Google Cloud organized an annual conference ‘Google Cloud Next Tokyo ’24’ on August 1 and 2, joined by JR EAST, Hoshino Resorts and LINE Yahoo that introduced how they use generative AI. It was clear to see that the use of generative AI in the tourism space is beginning to make steady progress.
JR EAST uses generative AI for travel planning support service
Katsumi Ise, JR EAST Vice President for Innovation strategy, explained that ‘JR East Travel Concierge’ a travel planning support service for inbound travelers to Japan was launched on July 29 this year, installing Google’s cutting-edge generative AI ‘Gemini 1.5 Pro.’
The service, jointly developed by JR EAST’s Digital & Data Innovation Center (DICe) abd Google Cloud, consists of ‘spot suggestion,’ ‘information delivery’ and ‘itinerary creation’ based on the user's chat-style dialogue and past experiences. The Gemini 1.5 Pro's unique feature enables the railway operator to narrow down suggestions by incorporating the user's reactions during dialogue, and to provide more personalized information and experiences in a timely manner.
"there are many challenges, such as language barriers, cultural differences, and a lack of information about local areas, which means that opportunities for regional economic development are being missed,” Ise said. “"With generative AI, we will broaden the scope of travel in Japan and allow inbound travelers to enjoy the good things about the regions.”
In addition, JR EAST uses Google technologies in its original apps that guide users to major stations, underground shopping malls as well as route searches that take train delays into account.
Hoshino Resort talks about merits of generative AI
Hoshin Resorts has introduced ‘Gemini for Google Workspace.’ Yoshiharu Hoshino, Hoshino Resorts CEO, said, “Generative AI will transform the tourism industry. Earlier entry into the market will allow us to grow further and maintain our competitiveness. I think it is important to have generative AI as a partner,”
Hoshino explained four perspectives as reasons for installing generative AI: (1) foreign languages, (2) accuracy, (3) breaking away from stereotypes and (4) stimulating creativity.
Breaking away from stereotypes is important for daily decision-making by staff, he said. “Market research is done for big decisions, but when staff make small decisions, there are limited budgets and time, and what they can do is to trust their own instincts. I thought that would be a bottleneck. In real market, different things from each person’s instincts are happening. So, I expect generative AI in breaking away from stereotypes.”
Furthermore, generative AI enables them to use time, energy and cost in different spaces. “Generative AI is a tool that everyone in the market can get. Originality and how you create your own strategy will be important,” Hoshino said. “We have to decide what to choose from the options presented by generative AI, and create our branding and personality ourselves.”
LINE Yahoo delivers more convenient shopping experiences
Using Vertex AI Gemini API that processes multi-information including texts and images, LINE Yahoo provides a function automatically generating product descriptions from the product name, category, and image entered by the user when listing for Yahoo! Flea Market service.
Gen Miyazawa, LINE Yahoo senior executive director for generative AI, explained that text generation speed is about 5 times faster, at an average of 4 seconds, than the previous model, and the listing completion rate has also increased by 3%.
“We think that generative AI’s strength is to connect multiple services to provide more comfortable user experiences,” he said. LINE Yahoo plans to provide chat-based shopping experiences even in the search areas, such as Yahoo! Travel or Yahoo!Shopping.
Generative AI is shifting from the test phase to the usage stage
Tomoyuki Hirate, Google Cloud President Japan, said “AI is definitely shifting from the trial stage to the usage stage in the business world,” appealing Gemini 1.5 Pro that can handle large amounts of context and simultaneously process data in different formats such as text, audio, images, and videos.
He also pointed out that the usage way is rapidly shifting from the use of models to the use of generative AI agents, an autonomous system that thinks about and executes multiple tasks without human intervention to achieve a specific goal.
“By linking with existing applications through AI agents, the entire business can be covered, and business efficiency and automation can be achieved,” Hirate said. “Many companies will begin to build and use generative AI agents in earnest in the future.”