know are people still using ChatGPT you know like it's Google right where you put in a query and oh good I get a response I copy and paste that somewhere else are you using it like the you.

You know, fancy tool you can use to rewrite your emails so you don't sound so bitchy, right? Like yes, and I still use AI in that way. But there's then a whole other range of ways to use AI. And so, as somebody who's a little more advanced with it, it's hard to sit through a presentation where I don't feel like the more advanced use cases are talked about. So this is going to do a bit of both, and we're going to talk about the front door and the back office. So we were just having a conversation about, oh, God, is AI going to kill all of us? Are the data centers going to destroy the environment? Like, there's certainly a lot of concerns about AI out there, and I'm sure there's some people in the audience who are like, I don't really feel great about using it, which that is fine, right? Like, I am not here to bang the door that you absolutely have to start using AI.

But at a bare minimum, we need to understand that our travelers are using AI in order to discover what destinations we are going to visit and what we are going to do when we get there. And so we're going to talk about both sides of this equation. How can we prepare for the changes in the traveler journey? That's the front door. And then how can we prepare, like at work, to take advantage of what I consider to be an incredible tool for getting more done. Right and for providing more value to our stakeholders. So we'll talk about both today. And to get started, I would just love to know where are you on that spectrum. So who here has paid access to an AI tool at work? That's actually that's pretty good. Who's using ChatGPT primarily? Do I have any Claude fans? Yes.

Who is building stuff with Claude? Who is making apps and creating websites? And okay, great. So we'll have some. And who, if you don't mind sharing, feels like you're more on the beginner side, or maybe more of a, more of a doubter, more of a you. You don't feel great about using this technology. Perfect. We will have a little something for everyone. There's going to be a bit of time for Q and A at the end, so if. I'm not really going into danger or the environment in this particular presentation. If that's something you want to talk about, we can go into it at the Q&A. And then to get started, it's probably a good moment to let you know that Claude created this presentation. Like obviously, with my help, I didn't say make it and then walk away. So the images. This isn't a real picture of me as a fighter fighter pilot.

The images you see in the presentation are all AI generated, and I will say, as a tool for making presentations, AI makes it so easy to to build this stuff out. I find it to be a great help in terms of putting together the arc of the story I want to tell in a presentation, and then actually doing the slide design. If you're not familiar with Brand USA, We are the destination marketing organization for the United States, so we're focused on bringing international inbound travelers to the U.S. And then, in my role as Chief AI Officer at Brand USA, I'm trying to both build up our staff and our ability to use AI to do our jobs. Shout out, Danny! You are a terrific advocate for AI use at Brand USA, so thank you for that.

I want to increase the capabilities of the industry because, like we're competing against every other country in the world for that international inbound traveler, and we're not going to be able to spend our way out of it. Right, we're not going to outspend the Middle East when it comes to attracting those visitors, but we can operate smarter and as efficiently as possible if we are all more familiar with how to use AI. And then the third piece of my work is using AI to make the United States more discoverable and more bookable through AI systems. And so that's the piece that we're going to start with, the front door. Yeah, it's a fake octopus. It's not real, guys. So how can we use AI to help bring in visitors? Maybe. I don't know why that isn't working.

Oh, because there's a little error message on the computer. There we go. So we know travelers are using AI to book. This isn't even a new survey. This is from last year, but I use it because it's very global, and I want to understand how people internationally are using AI. So as of a year ago, Booking.com did a huge study of 38,000 people that from around the world. 62% of them at the time were using AI in some fashion as part of their travel planning process, and I'm confident that that number is higher today. And we know that people are planning their trips in a chat window. They are going to tools like ChatGPT and Claude in order to do their planning. We know that if you go to Google, those AI overviews are answering your question directly in Google, so that.

They're not necessarily those travelers aren't always making their way to our websites, which is really annoying, right? We also don't know when Google surfaces information from our websites in these AI overviews very reliably, so we don't know what kind of content we're even supposed to be creating, because we can't rely on which pieces of content are getting the most traffic on our website to make more content like that. We know that ChatGPT is able to take external. They have a program called Apps that brings external information into your conversation. So this was probably eight months ago now that they launched this, and so this was a partnership with both Booking and Expedia, where inside of a chat you could get real-time information on hotel prices or on flight prices.

And then, if you're not familiar with this, this the new part, Jay. This Little Pickle is the name of my Muse agent because I have another agent that I call Mr. Pickles, so I wanted them to be related but not confuse them. So this a meta product that you can download the app on your phone. It's the most downloaded app in the App Store, I think, for the last two weeks since this came out. You can connect it to your email and to other you know apps on your phone. I would not use this for work unless this is something your IT team is installing for you. And I will say it's a little risky probably for personal use. So I'm you know caveating how much information do you want to give a company like Meta. You also want to go into the settings and turn off the ability for them to train their models on your data.

So obviously things like your personal email, we want to keep that personal. But it's sort of insane what this able to do. This an app on your phone where the agent will open up its own computer in the cloud to do work for you. So in this example that I have the screenshot for the screen video recording of. I'm I'm coming into Memphis yesterday in order to get to Oxford, and I don't want to rent a car. What are my options? And it went and found a car service that would do you know do those transfers that would take me here, and it looked at the first one, and the first one blocks spots on their website. So little pickle came back to me and said, I can't book that car service for you. I'm going to find another one.

And found the next one that you send a message on the website to the car service company, and they texted back in that message, and this agent was able to act on my behalf to book the car for me, and I was on a flight home from Amsterdam, so I'm over you know the Atlantic Ocean. I ended up just going with Uber because then this you know the woman who runs this company. You know, oh, I need to talk to you. Can you call me? I'm like, well, I can't call you right now. I okay, well, call me as soon as you land. And then she called me like three times to say, what do you mean you're only staying for one day? And I'm like, that's going to be so much conversation, an hour long. Like I could already tell she was a talker. And you know, thinking back to our presentation from this morning, sometimes I just need the quiet downtime. So I went the Uber route.

And skip the southern hospitality on that flight or on that that trip to Oxford, but this is an example of where AI is going, the power that it is going to have to help individuals start planning their travel. What we're discovering here is that discovery layer is being rebuilt in real time, but that's fine because that creates an opportunity for tourism. So, if you're trying to become more discoverable in an AI system, if you want ChatGPT to suggest your destination or your tourism business, think about content for that website that only you could have written. So, something with a very specific, unique point of view. You want it to be authoritative and you want it to be useful. So, AI isn't going to reward you with traffic if you're.

Giving a list of top ten things to do in a destination, right? Because hundreds of people have already created that kind of content. You don't need the frequently asked questions. What you need are the infrequently asked questions. What are the little details that the AI can't find anywhere else on the internet, but could pull from your website about your business or about your destination. And then you want to make sure on the back end that you have structured content so that you are making it clear to AI, hey, my site is accessible to bots, and this the information that you are looking for. This the type of website that I am. On pages where you have content, you want to have information about the person who wrote that content. And I know, you know, frequently on a destination website, it's just kind of the content is written by a staff member.

But what AI is really looking for is, oh, this written by this person with this unique expertise, and this why this content is something valid that you should pay attention to. And then in schema on the back end of the website, so that's the part that the robot reads, but the human never sees. There are schema standards you can follow to say person equals X, and you are then spelling out for the machine. What is you know why is this particular person authoritative in this particular way, and then of course what the press says is really important. So if you read about how people are adding plus Reddit to Google searches because you don't want that AI overview, you want to see what humans are saying. AI is looking for those signals as well. So what other people say about you is now it's always been really important.

It is more important than ever. So things like how do you appear in reviews, how do you appear in the press—that's going to be important to AI. And if you want a little tip for how does your website actually look to AI bots. So this will take you to a widget on my personal landing page. So it's janetteroush.com. If you wanted to look it up later, that will show you all of the different like how much schema do you have on your website? Is it possible for a bot to access your website without jumping through hoops? Right? Is Cloudflare or Akamai blocking bots from visiting your website? Right now, all web traffic.

I think for the first time, it's 51 percent bots as all of the traffic on the web. There is more web traffic to robots than to humans right now. So that is the new reality that we are we are all operating in. And of course, that then creates some questions, right? Like, are websites obsolete? The way that we have, you know, people excavating Google with its ten blue links, which is now dead and gone. And my opinion for the moment is, no, websites are not obsolete. We still have to have the website because that is our own source of truth. That is where these bots can go to get the information in order to put into those AI overviews. You become invisible if you do not have a website for.

Both people and bots to refer to, and of course human behavior takes time to change. So we are not yet at the place where humans are not going to websites at all. Plus, a website could be an opportunity to have AI actually on the website. So this is something that we have at visittheusa.com, which is our consumer website. We have a partnership with a company called Mindtrip, which does AI. AI itinerary building. This is really interesting because that becomes first-party data that we get to collect. So Brand USA can read all of the chats that these people have with our chatbot about what to do in the United States. So while website traffic may be smaller today than it was a year or five years ago, it's possible to know more about that person visiting your website today.

Than we ever could have known before, so that is a really interesting opportunity I think for you to learn more about your website visitors. If you have and there's a number of different services that do this, or you can build one yourself, and then that allows you to collect this information. But not everybody is going to go to a website in order to interact with their chatbot. Where I see all of this going is that just like with Muse. And my personal agent on my phone. I think we are all going to develop a personal agent that runs pieces of our lives for us, and that could be ChatGPT on your phone or Claude. It could be in the future, just like an open sourced model that you host locally on your phone, where the information never goes into any cloud anywhere. It's something that you own completely. But what happens?

When people are planning a trip, in that scenario, right? And we know now this is the same study from Booking that only six percent of people a year ago were fully trusting AI solutions, and I'd say that six percent too many. Like even today, I would not fully trust an AI solution for booking things for me, right? So, how do we kind of? Surpass that trust barrier. We need to think about that content on our website as how can we make it available to people even if they are not on our website. How can that content become portable so that if I am using my personal AI on my phone, people are still able to access that information. And I think that people will want to access that information. I think your unique point of view.

Is really valuable. So, looking at who in the room is a DMO or representing a city. So, right, we have this question about our websites, and it's hard to track what information is being used. And we have those questions about what should we be doing with those websites. But I think DMOs sit in a really unique place in online content because we are the only people who sit at that intersection of Government, the local government, the visitor, the tourism business, and the resident. So that's a really unique and interesting perspective. So, this fake photo is representing when I worked at New York City Tourism. We did a partnership with Sony for a new Ghostbusters film that came out a few years ago. So the Ghostbusters Firehouse in New York City, which is not open to the public.

Is surprisingly to me like it's a big tourist destination. So not only do people still 40 years after that film came out, want to go and see the real Ghostbusters firehouse. Like they have an event every June where they do Ghostbusters Week, the week that the film came out in 1985, where and they do parties in the street about it. Like. So we, as part of this partnership with Sony, we made all of this content for the website. We and went and did like interviews with the working firefighters who actually work in that firehouse and what is it like to do that particular job, and kind of represent the city and a film franchise in that way. Nobody else is making that content because nobody else is going to monetize that content. There's nothing in it for them. So.

That unique point of view is something that I think belongs in tourist itineraries. Right, for the right person, they need to be able to access that kind of storytelling. So, I think the future of what DMOs will look like is that we need to serve as a source of truth in some way for AI systems, and honestly, not just for AI systems, but For social, you know, for influencers as well. So, in New York City, I just learned this like a couple of years ago. There is a whole thing where somebody, some years ago on TikTok, said the best place to see the fireworks at midnight on New Year's Eve is you know Brooklyn Bridge Park overlooking the Brooklyn Bridge. There's no fireworks in New York City on New Year's Eve. We do that on Fourth of July. So.

There are people who gather in the park and stand around, and then at 12:05 they're like, "Oh, gee, I guess there's not any fireworks happening," and I spent my evening outside. And so it becomes a reputational risk for destinations when when influencers are spreading untruths or when AI is making things up. And there's a number of different examples of AI sending people to hot springs that don't exist and putting visitors in kind of dangerous situations. So I think one job of the DMO in the future might be to serve as a source of truth with our websites, with what information about our destinations is correct or not correct, and so.

I have kind of a bigger idea of what this could look like that I'm calling destination as a service. So it's the idea that our content, yes, it needs to live on a website, but it also needs to be available inside of my ChatGPT account or inside of you know from a B two B perspective. If I'm working with tour operators in Italy and they're building itineraries, I want DMO content. To be accessible to those tour operators, and so the way that I see that kind of coming to life is that an AI tool. Let's say I'm standing in Times Square. I'm like, okay, what should I do next? Should I go to the Empire State Building or visit the Statue of Liberty on the Staten Island Ferry? I think that people will use something like ChatGPT in the future to ask that question.

And then it could pull in information from New York City Tourism, and the subway system, right? The MTA and a weather app to say, okay, well it's kind of cloudy right now, so it's not the best time to go to the Empire State Building. But I can see from the foot traffic camera in Lower Manhattan that it's less busy than usual. So this is the subway that you are going to take to go to Lower Manhattan to do this event. And oh, here's an API to connect you to the Empire State Building to buy tickets for later in this afternoon. So I think that is when we talk about AI and personalization. I think that's how it is going to come to life. And I think the way that DMOs participate in that is that we need to take the information on our websites and set it up with a particular like schema in the background.

That tells ChatGPT or tells the tour operators, you know, AI system that they have set up. I am an authority on these things and these areas, and this is where, on my website, you are going to find that information. And so, in the future, what I think this means is that we have to know that the AI systems for travelers and for B two B partners. They may not only be visiting our websites; they're going to be visiting the data that exists on our websites, and they're going to be pulling that into their own tools. And we have to start getting ready for that. I think we need to publish our information to be machine readable, and that has to be a priority for us. I think there's an opportunity for DMOs to become trust brokers.

Our destinations, so that in a world full of lies, you know you are going to find the truth about a destination in this one specific place. And I think that we can create this verified, high quality content, even if it is only for a B two B perspective. I think that's incredibly important in the short term for how we think about using AI. And so with that. That is how I am thinking about the customer journey and how they're engaging with AI right now. Now I want to move to the back office, which is the piece of how are we using AI for the work that we do every day. And I'm going to walk through this in kind of four different categories. So kind of starting with the simplest and then moving to the most complex. So we're going to start with a task. Then we're going to move to a tool. Then we're going to talk about a.

Process that you can reinvent with AI, and then finally ending with an institution. What's like an entirely different way of thinking about doing your work with the help of AI. So starting with the task, this a tool called Descript. So this if you're primarily a ChatGPT or Claude user. This happens to be a separate subscription to a separate AI tool. I will say what I love about Descript is that you can use a connector to connect it to your ChatGPT or Claude account. So now, when I am editing videos, I just open up Claude and say, "Hey, can you find the video from this webinar I did today? And these are the edits I want you to make to it." And I tend to make the same edits every time. So now that becomes a skill.

In Claude, so which is a repeatable workflow, so it knows okay you're going to create a card to go at the beginning of the video that has the name of the webinar on it, and you're going to add in this little clip of background music to make the video feel like it's a little bit professional. Like, is it as good as what a real video editor would do? Like, no, I'm not going to pretend that it's a one for one exchange, but I wasn't. Hiring a real video editor to do this previously, so nobody wants to click onto a website and you have to wait through the three minutes of okay, great, just waiting for everybody to join. You know the typical webinar launch stuff. You want it to feel like you're on YouTube. You have the music, you have the title card, and then you go straight into the content. And so now that's something I can do on my own with no video editing skills whatsoever.

So this example, I'm going to show a short video. At Brand USA, we do a lot of international trainings for tour operators and for travel agents. And a lot of the companies that we work with don't have people who speak multiple languages, but they still want to offer training videos on their products or their destinations right to these international offices. So this is something we actually launched for South America, where. The woman who manages our South American outreach will take videos and then put them into Descript, which will both translate the captions on the video, and will redo the voiceover using your voice in order to put it in language for the people who are watching.

So that's my voice saying this. Does anybody in here speak French? So I know the people who do speak French. What you may have noticed in that video, right, is that that's a French Canadian accent. So Descript, for whatever reason, they don't have a French speaker as the basis for their French language. It's somebody from Montreal. And that's not something you know. I'm monolingual, so that is not something I would have ever picked up on. I'm like, I don't know. It sounds pretty good to me, but it's a sign that you want people to stay in the loop when we do this stuff. When you hear human in the loop, we can't get rid of people reviewing these materials before we put them out in the world. Because now I can apologize for the French Canadian accent, or if I'm in Canada, I'm like, look, I nailed your particular accent.

But you need to understand that that's what is happening here, and if you don't have a real French speaker listening to the video, you may not know those nuances. So, whatever you do with AI, you have to stay very involved with the work that you're doing. So now I'm going to move on to using AI as a tool. And as a side note here, I'm not a tech person. Like I worked on Broadway for 20 years. I was a theater major. I think my introduction to technology was that in my very first internship, it was a job I got at my stepmother's company because there was an EA who didn't want to learn how to use the fax machine. So, so I was the tech guru who sent the faxes for this person who didn't want to learn that technology, right? But it's like.

That's also why we kind of need to learn how to use AI because what you're going to hire the intern to press the AI button for you? No, that doesn't make sense. And also, people who have done their jobs for a while are the ones who are most equipped to use AI. I'm like I'm Gen X. I like to say this is the superpower of Gen X because we've been through these technological evolutions. We know what it's like when something doesn't work. First time, and you have to take the cartridge out and blow on it and put it back in again, and that's AI. Like it does not work the first time. You have to have the patience to keep working with it until you get the output that you want. But I say that because I'm using AI to build some cool things, and this is available to all of you. I do not have some kind of special magical tech skills here. I, you know.

I I am exactly like all of you. I promise. Oh my gosh, this error message that keeps coming up. So this is something that I built for Brand USA's Global Ambassador Program. So we have 15 international offices, and we wanted to create a program around America 250 this year, where we bring in 250 travel advisors who are also content creators. And give them the title of global ambassador, and then work with them on creating content about visiting the United States. And originally, when we were going to take these applications, the team built a Google form for these applications, right? And a Google form's fine, but it also it looks like a Google form, right? And the team came to me, and they're like, we know there has to be a cooler way to do this.

So I just went to Claude. I'm like, there's got to be a cooler way to do this, right? Like, what should we do here? Oh, well, I can build a front end for a Google Form. It still works exactly like a Google Form. The data still goes into a Google Sheet, but it can look like a beautiful website that people can fill out. And then on the back end, so this is again a regular old Google Form that's collecting the information. It is able like what the process was going to be is that somebody on our team would have to check every single day to see if any new applications had come in, and then forward that information to the correct person at one of fifteen international offices in order to move that application forward. Instead, this automatically sorting, and this fake information, by the way. So I'm not sharing PII.

I'm like Claude. Make make a fake version of this spreadsheet that I can put on a room and put in front of people in a room, and you'll see at the bottom. It's like every single market that we work with. So your application automatically gets sorted to the correct page in this spreadsheet, and then it automatically sends the application by email to the person in the office who is responsible for vetting it. So now this thing that would have You know, over time, like maybe initially, right? You'd remember to check it every day, but it's still live now, and this is something we launched in March. Now we don't have to worry about checking it every day. Everything is automated. Plus, the spreadsheet serves as a backup. If we worry that, oh gosh, did something go wrong or something get lost? No, you can come to the original spreadsheet and make sure that everything is still working. So.

Like this is something that now operates in a really slick way that it didn't necessarily work so slickly before, and that's possible because of AI. Now, the piece that I am going to walk through is a process. How can you reimagine a process at your company with the help of AI? And the process I'm going to share is working on your annual plan or the strategy for your company. So, who here has been involved in a strategy or an annual planning process, right? So, when I was in New York City, that's something that involved you know you had to go around to every single department head and say, okay, here's the PowerPoint template that we need you to fill in with your objectives for the next year. Please make sure they ladder up to this larger idea, and then people who.

Like maybe this isn't their primary job. It's to fill out powerpoints about strategy. You have to kind of keep chasing them to get the information, and then it you know gets handed off to an intern to put together in one giant deck, and then you file the deck in your hard drive, and you look at it next year when it's time to start the process all over again. So we wanted to approach the. The process of putting together an annual strategy for the company a little bit differently, and so I started by recording the conversations that we were having. So I'm like I'm wearing a million different little microphone things right now. One of them is what is called a Plaud Note. So obnoxiously, it rhymes with Claude, but it has nothing to do with Claude. So it's spelled Plaud, like like applauding.

I think is the shtick that they are going for. It's just a personal recording device, and it syncs to my phone, and then the phone pushes it to you know the website, to a a Claude website that then connects to my Claude account. So I can leave this meeting, and I can get the transcript of everything that I said today, and I could turn it into a website for all of you to look at. If I wanted to, like I know, often people say, "Oh my goodness, can I get a copy of your slides?" I don't know what you're going to do with my slides, right? Like they're cute, maybe, but they're not going to tell you what I talked about. Claude could create a website that has my talking points and then uses the slides as like decoration or to illustrate the different points. So, like that's the kind of thing that's possible when you start collecting transcripts for what you're talking about. So here.

I start collecting the transcripts of our strategic conversations. What do we want to accomplish in the next two years? And then I am able to take that information and I used it to create a password protected website that captures what we are working on as a company as our four strategic pillars for the next two years. And what I really like about the live website. Is that when you're working from PowerPoint, you know it's like that saying that if you have a hammer, everything looks like a nail. With a PowerPoint, everything has to fit in that PowerPoint slide, and you'll spend a lot of time. Like I know over my career, it's like, okay, am I cutting words to make this thing fit on the slide, or should this strategy be two pages? But then the other one was only one page. You spend so much time thinking about formatting that you aren't.

Focused on the ideas, so when something is a website, if you need to scroll down in order to get more information, you're able to do that. If you need something to be, you know, have a link, so if you want to read more, you can, you're able to do that. So sharing information in the format of a website is a really useful way of approaching your work, and so I've gotten a ton of value out of.

Making websites for everything because it's free, and when you're done with the website, you just tell your AI tool, "All right, take take it down now," and it's a more visual way of taking in information. What I also like about this is, as things change throughout the year, I can automatically update the website to reflect. Oh, here are some new tactics that we have introduced for this particular program. So it's not something that sits in a file and you know gathers dust. It becomes something that we use in every single internal strategy meeting, in order to understand what are we doing and where are we going. And you know what? We have a new CEO who's going to be starting in about a month, and I'm really glad that we already have the groundwork laid to bring her up to speed on what is the company focusing on. So we don't have to scramble and put something together.

Everything already is crisp and ready to go. And so now the final piece that I'm going to walk through is as an institution, how could AI change the way that we do our work? And so for this, I'm really thinking about AI first companies. So the one that I think about the most is a company called Y Combinator, and so that is a tech like accelerator. Basically based out of San Francisco, Sam Altman, who is the president of OpenAI, used to be the head of Y Combinator. So this a company that is advising startups on how to 10x their growth, and they obviously want to be an AI first company. So this the way that they operate. So the way that I have copied how they work is I've looked at.

Our 15 international offices that I mentioned earlier, reporting from 15 international offices is a lot to keep on top of because that's they're sending us a PDF file once a month that says, "Oh, here's all the work we did for you last month." They're having weekly phone calls. They are having monthly check-in meetings. They are having quarterly kind of big report outs. You know, to you know the most senior people at the company. And that's happening for the people at Brand USA who are managing these offices. Like that's 15 different offices operating, you know, around the clock. So scheduling these meetings can be difficult because it may be midnight for you when you're jumping on these calls. And again, it turns into reports getting sent, and then they they go into a folder on we use box for our filing. So you might use.

SharePoint, you might use Google Drive, but then the files kind of sit and have the chance of gathering dust, and that's not something that we want. And so I've created a system that, and Claude was very insistent that it create an imaginary example and that it tell you that it's an illustrative example and these aren't real numbers. So if you're familiar with working with Claude, it's always a little bit like, don't you do that? That's going to be wrong. Like thank you yes I won't share confidential information at the conference, but like I want to know not if I wanted to ask the question like what is the luxury traveler doing, right now if I wanted to find the answer to that question as me as the chief AI officer, I would send a Slack to the travel trade team or I would send emails to individual people to see you know what are you hearing from your markets or I could go into Box.

And I could find the France file, and then I could read the last three or four monthly reports from France, and maybe I repeat that for all 15 markets. So it's a little bit of a clunky process. And if I just relied to asking our internal folks, we're relying on—I hate to say it—right, our unreliable meat brains, like humans. Aren't necessarily perfect at everything, and remembering something that happened on a phone call that might have been at midnight your time, that's not going to be the strong suit of a human being. So I created a system where we organized all of that information. So here.

The monthly reports, right? This what our this a screenshot of Box. Like they get saved in all of these different folders. Every single folder has a different layout. So some are nested, some are organized by fiscal year, some have all of the reports in one folder. And Claude went in and created an organizational system for me. So it's created. This little system here, so that it has all of the information that it needs on these reports, and then it made an index. So Claude went in and read every single PDF file, and then in a new document, it wrote down: Okay, for this file, which is saved in this place, so it's the France May two thousand and twenty six file. These are the six topics that were discussed in that report, and here is a link to that report.

And then I connected all of that to a project in Claude, and that's something ChatGPT also has projects. It's like a folder in the AI tool that allows you to organize the information, and then you can share it with multiple people in your company. So if a person at Brand USA has access, has permission to access this particular box folder, and they are also I have invited them to be in this project. In Claude, then they can ask a question like, you know, who's talking about AI during you know our our meetings? Who who's who's addressing AI, or any markets discussing this? Claude will go to this index file that is saved in Box, and it's going to look for AI. And every time it sees that AI has been indexed, it's then going to look at the source material and say, oh, South Korea.

Talked about how travelers were using AI in the report from July, and it's going to collect all of that, and then it will come back and say, "Great, here is all of the information. Do you need this formatted as a PowerPoint? Is this an executive summary that you need to give to your CEO? Is this, you know, a PDF file? Are you trying to write a LinkedIn post? Because the AI tool can create any format of output that you want. It's just a matter of giving it the right kind." Context to put into that information so that it's not making things up, it's not hallucinating, it's looking at your real information, pulling quotes for you, and then telling you where those quotes came from, so that I can go back and read those source documents myself if I don't trust that Claude is telling me the truth. And I think that in the future, that's how organizations are going to work. That is exactly how.

Y Combinator works is that you have this system in the middle. So for us, it's this index of the reports from our international offices, and then people sit around that system, feeding it in information and pulling out information. So like the next step in this journey is having everybody who works at Brand USA who have meetings with the offices. Use a tool like Granola or a tool like Plaud to record phone calls. Obviously, with permission, but to take those transcripts and then have those added to the index as well, which then starts to create this much richer, living source of truth about what is happening in all of the international markets that we're working with, and.

If you're trying to, if you're struggling figuring out like how does this apply to you or your organization because maybe you don't have 15 international offices, you could just start with any source of information that like an event. So the organizers of this event, if there is different folders you know in your computer where you are saving post-event surveys and all of the contracts that you fill out every year and all of the work. That happens around producing a governor's conference on tourism. You can use that to create an index, and then you have one source of truth where you can have Claude read the actual files that you have saved on your server about this particular conference and say, "Oh, what was that thing that we did three years ago that everybody liked so much?" And instead of digging through files to find that out, Claude will look at the index and pull that answer out for you.

And what all of these examples have in common is that nobody had to make massive changes to the way that they did their work. So for the translation example, the person doing the webinar didn't have to learn French. They didn't have to change the way they do their work. AI filled in the gaps for them. For the application, the people at Brand USA didn't have to change the way they work. In order to process these applications, the applications came to you in the way that you expect to do the work. The person filling out the form got to fill out a nicer form than just your standard Google form. Like they weren't asked to do something differently. In the Hive Mind example, the international offices collating their information, the documents get saved in the exact same folders that they always have.

The AI just learns where is that information saved, and it has a record of that, so it doesn't have to keep searching for everything, you know, from scratch every month. It has a directory of where everything's saved. But we didn't have to tell the offices, oh, there's a new system, you have to save it here so AI can read it. No, it doesn't require everybody to immediately change the way that they're working, and that becomes really valuable. For making adoption happen more quickly, so that you can get to the end result of oh my goodness I'm in Claude and I'm getting this question answered that would have taken me three hours to answer before now it's happening in two minutes and that starts to show your team what the power is of working with these tools.

And it allows you to spend your time doing the work that you actually enjoy doing. So my theory is that maybe I'm living a couple of years in the future because it takes time for people to change. I think it takes time for systems to change, but I don't think I am working 20 years in the future. I think that. This is the direction that work is going. That we are going to start making transcripts of more of our calls in order to create more of this centralized intelligence. We're going to see silos in our companies start to break down a little bit more because it's easier to work that way. And the old silos of every department doing their own thing—it's going to stop making sense over the next couple of years.

And I think this idea of personal AI that is only going to expand over the next couple of years as tools like Muse come out that make it more obvious to people what the power is of having an AI assistant helping you with your work. And so, all right, that's a lot, right? How on earth do you get started with this if you're not already super AI pilled like I am? Say first, you want to stay focused on your mission, whether as a nonprofit, right? You are expanding economic opportunities in your destination, or you are a company that is trying to move tickets. What are you trying to get accomplished? Rather than just throwing AI at something for AI's sake, if this, if you are in an opera, you know, a position to do this.

Get a paid account for your team because that's what allows you to turn off training on your data, so your information stays secure. Like if you are just letting everybody at your company kind of do BYO AI, that means when they leave their company, they get to keep all of that information that you have put into the system. And we don't do that with email. You don't get to keep your email when you leave your company, so you shouldn't get to keep all of the information that you put into AI. So it's like. We're a few years into this now. It is time for everybody to have paid access to an AI tool. It's twenty dollars a month. It is absolutely worth it. And when you get back to work on Monday, I suggest keeping your preferred AI tool open in a second screen on your computer and just trying something new with it every day, and knowing it's not going to work perfectly the first time.

That doesn't mean that the AI is incapable. It just means you need to learn the edges of what is it good at, and where do you still need to do that work. And you need to know that that's going to change month over month because AI really is changing so quickly. You can't wait for a perfect use case. I've heard that this is a Gen X thing, particularly that people like oh I'm. I'm totally on board with AI. I have the perfect way that I'm going to use it, and so next week when that comes up, I'm going to give that a shot. Like, no, just start using it imperfectly now, and share what you are doing because OpenAI isn't going to make a guide for tourism on how to use ChatGPT. Right? It is up to us to figure out ways to share what we are doing, just like I did today.

So that people understand what the opportunities are and start to take advantage of them, and finally, please remember when we are at work, it is not cheating to use AI, and that can be a little bit of a mental trip up for us as well, right? That's like, oh, I I should do this the right way, the way that I was trained. I will say, if you work for a nonprofit, like we are, we are spending taxpayer money. Right, we are stewards of those funds, and I think we have a responsibility to spend them in the most responsible way that we can, and that includes if there is a way to do something better or faster or less expensive, we should take advantage of that. And then, if you're a commercial company, if you don't start learning these tools, somebody is going to come around that do your job, but better, faster, or cheaper.

And it will put you out of business. So, like, regardless of what side you sit on in that equation, we have got to figure out how we can take advantage of these tools. They are love them or hate them. It's not going away.