The AI that the local life service industry has collectively placed bets on may still be a bit useless at present.

After AI e-commerce, AI has also started to penetrate the local life service industry.

On June 26th, Meituan officially launched its first AI digital employee for service retail merchants and will provide each store with a professional AI business assistant. Meanwhile, Wang Puzhong, the CEO of Meituan's core local life business, revealed that Meituan invests over 10 billion yuan in AI every year.

Coincidentally, on June 24th, Ele.me announced the launch of the first domestic rider-side intelligent agent based on a large model, Ele.me's AI assistant "Xiao'e", which provides comprehensive support and services for millions of blue riders.

Douyin Life Services has also been active. It launched an AI product called "Tanfan", which uses AI technology to recommend local food restaurants to users.

Obviously, the giants in the local life industry are collectively flocking to AI, integrating the technology into every aspect from merchant operation, rider dispatching to user recommendation.

Fill the entire local life chain with AI

This AI competition covers the entire chain of merchants, riders, and users.

On the merchant side, AI mainly plays the role of a digital employee to reduce costs and increase efficiency, helping merchants with daily operations.

Meituan launched its first AI digital employee for service retail merchants, providing services such as business diagnosis, growth opportunity identification, operational support, intelligent scheduling of staff shifts, and customer service reception. Recently, it has also been conducting internal testing on the Kangaroo Advisor, which, relying on the massive information covering 4 million stores across the country and over 10 years of online catering operation experience, provides merchants with industry insights and practical business suggestions. It mainly covers four scenarios of intelligent business operation, including track selection, store location selection, dish development, and store operation.

Ele.me launched the "Merchant Settlement AI Assistant". From the pre - settlement stage, to store decoration, and then to post - operation, the AI assistant can help merchants with intelligent product selection, intelligent picture beautification, and diagnose the store's business situation.

Douyin's Douyin Laike in Douyin Life Services and Kuaishou's Keling in Kuaishou Local Life both provide relevant AIGC intelligent marketing and creation services for merchants, capable of intelligent video editing and intelligent graphic creation. Kuaishou Local Life has also launched the Nüwa digital human, which can conduct all - day live broadcasts and interact with the audience.

In addition, Kuaishou Local Life plans to launch tools such as "AI Intelligent Product Listing" and "AI Intelligent Product Selection" in 2025, which can intelligently determine when, where, and what products to list.

On the rider side, AI is committed to ensuring delivery safety. Ele.me's AI "Xiao'e" can directly complete operations such as order receiving, arrival confirmation, and activity query. It can also push content prompts such as weather warnings and road closures, helping riders improve operation efficiency and reduce delivery risks. In addition, "Xiao'e" can also achieve the "heat map warning" function based on the rider's historical data and the surrounding order heat map, providing intelligent analysis such as "where there are more orders" and "current income estimate".

Meanwhile, the country's first takeaway delivery vehicle equipped with an "AI brain" jointly developed by Beidou and Yadea was unveiled. The vehicle is equipped with a Beidou dual - frequency chip, realizing intelligent management of the binding between the person and the vehicle. It has functions such as identity recognition, automatic speed reduction, route optimization, and full - process traceability, aiming to solve the problem of "difficulty for delivery riders to enter residential communities".

On the user side, intelligent customer service is a relatively common application of AI technology, which can quickly respond to user inquiries, answer common questions, and improve the platform's service efficiency.

In addition, the platform is also exploring more grass - planting functions, making AI act as a decision - making assistant.

Meituan's "Ask Xiaodai" can provide users with catering recommendations and gift - giving suggestions. In the Shenqiangshou channel, some dishes labeled "AI Smart Selection" will be accompanied by short AI - generated reviews. Meituan's "Dianzai" in the internal testing of Dazhongdianping integrates functions such as store - finding assistance, dish - matching suggestions, and even AI - assisted note - writing.

Douyin launched "Tanfan" based on the Doubao large model, which can recommend nearby restaurants according to the user's geographical location and conduct multi - dimensional comparisons. After determining the restaurant, users can directly jump to Gaode Map for navigation, achieving a seamless connection from "finding a store" to "arriving at the store".

It is worth mentioning that Ele.me also launched the "Holographic Shield" system, which is mainly used for the platform's compliance governance. Specifically, the system can scan all online stores 24 hours a day. Using image recognition technology, it can identify those "ghost restaurants". The processing speed of problem merchants has been directly reduced from half a day to 1 minute, and the efficiency has been significantly improved.

It is not difficult to see that AI technology is being implemented in various application scenarios in the local life field.

However, are these AIs with diverse names and rich functions really useful?

Still a bit useless at present

It is undeniable that the investment and application of these AIs do show the potential to improve efficiency in specific aspects. For example, intelligent customer service can effectively relieve the pressure on human resources in quickly responding to standardized questions. Route optimization and basic information push on the rider side can theoretically improve delivery efficiency.

However, when it comes to more core aspects such as user experience, merchant operation efficiency, and rider workflow optimization, "useless" has become the feedback from many users.

First of all, AI lacks understanding ability. On the merchant side, an independent coffee shop owner said that although AI can quickly generate copywriting, its understanding of local, personalized, emotional, and unstructured information is not deep enough. It is difficult to capture subtle "human touch" and contextual needs, resulting in advertising copy that fails to impress people and attracts few actual customers.

On the user side, some consumers pointed out that AI recommendations tend to pile up flowery adjectives, over - beautify products, lack real insights, and sometimes even provide false information.

Secondly, AI lacks professionalism. An operator of a chain brand admitted that the AI functions of most local life platforms are not very effective. Taking picture beautification as an example, the generated dish graphics often do not meet the requirements. They prefer to use more professional AI tools independently to beautify pictures.

Moreover, the platform's AI tools are difficult to match these professional tools in terms of professionalism, flexibility, and output quality in vertical fields. Due to their mandatory nature, they disrupt the original workflow of merchants, make the operation interface bloated and cluttered, and increase the risk of accidental touches.

A deeper concern comes from the ambiguity of traffic rules. Many merchants have raised the issue that whether it is the consumer - initiated AI search recommendation or the passive "AI Smart Selection" label, there is a lack of transparent standards. "What kind of products and merchants can be placed at the forefront? Will this evolve into a new traffic pool that requires additional payment?"

Finally, the interaction design of AI fails to meet the principle of "efficiency first" in the delivery scenario. A crowdsourced rider gave an example, "For example, sometimes there is a situation where the merchant delays the order. Once you leave the store, the AI assistant will keep announcing that the order has not been picked up. But if there is any noise, the AI will determine that the order has been picked up, which is very troublesome." He revealed that few riders around him use the AI assistant, and some even study how to turn it off.

Overall, users' perception of the value brought by AI is not clear. Merchants have not clearly seen an increase in revenue or a significant reduction in costs. Riders do not feel that delivery is safer or order - taking is easier. Consumers have not obtained a revolutionary improvement in experience. Therefore, the evaluation of "useless" actually reflects a certain gap between the current AI capabilities and the expectations of all parties for "intelligence" and the need to solve core pain points.

 Finally 

The application of AI in the local life field is undoubtedly a major trend in the current industry development.

Judging from the actions of Meituan, Ele.me, Douyin, and Kuaishou, everyone hopes to tell a good AI story. However, to make this story truly appealing, the key lies in breaking through the current "useless" dilemma, making AI evolve from "existing" to "useful", and finally to "easy to use" and "loved to use".

As Jack Ma said, in the future, it is not about AI replacing humans, but about AI liberating humans, understanding humans better, and serving humans well. In this local life AI competition, the one who can be the first to cross the "useless" stage and make the technology truly solve practical problems will be the one who can truly tell this new AI story about efficiency, experience, and trust.

This article is from the WeChat official account "Yilan Business" (ID: yilanshangye), author: Meng Jiayi, published by 36Kr with authorization.