The goal of Alibaba's Qianwen APP upgrade is very clear: to create a "super portal" like "Alipay" or "mobile Taobao" with daily life of more than 100 million yuan in the AI era. The goal is to create a truly "capable" super portal that deeply integrates AI capabilities into existing business segments such as maps, food delivery, ticket booking, office, learning, shopping, and health. 1. Core: Connect the "family bucket" together Alibaba's biggest assets are Taobao (shopping), Ele.me (food delivery), Amap (maps), Fliggy (ticket booking), etc. These businesses are located in different apps, and now Qianwen APP wants to be a super entrance. You just need to "open your mouth" and give instructions (such as "help me book a flight to Shanghai tomorrow and order takeout") to handle the matter all at once. 2. Purpose: Seize the "dialog box" In the future, the Internet portal may no longer be a search box or APP icon, but an "AI dialog box". Whoever occupies it, controls the users. Alibaba must seize this strategic high ground. 3. Key: Feed AI with 'doing things' AI that only "chats" is not smart enough. Only by allowing you to use it for "shopping" and "ordering meals" can Alibaba obtain the most authentic user data. These data are the best 'nourishment' that can make Qianwen AI smarter and more capable as it is used. 4. Defense: Defend your own territory Tencent (Yuanbao), ByteDance (Doubao), and Alibaba (Qianwen) have successively stepped down, which is also a defensive battle. If Alibaba doesn't do it themselves, users may use someone else's AI assistant to "manipulate" Alibaba's services (such as using another AI to place orders on Taobao), and Alibaba will be completely "passive". Qianwen APP is an attempt by Alibaba to create a future life "gateway" by repackaging all its services using a complete set of AI technology. This is a competitive advantage that Tencent and ByteDance do not possess, but the user experience of Alibaba's products still lags far behind the other two. They are not good at the To C market and do not have enough high-quality content to provide for AI learning and generation. In addition, I have also experienced Qianwen's products in practice. Perhaps due to the quality of the article content, the advantages of the algorithm cannot compensate for this disadvantage. The quality of the answers to the same questions is not as good as Yuanbao and Doubao. Answering questions with existing reference materials is better, and more emphasis is placed on the integration of Qianwen in our own business. Otherwise, Qianwen now looks more like a semi-finished product