--- title: "Qwen gives away 10 million cups of milk tea. Can this help us develop the habit of AI shopping?" description: "A few days ago, Yuanbao was still sending red packets in the WeChat group. Early this morning, the link changed to Qianwen inviting you to have milk tea. One after another, Brother Long is really worr" type: "topic" locale: "en" url: "https://longbridge.com/en/topics/38574023.md" published_at: "2026-02-06T16:07:47.000Z" author: "[林氪](https://longbridge.com/en/profiles/7312377)" --- # Qwen gives away 10 million cups of milk tea. Can this help us develop the habit of AI shopping? A few days ago, Yuanbao was still sending red envelopes in the WeChat group. Early this morning, the link changed to "Qianwen invites you to have milk tea." One after another, Brother Long is really worried sick. Before 5 p.m., the Qianwen APP announced that orders had exceeded 10 million. Although offline milk tea shops were hit by a saturation attack round, they also set a historical record. At the same time, every time I take the elevator recently, I can see Xiao Zhan's Meituan advertisement, "Eating, drinking, and having fun, Xiao Tuan will arrange everything." Meituan is also fully promoting its AI conversation service. This can't help but make people sigh: "Will it be enough to just move our mouths in the future? Will the apps of the old era all be disrupted?" Faced with this statement, we first have to pour cold water on it. We still cannot be fooled by technology and must return to the user scenario to think about this problem. 1\. Why can't LUI beat GUI? Imagine a fixed scenario: Every morning at 9 a.m., you want a cup of Luckin's iced Americano. In the App (GUI): You open the app, muscle memory makes your finger click "Order Again," pay, and finish within two seconds. The whole process requires no thinking, not even needing to see the screen clearly. In the dialog box (LUI): You need to type (or speak): "Help me order a cup of Luckin iced Americano, deliver to the office, half sugar, no ice." Then wait for the AI to parse it, the AI replies with a confirmation card, and you click confirm. See it? For high-frequency, certain, and standardized needs, the efficiency of natural language interaction (LUI) is actually regressive. The essence of a graphical user interface (GUI) is, "Don't Make Me Think." It encapsulates complex instructions into intuitive buttons, allowing people to complete operations based on intuition and habit. Forcing users to describe a muscle-memory action in language violates the basic principles of interaction design. So, expecting users to develop the habit of chatting with AI every day to order takeout? Probably very difficult. The essence of human nature is to seek simplicity. If it can be solved quickly with hands, we will never waste words. 2\. What is the real scenario for AI? Since simple ordering doesn't require AI, why are giants still investing heavily? Because today's food delivery and even local life platforms are no longer an efficient tool, but a tiring "infinite shelf." When you don't know what to eat, facing a sea of stores, complex rules, and hard-to-distinguish reviews, the decision-making cost is astonishingly high. You keep scrolling the screen, physically and mentally exhausted. This is where AI should shine: It shouldn't be an "order taker," but should become your "decision-making agent." The core scenario of the future is not "help me buy a cup of coffee," but handling those vague, non-standard needs: "I'm so tired from working overtime today, I want to eat some hot noodle soup, not too spicy, preferably delivered within half an hour, budget under 50 yuan." "Singing karaoke with friends on the weekend, see which nearby private rooms are available for booking." Using a traditional app, you need to manually filter categories, delivery times, and then judge taste and reputation one by one... the process is tedious. And a real AI assistant, based on semantic understanding, can instantly "collapse" the entire shelf and directly give you two or three of the most suitable choices. The value of AI is not to replace clicking, but to eliminate filtering. 3\. What will the future look like? So, what will the future ordering interface look like? I believe that a simple chatbot is just a transition. A more likely endgame is the combination of LUI (intent recognition) + Generative UI (dynamically generated graphical interface). A simple search box on the interface. When you type "light Cantonese food suitable for taking parents to eat this weekend," the app no longer spits out a string of links, but generates an aggregated page in real-time: Automatically stitches together pictures of signature dishes from three restaurants to form a beautiful menu. Highlights the label "Good reviews from elders." Directly pins the "Suitable for 3-person set meal" card to the top. The interface is fluid, dynamically reorganized in real-time based on your intent, everything broken down and then reassembled. The globally optimal choice. You spend half an hour choosing, only to find that it's not as good as the result the AI produced in three seconds. At that point, everything becomes irreversible. 4\. Conclusion Back to the question in the title, will people get used to ordering takeout like this? My answer is: People may not like chatting with machines, but they will definitely love the ease of not having to choose. Meituan's "Ask Xiao Tuan," Qianwen's large model... current attempts may still seem clumsy, sometimes even slower than directly opening the app. But this might be a battle for the "entry point" of local life. Whoever can first cross the stage of just being a chat companion and truly achieve "intent as a service"—you just say what you want, and I bring the organized solution to you—might get the first-mover ticket for the next decade of local life. The ultimate form of service will evolve from providing infinite choices to providing one of the optimal choices. This not only applies to food delivery but is also likely the future of e-commerce and even more industries. $BABA-W(09988.HK) $MEITUAN(03690.HK) $PDD(PDD.US) ### Related Stocks - [03690.HK - MEITUAN](https://longbridge.com/en/quote/03690.HK.md) - [83690.HK - MEITUAN-WR](https://longbridge.com/en/quote/83690.HK.md) - [MPNGY.US - Meituan](https://longbridge.com/en/quote/MPNGY.US.md) - [PDD.US - PDD](https://longbridge.com/en/quote/PDD.US.md) - [09988.HK - BABA-W](https://longbridge.com/en/quote/09988.HK.md) - [89988.HK - BABA-WR](https://longbridge.com/en/quote/89988.HK.md) - [BABA.US - Alibaba](https://longbridge.com/en/quote/BABA.US.md) - [KBAB.US - KraneShares 2x Long BABA Daily ETF](https://longbridge.com/en/quote/KBAB.US.md) - [BABO.US - YieldMax BABA Option Income Strategy ETF](https://longbridge.com/en/quote/BABO.US.md) - [BABX.US - BABA 2x Long Daily ETF - GraniteShares](https://longbridge.com/en/quote/BABX.US.md) --- > **Disclaimer**: This article is for reference only and does not constitute any investment advice.