Zhou Hongyi's Speech at Zhongguancun Forum: Intelligent Agents Reshape Industry Landscape; Six Key Areas Nurture New Unicorns

If you want to trade stocks, just rely on the Jinqilin analyst reports—authoritative, professional, timely, and comprehensive—helping you uncover high-potential theme opportunities!

（Source: Leifeng.com）

At the 2026 Zhongguancun Forum Annual Conference Global Unicorn Enterprises Conference held recently, Zhou Hongyi, founder of 360 Group, delivered a keynote speech. Centering on the new-generation agent technology represented by “Lobster” (OpenClaw), he systematically laid out the opportunities for industrial transformation brought by the accelerated evolution of artificial intelligence. He pointed out that the sudden rise of “Lobster” in 2026 marks the successful “crossing over” of agent technology—from the geek community to the general public. This not only completes mass adoption of agents, but will also drive a comprehensive reshaping of the entire internet infrastructure, the software industry, and even all related sectors of the real economy, where there are enormous opportunities for nurturing the next generation of unicorns.

First is the reshaping of AI-native infrastructure. Zhou Hongyi noted that today’s internet ecosystem is mainly built around people’s interaction habits. In the age of agents, browsers, search engines, databases, and even account systems all need to transition to “serving intelligent agents.” The very process of comprehensively rewriting and reconstituting the underlying infrastructure framework will, in itself, nurture new platform-style companies. Second is the evolution of the software industry toward “atomization.” Traditional software that clings to existing models will be hard to sustain; it is shifting from an end-to-end product form toward being foundational and modular—becoming “skill components” that intelligent agents can flexibly call. Under this model, software is like “raw materials,” while agents are “chefs” and dispatchers that can dynamically assemble, or even rewrite, functions according to needs. Without enterprises having to redo their software, service upgrades can be achieved—completely changing the logic of software sales and usage—thereby reshaping enterprise service forms and giving rise to a new generation of software giants. Third is the rise of “custom agents” in vertical sectors. Faced with the stringent requirements of enterprise work for stability, consistency, and determinacy, today’s general-purpose agents (such as general “Lobster”) still lean toward tools for geeks, making it difficult to directly meet complex standardized, process-driven business scenarios. At the same time, enterprises generally lack the capability to build proprietary agents, which means that the customized business for specialized agents and vertical-industry agents has a huge market. This trend will accelerate the rollout of agents in fields such as finance, manufacturing, and government affairs.

Fourth is the emergence of new organizational models such as “one-person companies (OPC).” As agents become widespread as “silicon-based partners” and “digital employees,” it will drive the flourishing of the “one-person company” model. Entrepreneurs can complete complex tasks by directing “digital teams” composed of multiple agents, achieving higher efficiency with fewer people. In the future, the headcount of unicorn enterprises may be reduced by 10 times compared with today. With this lightweight, high-output organizational model, business innovation will reshape the basic logic of innovation and entrepreneurship. Fifth is the formation of a “machine economy” ecosystem driven by agents. Zhou Hongyi said that in the future, agents will act as a “new species” on the internet and gradually gain independent account, communications, and even payment systems, enabling them to replace humans in automatically handling information processing and business decision-making, as well as transaction performance. As large numbers of agents coordinate with one another, an exclusive agent-economy ecosystem is expected to take shape quickly.

Sixth is the coordinated upgrade of underlying industries such as computing power and energy. Always-on agent applications will trigger a hundredfold-level growth in compute consumption. This will promote the separation of inference computing from training computing, directly giving rise to an independent “inference chip” industry. Meanwhile, the massive demand for compute will further be transmitted to the electricity and energy systems, unlocking trillion-level markets across power supporting industries and upstream and downstream sectors such as green electricity and nuclear power, forming new growth momentum. Zhou Hongyi said that artificial intelligence is moving from point breakthroughs toward system-level reshaping, and agents will become an important vehicle connecting technological innovation with industrial upgrading. In the future, 360 will continue to increase investment in the Zhongguancun artificial intelligence track, and will deeply develop subfields such as humanoid robot R&D and AI services. By promoting deep integration of technology and industries and fostering an innovation ecosystem with an open-and-win mindset, it will fully empower individuals and organizations, and work together to incubate more super enterprises of the intelligent age—driving the emergence of a new round of unicorns at an accelerated pace.

Massive information, precise interpretation—available on the Sina Finance APP