Imagine a dark ride featuring an infinite number of storylines. Not just a few, carefully-scripted branches—but an infinite number of outcomes and interactions along the way. Sound unbelievable? Or at least ridiculously expensive to produce? With the latest advancements in artificial intelligence (AI), an attraction like this may not be far away.
“The idea of having custom content for every guest, where they can have an ongoing, personal relationship with characters, that’s always been the dream,” reflects Michael Libby, a technopreneur who is the founder and CEO of an end-to-end show control solutions provider Worldbuildr. Libby’s career has included work with Disney, Thinkwell, and Microsoft, where he worked on ChatGPT before it was released to the public.
“Up until now, if you wanted any sort of interactivity or branching narrative, you’d have to develop multiple versions of the same content. Now, you can use AI to generate that content for you,” he tells Funworld.
Custom Content
An AI ride could be like a blank slate, with less hardscaping and more projection screens. Before the ride starts, riders will answer questions, which will generate answers by the AI, which in turn can be automatically converted into audio for animatronic figures or multimedia video that is projected at key points in the ride. Each vehicle will experience unique content as it passes scenes and screens.
“It’s kind of like Mad Libs,” Libby explains. “The prompt engineer has just put in blank spaces where you as the guest fill in the thing that makes you unique. That generates the content for the ride. It’s no different than rendering a video game in real time, except the content is based on answers provided by the AI instead of human-written code.”
Happening Now
All the technology that Libby describes is publicly available today. However, two challenges immediately present themselves. From a technical standpoint, there could be a small degree of latency that needs to be overcome so that the experience is seamless. Park operators also must possess a willingness to accept unscripted content. While this is at the essence of artificial intelligence, it is perhaps a risk too hefty for some intellectual property owners.
In the short term, it may be up to ride manufacturers to develop an AI dark ride that can be customized and sold (or leased) to park operators, much as virtual reality roller coasters were first introduced eight years ago.
“Whoever serves that need first, I think, will be very successful,” says Libby.
Transforming Customer Service with ChatGPT
ChatGPT, which stands for Chat Generative Pre-Trained Transformer, was “trained” on a massive 570GB of internet text databases. The computer programming language uses information it collects to conversationally address user inquiries.
Its large reserve of knowledge means that ChatGPT itself is the ideal source of suggestions on how artificial intelligence could increase park efficiency and generate higher revenue.
The first item on ChatGPT’s list when prompted on how to assist an attraction? Customer service.
“ChatGPT could answer frequently asked questions about the park, such as park hours, ride information, ticket prices, and provide guests with recommendations for rides and attractions,” it shares, referring to itself in the third person.
Virtual concierges and virtual tour guides supported by GPT already exist. Companies such as Deep Brain AI, Pantheon Lab, Soul Machines, and Uneeq create “digital humans” that can answer visitor questions. The avatars—designed with computer-generated imagery (CGI) or AI platforms—are built on top of large language models like GPT. These interactive virtual assistants can appear on monitors throughout an attraction or inside an app on a visitor’s phone.
Building a virtual concierge or tour guide costs around $75,000 and requires two to three months, depending on the level of complexity, according to Pantheon Lab co-founder Christina Tse. More than 70 languages and nearly twice as many accents and dialects are supported.
“Our virtual assistants are integrated with GPT,” Tse explains. “On top of this, we build another layer of company-specific information. It is this latter layer that allows a bot to answer questions like “How many fish are in your aquarium?” or “How fast is your fastest roller coaster?”
Deploying virtual concierges could reduce waiting time at information counters, enable guests to easily book in-park restaurant reservations, and provide multilingual guidance for guests visiting from destinations whose languages are not typically supported in an attraction.
Moving forward, artificial intelligence may also provide quick, affordable translation solutions for games, ride narration, and live performances.
Transforming Content
ChatGPT’s second suggestion for how attractions could use the tool to improve their business focuses on personalization.
“By asking guests about their interests and preferences, ChatGPT could provide tailored recommendations for rides, attractions, and food options,” it notes.
AECOM Vice President and Global Director of Leisure and Culture Services for Asia, Chris Yoshii, takes the suggestion a step further. He expects the AI chatbot to transform how consumers, who demand personalization, search for experiences—and attractions operators need to focus on how to show up in ChatGPT’s replies.
Consumers may ask ChatGPT, for example, “My 10-year-old son’s birthday is next month. What type of party can I throw him?” or “I’m going to be visiting Tulsa in August with my teenage daughter. What attractions might we enjoy together?”
Alternatively, a theme park may ask ticket buyers, “Who is your favorite character in our park and would you like to be greeted by them when you arrive?”
“There’s a business model here,” Yoshii believes, “that says, if you’re an attraction, and you set yourself up to cater to individualized experiences, you can monetize this through these AI search engines, as booking and payment platforms are integrated in.”
Nearly a dozen companies have already developed official plugins for ChatGPT, including several in the hospitality and tourism space. The first movers in this category are Expedia, Kayak, and OpenTable.
Without even investing in AI, attraction operators can reach these new markets by regularly adding targeted content to their websites so that they show up in next-generation searches. At the same time, manufacturers and suppliers—the makers of laser tag games, escape rooms, bowling alleys, media-based experiences, and midway games—need to explore how to make it easy for attractions operators to customize their products.
“This is going to be a very quick transition,” Yoshii says. “Within one year, I think that the business case for this will be obvious and then, after that, people have maybe another year to really adapt. If you don’t, then, redefine your attraction and think about time-sensitive issues and personalization, you’re going to lose business to others that have.”
