TRIP.COM(US:TCOM) Zhong Guo Jing Ying Bao·2025-04-29 21:57Core Viewpoint - The OTA platforms are intensifying their AI competition ahead of the "May Day" holiday, launching new AI-driven travel planning products to capture user attention and increase usage rates [1] Group 1: AI Product Launches - Tongcheng Travel has integrated the DeepSeek large model, while Fliggy has launched a multi-agent travel planning product called "Ask One" [1] - Tuniu has announced its self-developed AI assistant "Xiaoniu," and Ctrip has introduced "Ctrip Wenda" in 2023, indicating a rapid deployment of AI models in the travel industry [1] Group 2: User Experience Testing - A test was conducted comparing AI travel planning from Tongcheng, Fliggy, and Ctrip, revealing significant differences in recommendations and planning styles [3] - Tongcheng's "Chengxin AI" provided a comprehensive 5-day itinerary for Kunming, exceeding the budget of 5000 yuan, while Fliggy's "Ask One" offered a more extensive travel plan that included Dali and Lijiang, but was more demanding [3][4] - Ctrip's "Ctrip Wenda" provided a simpler 3-day itinerary without specific booking information, focusing on local attractions [4] Group 3: AI Product Limitations - The AI products are still in the experiential stage, with no clear winner among the platforms, suggesting that users may need to combine recommendations from multiple sources [4][5] - The current AI offerings serve as auxiliary tools without significantly enhancing efficiency in the travel planning process [5] Group 4: Data Quality and Challenges - The effectiveness of AI in travel planning is heavily reliant on data quality, with issues such as data fragmentation and noise affecting output accuracy [7][8] - Fliggy's "Ask One" utilizes a combination of real-time supply chain data, cleaned UGC data, and structured workflows from travel experts to enhance the accuracy of its travel plans [8] Group 5: Commercialization Prospects - The key to successful AI integration in travel lies in its utility for users and the ability to commercialize effectively [9] - Current AI models focus on improving user satisfaction rather than achieving significant commercial success, with platforms still in the early stages of AI model training [9] - There is a need for AI to bridge the gap between consumer needs and service provider capabilities, potentially enhancing personalized travel experiences [9]