Bei Jing Shang Bao·2025-05-12 09:32Core Viewpoint - The online travel industry is experiencing an unprecedented AI race in 2025, with platforms like Tongcheng Travel, Fliggy, and Mafengwo rapidly updating their AI products to gain a competitive edge in itinerary planning, intelligent customer service, and supply chain optimization [1][3]. Group 1: AI Product Development - All surveyed online travel platforms have integrated the DeepSeek model, with an average product iteration cycle of 1-2 months [3][6]. - Mafengwo launched its self-developed AI travel assistant "AI Xiaoma," which fully integrates DeepSeek and its own refined model [3][4]. - Fliggy's AI product "Ask One" has also accelerated its launch, showcasing a trend of rapid AI product updates across the industry [3][5]. Group 2: User Experience Enhancement - The upgraded AI products are designed to shorten user decision-making time and assist in itinerary planning and travel strategy formulation [6][7]. - Mafengwo's "AI Xiaoma" can provide real-time Q&A, itinerary planning, and personalized recommendations, significantly reducing the time users spend browsing [6][7]. - Fliggy's "Ask One" includes various specialized assistants, such as budget managers and itinerary assistants, to streamline the decision-making process for users [7][9]. Group 3: Talent and Investment Trends - The demand for technical roles such as algorithm engineers and AI product managers is increasing due to the growing investment in AI by online travel companies [9][10]. - Companies like Ctrip and Tongcheng Travel are actively hiring for these positions, with salaries ranging from 20,000 to 50,000 yuan per month [9]. - Ctrip's R&D expenses for Q4 2024 reached 3.4 billion yuan, a 16% year-on-year increase, indicating a significant investment in product development [9][10]. Group 4: AI Limitations and Future Outlook - Current AI technology is still maturing, with the ability to replace only simple tasks in the short term, while complex decision-making still requires human intervention [12][13]. - Experts suggest that AI can assist travel agents in creating itineraries but cannot fully replace the human touch in customer interactions [12][13]. - The future of the online travel industry may involve a high-frequency service model, where AI enhances user engagement and decision-making frequency [15][16].