网络通信技术(如网络带宽、芯片内的晶体管数量和软件代码数等)经过几十年的指数级增长之后，在未来二十年前后会碰到七大“技术墙”，包括移动通信技术墙、IP数据通信技术墙、光纤通信技术墙、集成电路技术墙、软件复杂性技术墙、大数据处理技术墙和能耗技术墙。这七个“技术墙”相对独立、相互影响又自成体系。翻越每一个“墙”的难度巨大，而同时翻越七个墙的难度就更可想而知。由此引发我们做出战略性和全局性思考，除按原来的方式继续发展(趋缓、改变增长方式或用新模式新理论新方法)外，是否还有另外的路径可走，而不用翻每一道墙。从未来发展看，这一路径可能是网络通信的人工智能化，其会逐渐逐步地发展几十年，从低级、中级、高级到超级，作为一种可能的选项用以弥补“带宽等无法达到指数级增长”之不足。文章最后简述了人工智能(AI)对网络通信产生的影响，并对网络智能体趋势做了初探。

This paper systematically and roundly proposes the seven “technical walls” of network communications technologies we face after decades of exponential growth of the network bandwidth, number of transistors in the chip and number of software codes, including the “technical wall” of wireless mobile communication, IP data communication, fiber communication, Integrated Circuit (IC), software complexity, big-data processing and energy consumption. These “technical walls” are relatively independent, interactive and itself- systematic. As you know, it is very difficult to turn over a “wall”, and it is more difficult to turn over the seven “walls” at the same period. This leads us to think strategically and globally. In addition to developing in the original way (“the trend is becoming smoother”, “changing growth patterns”, or “using new mode, new theory and new method”), Is there another way to go? Instead of turning to every wall. From the perspective of future development, this path may be Artificial Intelligence (AI) of network communications, will gradually develop for decades or more, from low, intermediate, advanced, to super, so as to compensate for the lack of bandwidth without reaching exponential growth law as a possible way and option. At last paragraph, the network communications may become intelligent and the effect of AI is also given as “network AI body” to solve the operational problem and to greatly raise efficiency uninterruptedly for decades or more.