2.118

影响因子

    高级检索

    超大预训练模型在指挥控制领域的应用与挑战

    Issues and Challenges of ChatGPT-like Large Scale Pre-trained Model for Command and Control

    • 摘要: 超大预训练模型(large scale pre-trained model, LSPTM) 的发展在人工智能领域产生了意想不到的效果, 尤其是在自然语言处理(natural language processing, NLP)上 ChatGPT 的突破, 似乎打通了 “人工智能” 的任督二脉, 在短短的几个月内, 其智能实现了从人类可以理解的智能到无法理解的智能跨越. 以 ChatGPT 为代表的 LSPTM 即将开启一个全新的硅基智能时代, 指挥与控制(command and control, C2) 作为人类社会引以为豪的体现智慧的关键要素, 会面临什么样的机遇和挑战?以 C2 过程的基本范式和运行基本模式框架为指导, 全面分析 LSPTM 在 C2 活动的物理域、信息域、认知域以及社会域各方向潜在的应用, 阐述硅基智能时代, 人工智能从辅助工具角色向伙伴和替代角色的跨越, C2 领域发展的机遇. 军事领域对抗性 C2 的竞争不再局限于技术, 而是培育 LSPTM 的文化底蕴, 东西方文化与价值观的差异将决定两种不同 LSPTM 的智慧与智能.

       

      Abstract: In recent years, the development of large scale pre-trained model (LSPTM)has produced unexpected effects in the field of artificial intelligence, especially the breakthrough of ChatGPT in natural language processing(NLP) , which seems to open up the key to a door of "artificial intelligence", and in just a few months, its intelligence has achieved a leap from human understandable intelligence to incomprehensible intelligence. The LSPTM represented by ChatGPT is about to open a new era of silicon-based intelligence, and command and control (C2) , as a key element of embodying intelligence that human society is proud of, what opportunities and challenges will it face? Guided by the basic paradigm and operation mode framework of the C2 process, this paper comprehensively analyzes the potential application of LSPTM in the physical, information, cognitive and social domains of command and control activities, and explains the leap of artificial intelligence from auxiliary tool role to partner and alternative role in the era of silicon-based intelligence, and the development opportunities of command and control field. The competition of confrontational C2 in the military field is no longer limited to technology, but cultivates the cultural heritage of LSPTM, and the difference between Eastern and Western cultures and values will determine the wisdom and intelligence of two different LSPTM.

       

    /

    返回文章
    返回