Wang Xing, founder and CEO of Chinese delivery giant Meituan, has warned that the coming wave of autonomous AI agents will pose a larger disruption to his business than earlier large language models such as ChatGPT. The comment, made in internal and public remarks, signals a shift in the company’s technology priorities toward systems that combine planning, task execution and multi-step automation rather than stand‑alone conversational models.
Wang paired that technological prognosis with a cultural prescription: he urged colleagues to reduce what he called “登味” — a Chinese phrase that conveys ceremony, formality or excessive deference — and argued against honorifics such as “Xing ge” when addressing him. He proposed that staff use direct, first‑name forms of address internally to flatten hierarchy and accelerate decision‑making.
The distinction Wang draws between chat‑style models and autonomous agents matters because the two classes of AI have different business vectors. Chatbots mainly augment information and customer interactions; agents stitch language models together with tools, workflows and real‑world APIs to carry out complex tasks autonomously. For a platform built on logistics, local commerce and on‑demand services, agent technology can change routing, allocation, customer triage, merchant automation and — crucially — labour composition.
Meituan has long invested in algorithms for dispatch, pricing and local search; treating agents as a strategic frontier suggests the company anticipates a second wave of productivity gains and product innovation. Agents could automate many of the platform’s routine operational decisions, speed up onboarding and customer service, and create new customer experiences that combine planning, multi‑step task flows and real‑time sensing.
That technological ambition, however, carries trade‑offs. Faster automation raises regulatory and social risks in China’s sensitive environment for big tech: regulators have already fined Meituan in the past and scrutinise platform labour practices. Widespread deployment of agents could reduce demand for human couriers or reconfigure their work, exposing Meituan to labour tensions and public scrutiny if not managed carefully.
Wang’s cultural exhortation is equally consequential. Asking employees to drop honorifics and embrace more direct modes of communication is a familiar tactic in tech firms seeking speed and experimentation. In Meituan’s case the move reads as an attempt to break down internal bottlenecks that could slow agent experimentation: faster iteration requires junior staff to surface problems quickly and for managers to cede decision rights.
For investors and competitors, Wang’s comments are a public signalling of priorities. If Meituan moves decisively to integrate agents across logistics, merchant services and new retail experiences, it could widen the gap with peers that treat generative AI primarily as a customer‑facing layer. At the same time, Meituan will need to navigate regulatory expectations around employment, safety and algorithmic transparency as agents take on operational autonomy.
Ultimately, Wang frames the coming months as a race not only of technology but of organisation. Success will depend on whether Meituan can pair ambitious engineering with credible safeguards for workers and consumers, and whether a less ceremonial internal culture actually speeds productive change rather than becoming symbolic managerial theatre.