摘要
[研究目的]探讨ChatGPT对TCPED路径下情报工作的影响,为优化类ChatGPT生成式人工智能在情报工作的实际应用表现提供启示,促进我国类ChatGPT生成式人工智能模型的构建与优化。[研究方法]以ChatGPT技术迭代为研究起点,分析ChatGPT对情报工作各阶段的影响和ChatGPT技术风险与考验,并以美国情报界现行TCPED全路径情报视角解析ChatGPT对TCPED路径下情报工作的影响。[研究结论]研究表明,ChatGPT在介入TCPED全路径情报工作时仍面临响应深度不足、负面数据危害严重、意识形态渗透等挑战,启示情报机构通过搭建领域特定模型、增加防护能力储备、加强数据前置审查等措施优化类ChatGPT生成式人工智能的使用。
[Research purpose]This paper discusses the influence of ChatGPT on intelligence work under the TCPED path,provides enlightenment for optimizing the actual application performance of ChatGPT-like generative artificial intelligence in intelligence work,and promotes the construction and optimization of ChatGPT-like generative artificial intelligence model in China.[Research method]Taking the iteration of ChatGPT technology as the starting point,this paper analyzes the impact of ChatGPT on various stages of intelligence work and the risks and tests of ChatGPT technology,and analyzes the impact of ChatGPT on intelligence work under the TCPED path from the perspective of the current TCPED full-path intelligence in the US intelligence community.[Research conclusion]Research has shown that when ChatGPT is involved in TCPED full-path intelligence work,it still faces challenges such as insufficient response depth,serious harmfulness of negative data,and ideological penetration.It is suggested that ChatGPT-like generative artificial intelligence should optimize the use of AI-generated content by building domain specific models,increasing protection capacity reserves,strengthening data pre-examination and so on.
作者
张国庆
肖延辉
田华伟
张猛猛
Zhang Guoqing;Xiao Yanhui;Tian Huawei;Zhang Mengmeng(Research Center for Public Security Information,People's Public Security University of China,Beijing 100038)
出处
《情报杂志》
北大核心
2024年第6期79-88,共10页
Journal of Intelligence
基金
国家重点研发计划项目“重要公共场所生物恐怖风险快速评估与联防联控响应机制研究”(编号:2021YFC2600502)
公安部科技计划技术研究项目“从噪声到噪声的无样本设备指纹自学习技术研究”(编号:2022JSYJC22)研究成果。
