Tencent Holdings Chairman and CEO Pony Ma recently addressed the public during the 2025 Performance Communication Meeting, detailing a strategic shift toward specialized AI applications referred to as "shrimp" or "lobster" models. This new direction seeks to move away from centralized ChatBot structures toward a more decentralized ecosystem that mirrors the architecture of WeChat’s Mini Programs. The initiative aims to integrate artificial intelligence across diverse scenarios while maintaining a balance between centralized oversight and the autonomy of third-party service providers.
From Centralized ChatBots to Decentralized Agents
Ma emphasized that the future of Tencent’s AI deployment lies in "raising shrimp"—a metaphor for developing numerous, specialized AI agents rather than relying solely on a singular, monolithic large language model (LLM). This strategy is designed to leverage WeChat’s decentralized concept, allowing developers to maintain their own traffic entry points and brand identity within the broader ecosystem. By utilizing distributed architecture, Tencent aims to prevent service providers from being bypassed or "commoditized" by the platform owner.
The "lobster" strategy focuses on:
- Adapting AI to rich, niche scenarios rather than general-purpose chat.
- Utilizing decentralized frameworks similar to the existing WeChat Mini Program infrastructure.
- Empowering independent service providers to control their own user interactions and data flow.
- Ensuring a hybrid approach that combines centralized resources with autonomous operation.
Strategic Implications for Web3 and Digital Ecosystems
The move toward a decentralized agent model has significant implications for the Web3 and blockchain sectors, as it aligns with the core principles of sovereign data and distributed networks. The mention of decentralized Mini Programs suggests that Tencent may be looking to provide a bridge between traditional Web2 services and emerging decentralized application (dApp) frameworks. Ma noted that the current development of the WeChat agent is heavily influenced by these considerations, ensuring that partners are not merely "called upon" by a central authority but operate as vital, independent nodes within the network.
This is a longer-term consideration... Agent partners hope to have their own traffic and entry points, but they also do not want to be simply called upon.
By merging centralized efficiency with decentralized autonomy, Tencent is positioning itself to lead the next phase of AI integration. The "raising shrimp" philosophy suggests that scalability in the AI era will depend on a diverse fleet of specialized agents rather than a single point of failure. As Tencent refines this hybrid approach, the industry will likely see increased competition in the development of decentralized AI infrastructures that prioritize the rights and visibility of third-party developers and blockchain-adjacent service providers.
Frequently Asked Questions
Quick answers to the most common questions about this topic.
For educational purposes only. Nothing here constitutes financial or investment advice. Crypto markets are highly volatile — always DYOR before making any decisions.
Fact-checked per our editorial policy. We maintain strict independence from advertisers. Spot an error? Let us know.