Empowering the Dissemination of the Party’s Innovative Theories with Generative Artificial Intelligence

Under the dual wave of information explosion and technological iteration, the dissemination of the Party’s innovative theories is undergoing a transformation from “broad coverage” to “precise reach.” The rise of generative artificial intelligence rapidly ushers us into the intelligent era. Compared with traditional data processing technologies, generative AI can not only integrate vast amounts of information but also, through natural language processing, affective computing, deep learning, and other technologies, create a “three-dimensional portrait” of the audience. From static characteristics such as age, occupation, and educational background, to dynamic characteristics such as reading habits, topics of interest, and value tendencies, and further to cognitive characteristics such as preferences for accepting theoretical content and points of comprehension difficulty, it forms a multi-dimensional, iterable cognitive model of the audience. By precisely delineating audience portraits, generative AI can lower the threshold of understanding using language familiar to the audience, evoke resonance with cases close to their lives, and promote practice with concrete and actionable guidance, thereby achieving a progression from knowing the theory to identifying with it to practicing it. This effectively resolves the “last mile” problem of theoretical dissemination.

在信息爆炸与技术迭代的双重浪潮下，党的创新理论传播正面临从“广覆盖”向“准触达”的时代转型。生成式人工智能的崛起将我们迅速带入智能时代。与传统数据处理技术相比，生成式人工智能不仅能整合海量信息，还能通过自然语言处理、情感计算、深度学习等技术为受众“立体画像”，从年龄、职业、教育背景等静态特征到阅读习惯、关注话题、价值倾向等动态特征，再到对理论内容的接受偏好、理解难点等认知特征，形成多维度、可迭代的受众认知模型。生成式人工智能通过精准勾勒受众画像，能够用受众熟悉的语言降低理解门槛，用贴近生活的案例引发共鸣，用具体可行的指引推动实践，实现从知晓理论到认同理论再到践行理论，从而有效解决理论传播“最后一公里”难题。

Use intelligent technology to vividly depict audience portraits. The data sources for audience portraits include both “explicit data,” such as users’ registration information, reading duration, and content bookmarks on theoretical learning platforms, and “implicit data,” such as users’ discussion topics on social media, emotional tendencies in comment sections, and search frequency for theoretical keywords. Generative AI extracts opinion tendencies from text through natural language processing technology, analyzes users’ focus on visual content with image recognition technology, and judges the audience’s emotional feedback on theoretical interpretations through affective computing technology. This multi-dimensional integration breaks the limitation of “single labels,” moving audience portraits from “flat” to “three-dimensional.” Dynamic generation and rapid iteration enable audience needs to shift from “static labels” to “real-time adaptation.” Audience needs change dynamically with social hotspots, policy advancements, personal experiences, and other factors. The advantage of generative AI lies in its timely and powerful continuous learning capability, which can capture new data in real time and continuously optimize portrait accuracy. This “dynamic iteration” mechanism avoids the drawbacks of traditional audience portraits, keenly captures such changes, and promptly adjusts portrait labels, ensuring that the customization of theoretical content is “neither lagging nor misaligned,” always closely connected to the audience’s current concerns.

用智能技术生动刻画受众画像。受众画像的数据来源既包括用户在理论学习平台的注册信息、阅读时长、内容收藏等“显性数据”，也包括用户在社交媒体中的讨论话题、评论区的情感倾向、对理论关键词的搜索频率等“隐性数据”。生成式人工智能通过自然语言处理技术从文本中提取观点倾向，借助图像识别技术分析用户对可视化内容的关注焦点，依托情感计算技术判断受众对理论解读的情绪反馈。这种多维度整合打破了“单一标签”的局限，让受众画像从“平面化”走向“立体化”。动态生成与飞速迭代可实现受众需求从“静态标签”到“实时适配”。受众需求会随着社会热点、政策推进、个人经历等因素动态变化，生成式人工智能的优势在于其及时且强大的持续学习能力，能够实时捕捉新数据不断优化画像精度。这种“动态迭代”机制避免了传统受众画像的弊端，敏锐捕捉这种变化并及时调整画像标签，确保理论内容的定制“不滞后、不错位”，始终与受众的当下关切紧密相连。

Adhere to a people-centered dissemination orientation. Only by always prioritizing the thoughts, concerns, and aspirations of the masses, and taking the masses’ cognitive patterns and reception habits as important considerations in dissemination work, can the Party’s innovative theories truly enter the hearts of the people. Generative AI plays a key role in this process. The process of constructing audience portraits is essentially a practice of using technological means to deepen the listening to the voices of the masses. Through precise analysis of massive data, it keenly captures the real needs of the masses at different stages, clearly presents the cognitive characteristics and reception preferences of different groups, enabling theoretical dissemination to truly “shoot the arrow at the target,” resonate with the masses, and become an ideological weapon that the masses can understand, are willing to accept, and can apply.

坚持以人民为中心的传播导向。始终将人民群众的所思所想所盼放在首位，把群众的认知规律、接受习惯作为传播工作的重要考量，才能让党的创新理论真正走进群众心里。生成式人工智能在这一过程中扮演着关键角色，其构建受众画像的过程本质上是运用技术手段深化倾听群众声音的实践，通过对海量数据的精准分析，敏锐捕捉群众在不同阶段的真实需求，清晰呈现不同群体的认知特点和接受偏好，让理论传播真正“有的放矢”、与人民群众同频共振，成为群众能理解、愿接受、可运用的思想武器。

Explain theories in the “language” familiar to the audience. The Party’s innovative theories are highly scientific and systematic, but their forms of expression need to “vary by group.” Generative AI can adjust and adapt expressions according to audience characteristics. For example, for youth groups, it can generate lightweight content such as “theory + animation”; for entrepreneur groups, it can produce in-depth reports like “policy interpretation + case analysis”; for elderly groups, it can push graphic content such as “theory + changes around them.” This “thousand faces for thousand people” expression transforms abstract theories into “matters close at hand” closely related to the audience’s lives, maintaining the seriousness of the theory while enhancing its affinity. Generative AI can also innovate content by combining different scenarios. The core of scenario-based content is the connection point between theory and life. The scenario-based dissemination of theories can significantly improve their acceptance. Generative AI can generate related content based on the audience’s daily scenarios. For instance, for teacher groups, it can combine their work scenarios of classroom teaching and home-school communication to generate micro-videos such as “The Embodiment of Quality Education in Teaching Practice,” using cases familiar to them, like “sharing high-quality lesson examples,” to explain educational concepts. This method of explaining theories within scenarios helps the audience understand the connotation of the theory in a familiar context.

用受众熟悉的“语言”讲清理论。党的创新理论具有高度的科学性和系统性，但其表达形式需要“因群而异”。生成式人工智能可以根据受众特点调整适配表达。例如，针对青少年群体，可生成“理论+动漫”等轻量化内容；对企业家群体，可制作“政策解读+案例分析”的深度报告；对老年群体，可推送“理论+身边变化”的图文内容。这种“千人千面”的表达，让抽象理论转化为与受众生活紧密相关的“身边事”，既保持了理论的严肃性，又增强了亲和力。生成式人工智能还可以结合不同场景进行内容创新。场景化内容的核心是理论与生活的连接点，理论传播的场景化能显著提升其接受度。生成式人工智能可根据受众的日常场景生成关联内容，如针对教师群体，可结合其课堂教学、家校沟通的工作场景生成“优质教育在教学实践中的体现”等微视频，用其熟悉的“优质课例共享”等案例阐释教育理念。这种在场景中讲理论的方式，有助于受众在熟悉的语境中理解理论的内涵。

Deliver theories to the audience’s “fingertips.” There are significant differences in the information acquisition channels of different groups, and precisely matching channel preferences is a prominent highlight of generative AI. Relevant research shows that middle-aged and elderly groups prefer WeChat public accounts and TV news; youth groups are active on Douyin, Bilibili, and Xiaohongshu; professionals pay more attention to industry journals and academic platforms. By analyzing the audience’s channel usage data, generative AI can achieve precise matching between channels and groups. For example, it can push in-depth articles like “The Party’s Innovative Theories and Discipline Construction” to university teachers; use learning apps to push interactive games like “20th National Congress Knowledge Challenge” to middle school students; and push H5 content like “Rural Revitalization Policy Implementation Guide” to township cadres through government affairs apps. This channel adaptation based on intelligent technology helps ensure that theoretical content “appears at the right time and in the right place.” The precise dissemination of generative AI is also reflected in its ability to grasp the “golden timing” of dissemination. For example, office workers are more receptive to short content during morning commutes and lunch breaks; student groups are more active in the evenings and on weekends; and farmer groups show higher attention to theoretical learning during slack farming seasons. Staggered push notifications can avoid the audience’s busy periods and, by combining with their psychological states during specific times, improve the “open rate” and “absorption rate” of theoretical content.

把理论送到受众“指尖上”。不同群体的信息获取渠道存在显著差异，而精准匹配渠道偏好是生成式人工智能的突出亮点。相关研究表明，中老年群体偏爱微信公众号、电视新闻；青少年群体活跃于抖音、B站、小红书；专业人士更关注行业期刊、学术平台。生成式人工智能通过分析受众的渠道使用数据能实现渠道与群体的精准匹配。例如，向高校教师推送“党的创新理论与学科建设”等深度文章；用学习类App向中学生推送“二十大知识闯关”互动游戏；通过政务App向乡镇干部推送“乡村振兴政策落地指南”的H5等。这种基于智能技术的渠道适配有助于确保理论内容“适时出现、适地出现”。生成式人工智能的精准传播，还体现在其善于把握传播的“黄金时点”上。例如，上班族在早间通勤、午休时段更易接受短内容，学生群体在晚间、周末更活跃，农民群体在农闲时段对理论学习的关注度更高。错峰推送既能避开受众的繁忙时段，又能结合其特定时段的心理状态提升理论内容的“打开率”和“吸收率”。

Form an intelligent dissemination closed loop of “dissemination – feedback – iteration.” The improvement of dissemination effectiveness cannot be separated from precise feedback. Generative AI evaluates the audience’s acceptance of customized content in real time through multi-dimensional data: judging content appeal from “behavioral feedback” such as reading completion rate, forwarding volume, and bookmarking volume; analyzing the degree of identification from “emotional feedback” such as comment section sentiment and bullet chat content; and assessing the depth of understanding from “cognitive feedback” such as online quiz accuracy and search volume for theoretical keywords. Dissemination strategies are also dynamically adjusted based on feedback. The value of feedback lies in “iterative optimization.” Generative AI can generate dissemination effectiveness reports based on feedback data and propose specific adjustment suggestions: simplifying expressions and adding cases for content with high comprehension difficulty; changing formats with low appeal to short videos or interactive Q&A; and adjusting distribution platforms for cases of channel mismatch. This mechanism of “promoting optimization through feedback” helps drive theoretical dissemination to form a virtuous cycle of precise delivery, effectiveness evaluation, and continuous improvement.

形成“传播－反馈－迭代”的智能传播闭环。传播效果的提升离不开精准反馈。生成式人工智能通过多维度数据实时评估受众对定制内容的接受度，从阅读完成率、转发量、收藏量等“行为反馈”判断内容吸引力，从评论区情绪、弹幕内容“情感反馈”等分析认同程度，从在线答题正确率、理论关键词搜索量等“认知反馈”评估理解深度。传播策略也会依据反馈情况动态调整。反馈的价值在于“迭代优化”，生成式人工智能可根据反馈数据生成传播效果报告并提出具体调整建议，对理解难度高的内容简化表述、增加案例，对吸引力不足的形式改为短视频、互动问答，对渠道错配的情况调整分发平台。这种“以反馈促优化”的机制，有助于推动理论传播形成精准投放、效果评估、持续改进的良性循环。

The “exclusive customization” dissemination empowered by generative AI is essentially a new exploration in the intelligent era to advance the popularization of the Party’s innovative theories. It is not a negation of traditional dissemination methods but a way to make theoretical dissemination more precise, vivid, and considerate through new technological means. This exploration not only helps enhance dissemination effectiveness but also highlights the vitality of the Party’s innovative theories. As long as we always adhere to the principle of disseminating for the people and innovate with technological empowerment, we will surely enable the Party’s innovative theories to radiate stronger guiding, cohesive, inspiring, and influential power in the intelligent era.

生成式人工智能赋能的“专属定制”传播，本质上是推进党的创新理论大众化在智能时代的新探索。它不是对传统传播方式的否定，而是通过新的技术手段让理论传播更精准、更生动、更贴心。这种探索不仅有助于提升传播效能，更能彰显党的创新理论的生命力。只要我们始终坚守为人民传播之正，创技术赋能之新，就一定能让党的创新理论在智能时代焕发出更强大的引领力、凝聚力、感召力、影响力。

(Author: Zhang Wenjun, specially appointed researcher at the Beijing Xi Jinping Thought on Socialism with Chinese Characteristics for a New Era Research Center, and Deputy Director of the Political Science Teaching and Research Department, Party School of the Beijing Municipal Committee of the CPC [Beijing Academy of Governance])

（作者：张文君，系北京市习近平新时代中国特色社会主义思想研究中心特约研究员，中共北京市委党校〔北京行政学院〕政治学教研部副主任）

Online Editor: Tongxin

网络编辑：同心