ITU 审稿人对“历史连续性 + 现实必要性”的期待。

中文版｜从 5G 标准史到 AI 协议问题的自然过渡

正如 5G 标准形成过程所揭示的那样，全球性基础设施的演进，并非仅由单点技术突破推动，而是依赖于可被反复验证、可被多方采用、并具备长期可扩展性的协议级共识。

当前，人工智能系统正处在一个与 5G 标准化早期阶段高度相似的历史节点。一方面，模型能力的快速提升推动了智能系统在通信网络、云平台与关键基础设施中的广泛部署；另一方面，系统之间在语义表达、行为约束与责任边界上的差异，正在显现出跨平台互操作性不足、可审计性不足以及规模化治理困难等结构性问题。

这些问题并非源于具体模型或算法本身，而是反映出在模型之上、应用之下，尚缺乏一个可被标准化的协议层，用以定义最小一致的语义状态、交互边界与治理约束。正如 5G 的关键挑战并非选择“最先进的单一技术”，而是构建一个能够容纳多种实现路径的统一标准框架，未来 AI 系统的可持续发展同样有赖于在协议层面形成跨组织、跨实现的结构性共识。

English Version｜Transition from 5G Standardization to AI Protocol Challenges

As illustrated by the evolution of the 5G standard, the development of global infrastructure is rarely driven by isolated technical breakthroughs. Instead, it depends on protocol-level consensus that can be repeatedly validated, widely adopted by multiple stakeholders, and sustained over long-term deployment cycles.

Today, artificial intelligence systems are approaching a historical moment comparable to the early stages of 5G standardization. While rapid advances in model capabilities have enabled widespread deployment of AI-enabled services across networks, platforms, and critical infrastructure, increasing disparities in semantic representation, behavioral constraints, and accountability mechanisms are revealing structural limitations in interoperability, auditability, and scalable governance.

These challenges do not arise from any particular model or algorithm. Rather, they point to the absence of a standardizable protocol layer situated between model implementations and application-specific logic—one capable of defining minimal shared semantic states, interaction boundaries, and governance constraints. Just as the success of 5G did not depend on selecting a single “best” technology, but on establishing a unifying framework accommodating diverse implementations, the long-term viability of AI systems similarly requires protocol-level alignment across organizations and deployments.

为什么这一段在 ITU 语境中“非常关键”

你内部要清楚，这一过渡段完成了三件事：

1. 把“AI 协议问题”从技术争论，提升为“基础设施演进问题”
2. 明确说明：问题不在模型，而在协议层缺失（这是 ITU 的天然职责区）
3. 在逻辑上自然承接 5G，而不是生硬类比

换句话说，它在向评审隐含传达一句话：

“我们不是在要求 ITU 为某种 AI 技术背书， 而是在请求 ITU 讨论一个已经出现的结构性空白。”