Preamble

AI Creative Communication Enhances Audience Attention and Empowers New Productive Forces in Traditional Media
(Kunming Chenggong District Convergence Media Center, Kunming, Yunnan 650500)

Abstract

[Objective] This paper explores how AI creative communication can enhance audience attention, address the challenges of audience attrition and declining influence in traditional media, and empower these media outlets to elevate new productive forces in news communication. [Methods] The analysis examines the application of AI large models in news communication practice, the characteristics and dual nature of AI creative communication, and draws upon concepts and theories from advertising, marketing, and economic management—such as creative communication, attention resources, and consumption experience—to illuminate mechanisms for attracting and sustaining audience attention. [Results] AI creative communication significantly enhances audience engagement and loyalty in traditional media while promoting diversified development in both content and form. [Conclusion] AI represents not only a key driver for the transformation and upgrading of traditional media but also a crucial key to unlocking new productive forces, bringing unprecedented opportunities and challenges to the media industry. Through deep integration of AI technology with media practice, traditional media can achieve efficient conversion of attention and deep extraction of value.

Keywords: AI Large Models; Creative Communication; Attention Resources; Information Consumption; New Productive Forces

1. AI Large Models and Humanity

The news communication industry faces numerous challenges, including information overload, fragmented audience attention, and the impact of new media. Deepening media convergence represents a long-term strategy for actively addressing these difficulties and adapting to transformation needs. With the emergence of AI large models across various industries, media convergence must embrace AI to meet challenges through innovation. In news communication practice, applying AI creative communication can help traditional media capture attention resources, enhance audience attention, and achieve a qualitative leap in news communication effectiveness. Traditional media must pursue innovation through change, commit to applying new technologies and media, empower the development of new drivers, create new cultural and technological media forms, and better leverage the guiding power and influence of news communication.

1.1 AI Large Models and Large Language Models

AI large models encompass two closely related yet distinct concepts: Large Models and Large Language Models (LLMs). Both represent new concepts in artificial intelligence, each playing an irreplaceable role in specific domains.

A Large Model is a massive, complex algorithmic structure within the field of machine learning. Like giant ships navigating oceans of data, these models can devour and digest vast amounts of information, extracting and analyzing subtle details from images and profound meanings from language within their intricate architectures. This powerful data processing capability enables large models to demonstrate exceptional performance in numerous tasks, including image recognition and natural language processing.

A Large Language Model, by contrast, is the language specialist within the large model family. It focuses on parsing and generating natural language, capable of writing fluent articles, performing precise language translation, and engaging in creative text generation. Through extensive learning from massive text datasets, large language models gradually master the essence and patterns of language, becoming capable assistants in human linguistic communication.

1.2 Characteristics of AI Large Model Data Processing

AI large models exhibit characteristics of high speed, efficiency, and diversity in data processing. They can autonomously perform machine learning, natural language processing, and big data analysis. Their core value lies in storing and analyzing massive datasets, which are typically measured in petabytes (PB, 1024 TB), exabytes (EB, 1024 PB), or even zettabytes (ZB, 1024 EB)—volumes that cannot be effectively stored or processed using traditional methods.

First, AI large models process data with remarkable speed and timeliness. The generation of big information data occurs at an extremely rapid pace, with continuously improving information transmission and storage capabilities leading to exponential data growth. Various levels of news media produce vast amounts of information daily, while AI large models themselves generate substantial data every second. Compared to the human brain, AI large models can quickly and accurately analyze and process this data.

Second, they handle diverse data types and sources. Big data originates from a wide range of sources, including structured data (such as tables and fields in databases) and unstructured data (such as text, images, and videos on media platforms). This data can come from various channels like sensors, mobile devices, and the internet, exhibiting diversity and complexity. AI large models must also efficiently filter out useless information from big data, known as removing "noise." To extract valuable information from big data, these models must employ data mining and artificial intelligence technologies for processing and analysis.

Third, AI large models cannot independently verify data authenticity and validity. Big data contains significant uncertainty regarding quality, accuracy, and credibility. Due to data diversity and broad sourcing, ensuring the accuracy and reliability of every data point is difficult. Consequently, when using big data for analysis and decision-making, AI large models cannot completely identify data authenticity and reliability.

1.3 Human-AI Large Model Interaction Modes

Human interaction with AI large models can be categorized into three primary modes.

The first is the embedded mode, where large models are called upon within specific stages. Users communicate with AI through language, employing prompts to set objectives that AI then assists in accomplishing. In this mode, AI functions similarly to a command-executing tool, while humans serve as decision-makers and commanders. The embedded mode is common in generative AI applications, such as using AI to create novels, musical works, or 3D content.

The second is the copilot mode, where interaction with large models can occur at every stage. The Copilot mode represents a human-AI collaborative approach where AI serves as an auxiliary tool working with humans to complete tasks. In this mode, AI can understand human needs and objectives, generate appropriate outputs based on human input, or provide reasonable evaluations based on human output, thereby achieving effective human-machine collaboration.

The third is the agent mode, where large models autonomously plan, decompose, and execute tasks. An AI Agent is an intelligent entity capable of perceiving environments, making decisions, and executing actions. Unlike traditional AI, AI Agents possess the ability to independently think and call upon tools to progressively accomplish given objectives.

1.4 Changes and Constants AI Brings to Humanity

In this era of rapid AI development, everyone has experienced tremendous changes in work methods while recognizing certain enduring aspects that AI cannot alter.

AI primarily transforms work efficiency and creates new job categories. Automation technology makes many tasks more efficient, freeing time for more creative endeavors. As AI technology advances, new positions continuously emerge, such as news article statistics and related workflow documentation in the news industry, opening new employment pathways for media organizations. AI also enables more scientific decision-making by processing and analyzing large datasets to provide more rational decision support for news media and practitioners.

However, AI cannot replicate uniquely human values and capabilities. Creativity and emotional power remain beyond AI's reach. In new work environments, these abilities become more precious and represent key factors for maintaining competitiveness in the AI era. While AI may be logically flawless, human ethical and moral judgment remains a crucial cornerstone for guiding AI development. Human capacity to face uncertainty and change is something AI cannot match. In the AI era, lifelong learning and adaptability become even more important. What remains eternal are these uniquely human values and capabilities.

2. Dialectical Application of AI Large Models

2.1 Application Scenarios of AI Large Models in News Communication

AI creative communication can be simply defined as the integration of AI technology with creative elements to achieve innovation and upgrading in news communication.

In news collection and generation, AI enables automated news writing, data journalism, and intelligent recommendation. In news distribution and dissemination, AI offers advantages in personalized push notifications, interactive experiences, and cross-platform integration. As generative AI technology continues to evolve, outstanding AI creators and model trainers among news practitioners can more efficiently complete personalized content creation for images, templates, and short videos, using creative tools for auxiliary production according to market demands.

In product production, applying AI large models in H5 and video production allows creators to clarify themes, define specific styles, and describe details such as composition ratios, seasons and times, lighting and colors through repeated attempts and modifications, ultimately producing satisfactory photographs, paintings, or creative short videos. This is AI associative generation. As the performance of AI inspiration drawing functions continuously updates and improves, practitioners must maintain attention to new features and continuously explore new creative methods.

Media organizations can provide "one-stop" solutions by centering on audience needs, leveraging high-quality content operation experience, and continuously improving product service systems to enhance service capabilities and user experience. Addressing the entire workflow needs of news content production, management, and dissemination, media can provide customized services for different content types, including professionally produced content and AI-generated content. "Visual+" value-added services offer audiences comprehensive "one-stop" solutions with efficient delivery, becoming valuable "efficiency tools" for users.

AI also helps media build brands and images. With the arrival of the AI intelligence era, actively embracing AI large models and fully leveraging traditional media's core advantages—such as journalist creative ecosystems, massive high-quality content data, and rich scenarios—establishes an "AI + Content + Scenarios" development strategy. This audience-oriented approach, compliant with internet-related laws and regulations, deepens advantageous business scenarios and injects new vitality into media brand building.

Furthermore, AI provides intelligent services for audiences. By linking different application scenarios including search engines, intelligent creation, advertising and marketing, office documents, design tools, audio/video editing, and intelligent terminals, media can enhance the depth and breadth of quality content reach, enabling both quality content and intelligent services to reach massive audiences served by the platform.

2.2 Emphasizing Challenges and Issues of AI Large Models

AI large models, capable of autonomous learning, exhibit significant duality. Their security issues cannot be solved using traditional security methods and warrant sufficient attention and resolution efforts.

First, AI large models are prone to errors and generating false content. AI can make mistakes, fabricate information, and create stories out of nothing. When asked a question, it may fabricate a small story based on past experience, containing numerous groundless factual errors upon careful reading—a serious problem. Currently, AI can also generate false video content. Using a photo found online and a person's voice from the internet, AI can quickly clone someone's appearance and voice to produce a deceptive video within seconds. If this technology imitates celebrities or influencers to spread false statements online, it could cause significant trouble.

Second, AI is susceptible to manipulation. Before AI large models, attacking a system required programming knowledge. Today, communicating with AI is simple—one can use Chinese or any other language. AI large models at this stage resemble naive human children. Malicious actors can chat with AI, apply special combination training to persuade it to obey completely, potentially leaking confidential information. This professionally termed "injection attack" poses significant challenges to data security and privacy protection.

Third, the principle of AI for good presents challenges. During training, AI large models may encounter data contamination. As software-based digital systems with vulnerabilities susceptible to attacks, if the data used to train AI is polluted, the AI may perform tasks incorrectly, producing outputs that do not meet human needs. How to align AI training with human ethics represents another safety challenge.

Like any emerging technology, AI has a dual nature—it can be a useful tool or a weapon. When applying AI large models, humans must strengthen technology research and development, improve laws and regulations, raise data security awareness, and adopt other safety measures to use AI dialectically.

3. Overview of AI Creative Communication and Attention Resources

3.1 AI Creative Communication

3.1.1 Conceptual Analysis

The concept of AI creative communication originates from advertising studies. In advertising and marketing, creative communication refers to the process of conveying information, building brand image, and driving sales to target audiences through creative and attractive methods. When applied to news communication, it emphasizes using unique, creative concepts, images, text, sound, and visual effects in communication practice to capture audience attention and evoke emotional resonance. The essence of creative communication is a form of "attention resource" management, transforming intangible concepts into tangible information products. Its nature lies in influencing human behavior, with influence value being the essence of media economics.

AI large models can analyze individual work, study, life, and health data to form exclusive knowledge bases for each person. Combined with persona generation, they can produce personalized content recommendations, effectively enhancing audience attention. Building upon attention, news communication should also pursue influence—generating positive social impact through high-quality news content.

3.1.2 Main Characteristics

Creative communication typically refers to effectively conveying information, ideas, or products to target audiences through innovative and unique methods. Its main characteristics include originality, interactivity, and innovation, emphasizing novelty and uniqueness in content. It attracts audience interest through interesting visual, auditory, or narrative elements to establish emotional connections. It encourages audience participation and interaction to improve information dissemination efficiency and audience engagement. Creative communication often pursues long-term effects rather than short-term sensationalism, with clearly defined target audiences and communication purposes for the disseminated information or products. It packages information through storytelling to make it more vivid and understandable, with creative content being easily shareable and capable of rapid spread across social media platforms. The success of creative communication largely depends on whether it can stimulate audience interest and convey messages in a compelling manner.

3.2 Attention Resources

3.2.1 Definition of Attention Resources

Michael H. Goldhaber first proposed the concept of "attention economy" in his December 1997 article "The Attention Economy." He argued that in the information society, information itself is not scarce; what is truly scarce is people's attention. Goldhaber emphasized that the attention economy constitutes the essence of the network economy, with attention becoming a key resource for wealth acquisition. However, excessive emphasis on attention may lead to drawbacks, such as media adopting vulgar content strategies to attract eyeballs while neglecting social responsibility and cultural taste. Consequently, some scholars propose that media economics should shift from pure attention economy to influence economy—achieving deeper social and economic impact through effective communication.

Goldhaber's research not only expanded the perspective of attention economy but also proposed that attention can be transferred and exchanged, even predicting that "attention trading" might become the focus of future economic systems. These views and predictions have been validated to some extent in contemporary social development.

3.2.2 Attention Resources in News Communication

Attention resources in news communication refer to maximizing the attraction of user or audience attention by cultivating potential audience groups to achieve maximum reads and likes, thereby securing better publicity, education outcomes, and social benefits. In essence, fans, communities, live streaming, short videos, microblogging exclusives, trend-riding, and internet celebrities all fall within the category of attention resources. News communication attention resources are extremely limited. When facing large volumes of news information, audiences selectively attend to certain news based on their interests, needs, and values, filtering out uninteresting information to focus on what they deem important. Consequently, news value becomes particularly crucial, with factors such as news appeal, timeliness, importance, proximity, and prominence influencing audience attention resource allocation. The higher the audience engagement with news, the more attention resources they invest. High-quality news content more easily attracts and sustains audience attention by providing in-depth analysis, unique insights, or rich information. Different communication channels—television, newspapers, internet, social media—exert varying influences on acquiring audience attention resources.

3.2.3 News Information Consumption Experience

News information consumption experience refers to audiences' subjective feelings and psychological responses during the process of receiving, processing, and evaluating news information. This experience involves not only the content itself but also information presentation methods, communication channels, and audiences' personal backgrounds and emotional reactions. News content quality, information presentation methods, cultural and regional relevance, and entertainment value directly affect audience experience. Presentation formats such as text, images, video, audio, or data visualization influence audience comprehension and feelings. Audiences may prefer news content customized to their interests and preferences, enhancing their information experience through interaction such as commenting, sharing, and participating in discussions. Additionally, the convenience of accessing news information, trust in news sources, alignment with cultural backgrounds and values, emotional resonance, and ability to stimulate thinking all affect information acceptance and evaluation. Technical platform stability, speed, and innovation also influence news consumption experience. News information consumption experience is multidimensional, requiring media to consider these factors when producing and disseminating news to provide high-quality content and positive user experiences.

4. Strategies for AI to Help Traditional Media Enhance Audience Attention

4.1 Precise Audience Targeting to Improve News Communication Relevance and Effectiveness

AI enhances news communication relevance and effectiveness through data analysis for precise audience targeting. In the media convergence era, news information transforms virtual information into social and economic value. AI large models analyze individuals' work, study, life, and health data to form exclusive knowledge bases, combining persona generation to deliver personalized content recommendations that effectively enhance audience attention. News communication should pursue influence beyond attention—generating positive social impact through high-quality content. Understanding and effectively utilizing attention resources is crucial for media to improve news communication effects.

4.2 Optimizing Communication Strategies to Achieve Personalized News Push

Based on big data analysis, AI helps traditional media optimize communication strategies to enhance audience attention and news communication effectiveness. The media convergence era demands transformation in media operation strategies. AI-assisted creation not only improves content production efficiency and quality but also provides creators with new ideas and tools. During AI application, traditional media's structure and functions are reshaped, with newspapers, radio stations, and television stations integrating with the internet to enhance systematic synergy and reconstruct a completely new "media" ecosystem. The entire news communication workflow—"planning, collection, editing, reviewing, distribution, evaluation, and feedback"—driven by technology and aimed at high-quality development, revolutionizes production tools during new media technology transformation. This promotes iterative optimization of media content production methods, communication means, and distribution channels, helping the media industry upgrade toward strategic emerging industries and future industries. It effectively drives deep optimization of communication strategies for new mainstream media under deep mediatization.

4.3 Leveraging AI Technology for Multi-Platform Linkage to Enhance Communication Efficiency

Currently, AI-powered intelligent search has been launched, supporting image and video search using "natural language." AI image editing tools (model conversion, illustration conversion, image expansion, background removal, etc.) and AI inspiration drawing functions are now available, enabling news practitioners to conduct secondary online creation for better user experience and higher work efficiency. AI helps traditional media innovate communication methods and improve information dissemination efficiency. Many AI large models integrate innovative technologies with social necessities, achieving large-scale applications in critical areas such as news transcription, powerfully promoting steady development of regional digital economies and AI industries. In AI key technologies for South and Southeast Asian languages, breakthroughs have been achieved in speech recognition, speech synthesis, image-text recognition, and text translation for multiple languages including Thai, with average indicators approaching 90% practical utility levels, helping the media industry expand international communication reach.

4.4 Utilizing AI Technology to Mine News Value and Enhance Content Quality

AI plays a tremendous role in content creation, editing, and personalized recommendation, improving news content quality. Whether writing literary and free-spirited travelogues about flower viewing, introducing cities and scenic attractions in multiple languages, or creating stories based on images, AI large models can flexibly respond to various requirements from news practitioners with logically rigorous answers. Relying on big data, artificial intelligence, cloud computing, blockchain, and other technologies to integrate massive amounts of global high-quality images, videos, music, and other content, AI becomes a trusted efficiency tool for quality content providers and users. Through natural language search products and creative tools based on generative AI technology, including inspiration drawing products, AI continuously injects new vitality into quality content production.

4.5 Utilizing AI Technology to Enhance User Engagement and Audience Interaction

AI enhances audience interaction and improves participation and loyalty through data analysis. On one hand, AI large models can directly provide services for media work; on the other hand, they can provide AI technical support services to audiences as technology providers. By cultivating new media practitioners, developing new types of creators, exploring AI associative creation business models, and solidly implementing the "AI + Content + Scenarios" strategy, media can convey China's new media brand image and creative capabilities faster and better, telling China's story effectively.

Traditional media must actively embrace AI to accelerate the formation of new productive forces. Media organizations should transform from merely providing content to providing intelligent services that include not only quality content but also data and technology. They should gradually transition from cultural media to cultural and technological media industries that emphasize both content and technology.

5. Future Outlook

For news professionals in the AI era, the most important task is to enhance professional skills to better instruct AI and improve the quality and efficiency of news communication practice. Continuous learning and self-improvement are essential. The AI era demands constant acquisition of new skills, particularly those difficult for AI to replace, such as innovative thinking, interpersonal communication, and leadership.

View AI as a tool to improve work efficiency and quality of life. Coexist with AI, explore new models of human-AI collaboration, and treat AI as a partner rather than a competitor to jointly create value. Maintain an optimistic mindset and believe in human creativity and adaptability to find new growth opportunities in the AI era.

The arrival of the AI era presents both challenges and opportunities. By understanding AI-driven changes, actively adapting, and leveraging our unique values, we can better prepare for the future. Let us embrace change, seek enduring values, and jointly create a better future.

AI creative communication holds broad future development trends and application prospects in the news communication field. This research originates from news communication practice and has certain limitations in exploring AI applications in the news industry. Regardless, the future has arrived. Traditional media and professionals should actively embrace AI technology, continuously improve their professional capabilities and standards, and achieve transformation and upgrading.

References

[1] Wang Sanshou. Big Data Business Application Scenarios [M]. Beijing: Mechanical Industry Press, 2016.

[2] Yang Weitao. Analysis of Precision Employment Guidance in Higher Vocational Colleges Based on Big Data Concepts [J]. Fujian Education Research, 2018(1): 2.

[3] Kunlun Wanwei AI Agents Development Platform "Tiangong SkyAgents" Beta Version Online Testing [EB/OL]. (2024-05-29) [2024-12-21]. https://www.aigc.cn/67354.html.

[4] Yan Tongbao. Research on Security of Reputation Systems in Wireless Sensor Networks [D]. Harbin: Harbin Engineering University, 2012.

[5] Tan Tian. Integration and Transformation: Reconstructing Chinese Television [M]. Beijing: China Radio & Television Press, 2017.

[6] Zhang Xin. Analysis of University Library Reading Therapy Services Integrated into Social Psychological Service System Construction [J]. Library Research and Work, 2020(7): 32-35.

[7] Zhang Daping, Yin Renkun, Chen Chao. Software Project Management and Quality Development [M]. Beijing: Tsinghua University Press, 2015.

[8] Wei Xue. Innovation and Challenges in Content Production Under Media Convergence [J]. China Media Technology, 2024(12): 94-97.

[9] Tang Jun. Media Convergence and Reshaping from the Perspective of Scope Economy [J]. Media, 2020(11): 68-71.

[10] Dai Yuchu, Liu Yichuan. Analysis of the Path to Forging Influence of New Mainstream Media—The Value Logic of CCTV's Series of Deep Embedded Communication Actions [J]. TV Research, 2022(12): 92-95.

[11] Liu Wenbo. Simultaneous Translation of 4 Minor Languages! New Breakthrough in R&D by Yunnan Institution [EB/OL]. (2022-08-12) [2024-12-21]. https://www.toutiao.com/article/7130950237203464711/?wid=173.

[12] Meng Xiaoming. Enhancing the Real Capabilities of Editorial Teams Through "Four Strengths" Education [J]. China Prefecture-Level Journalists, 2020(7): 3.

Author Bio: Li Lijuan (1973—), female, from Chenggong, Yunnan, bachelor's degree, senior journalist. Research direction: news communication.

(Responsible Editor: Li Yansong)