Robin Li谈token成本高企：企业换了电机，却没重构工厂
Robin Li 用了电机和工厂的老比喻来聊token烧钱不赚钱的问题，观点很直白：光买算力没用，得把销售、客服这些流程按agent重造一遍。做企业AI落地的人看完会有共鸣。
Robin Li 在 X 上回应一个行业问题：AI token 消耗持续攀升，但收入和利润没有同步增长。他的解释是，企业只是把蒸汽机换成电动机，却仍沿用旧厂房布局——即只采购 AI 算力，而未围绕 AI 重新设计工作流程。
他把 AI 类比为新一轮的"电力"，提出"智能组织"才是 AI 时代的工厂。在这一组织中，生产力提升来自围绕 AI 重构工作方式，具体做法是把 agents 嵌入销售、客服、研发和运营等业务环节。
他给出的判断是：换电机只是起点，重构工厂才是转折点。这一观点对应了当前行业对 AI 投入产出比的普遍质疑，也延续了 Baidu 推动企业级 agent 落地的叙事。
Spending on AI tokens keeps climbing. Why aren't revenue and profits keeping pace?
Robin's answer: companies have replaced the motor without redesigning the factory.
With AI as the new electricity, the intelligent organization becomes the factory of the AI era. There, productivity gains come from redesigning work around AI. That means embedding agents in sales, customer service, R&D and operations.
Replacing the motor is the starting point. Redesigning the factory is the turning point.
This report is categorized under AITOP's "Industry" desk, focusing on China's AI industry and domestic foundation model ecosystem. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Baidu, Robin Li, 智能体, AI商业化.
Source attribution: curated from 百度 AI Baidu (@Baidu_Inc).
