Introduction
In recent years, the technology of Artificial Intelligence Generated Content (AIGC) has shown great potential in various media fields such as news editing, virtual anchors, and documentary production. This article focuses on the practical application of AIGC technology and discusses how county-level converged media centers can innovate personalized content and improve content production efficiency using AIGC technology, further analyzing the specific implementation path of AIGC technology in county-level converged media centers.
County-level converged media centers aim to integrate local radio, television, newspapers, and new media resources, providing diversified converged media services including media, party-building, government affairs, public services, and value-added services. In August 2018, General Secretary Xi Jinping delivered an important speech at the National Propaganda and Ideological Work Conference, emphasizing the need to solidly promote the construction of county-level converged media centers to better guide and serve the masses. In recent years, county-level converged media centers have rapidly developed across the country to bridge the “last mile” of information services. However, due to differences in economic, resource, and technological levels across regions, some county-level converged media centers face numerous challenges, with the key to solving these problems lying in unleashing and developing digital productivity. AIGC (Artificial Intelligence Generated Content) is a technology that generates relevant content by learning and recognizing existing data patterns and has a moderate ability to generalize. This technology can generate various forms of content, including text, images, audio, and video, and can achieve conversion and combination between different modalities. The launch of ChatGPT by OpenAI at the end of 2022 marks the arrival of a new era of AIGC and is regarded as an important turning point in the development of artificial intelligence. Initially, ChatGPT 3.5, as a single language model, could not handle complex tasks such as images and videos. However, OpenAI’s release of the multimodal large language model GPT-4 in March 2023 addressed this limitation, enabling applications in news production such as image processing, audio generation, and video editing. With the continuous transformation and development of AIGC technology, the internationally renowned market research firm Gartner has rated it as one of the “most important strategic technology trends of 2022”. In its latest research report, Gartner predicts that AIGC will become the mainstream of online content, expected to account for 30% of the total online content by 2025. The rise of AIGC technology marks a shift in AI from a traditional passive role to a more proactive role of “generation and creation”, providing new technological momentum for the generation and updating of digital content. In the context of integrated intelligence, county-level converged media centers should regard AIGC as a key development direction. By utilizing this new technology, county-level converged media centers can further achieve cost reduction, efficiency improvement, and industrial innovation, thereby enhancing news production efficiency and influence, and more effectively serving the masses.
1AIGC Technology Empowers the Development of the News Industry (1) Application of AIGC Technology in Text Production AIGC technology’s application in text production marks a significant transformation in news production methods, evolving from basic text generation algorithms to advanced automated news writing systems. Since 2018, the Associated Press has begun using the generative AI tool Wordsmith to automatically generate sports news and financial reports, having automatically written over 50,000 articles to date. Reuters has also implemented an AI news writing tool called Lynx Insight Service to assist journalists in data analysis, story ideation, and automatic report generation. In China, Xinhua News Agency developed a news robot named “Quick Pen Xiao Xin” that can perform tasks such as data collection, processing, automatic writing, and editing, achieving multi-step automated news production. AIGC technology not only increases the speed and efficiency of news reporting but more importantly frees up more time and energy for journalists and editors, allowing them to engage more deeply in news analysis and in-depth reporting, thus enhancing the value chain level of the news industry. (2) Application of AIGC Technology in Video Editing AIGC technology, through the application of machine learning and artificial intelligence algorithms, makes the video editing process more efficient and intelligent, providing powerful support tools for video content creators and media organizations. During the 2022 Beijing Winter Olympics, China Central Television successfully utilized an AI intelligent automated editing system to rapidly process a large amount of event resources, achieving automatic editing of key event moments and mass production of short video content, providing a more vivid real-time event experience for media and audiences. In addition, Xinhua News Agency collaborated with Xinhua Zhiyun to develop an AI platform called “Media Brain” which, supported by AIGC technology, provides editors with convenient tools for quickly selecting shots and generating content, allowing for one-click rapid publishing to multiple platforms. This technology not only simplifies workflows but also significantly accelerates content dissemination speed, providing media organizations with more efficient content production and distribution solutions. The widespread application of AIGC technology in video editing has become an important support tool for short video content creators and media organizations, gradually developing towards intelligent video editing, automated content generation, and more personalized video production. (3) Progress in the Application of AIGC Technology in Digital Humans Utilizing key technologies such as natural language processing, speech synthesis, and speech recognition, AIGC digital humans have made significant progress in the field of broadcasting and hosting. On January 22, 2018, the world’s first large-scale documentary “Innovative China” that utilized artificial intelligence to simulate human voice for dubbing was aired. This project used existing dubbing materials from voice actor Li Yi, employing big data collection and algorithm construction to ultimately achieve a digital dubbing effect nearly indistinguishable from a real human voice. In 2019, the digital twin AI host “Little Xiao Sa” created for the CCTV online Spring Festival Gala was supported by American AI company ObEN. Not only does it resemble Sa Beining in appearance, but it also greatly meets audience perceptions in dimensions of language, expression, and action. AIGC digital human technology has made significant advancements in news broadcasting, with future developments potentially encompassing higher levels of personalization and emotional expression, enabling AI digital humans to be more vivid and engaging when delivering news. (4) Development of AIGC Technology in Voice Applications In recent years, with the development of machine learning and neural networks, AIGC technology has rapidly progressed from basic text to speech synthesis (TTS) technology, and now to highly humanized voice generation, achieving rapid advancements in the naturalness of mimicking human voice expression, emotional richness, and adaptability to different contexts. Taking iFLYTEK’s SMART-TTS system as an example, this system can generate clear and smooth speech, and its tone variation, sentence pauses, and voice delicacy approach that of a real person. Additionally, the Himalaya App utilizes AIGC technology to analyze text context and emotions, creating voice models that adapt to various emotions and styles, significantly enriching the audience’s experience in news, novels, finance, and other content production. The music video “Rise” created by Xinhua News Agency’s digital reporter Xiao Zheng showcases the deep application and innovative potential of AI in audiovisual media. AIGC technology not only provides more possibilities for the creation of voice content but also offers strong technical support for producing personalized and customized voice content. (5) Breakthroughs in AIGC Technology in Video Generation The application of AIGC technology in video generation initially focused on simple video editing and effect enhancement. With the increase in computational power and algorithm advancements, AIGC technology has now been able to generate complex and high-quality video content directly from text, even transforming static images into dynamic videos. In September 2022, Meta launched the “Make-A-Video” AI model that can generate videos based on short text descriptions or create new content based on existing videos using advanced deep learning techniques. Google’s Phenaki model demonstrated the capability to generate videos up to 2 minutes long based on script prompts, proving its technological advancement and providing new possibilities for storytelling and creative expression. Domestically, Baidu’s VidPress intelligent video synthesis platform combines natural language processing technology to automatically convert text and image content into video, showcasing the application of the Wenxin large model in automated video content generation. As AIGC technology continues to mature and application scenarios expand, future video content generation will become more intelligent.
2AIGC Technology Empowers the Innovative Development Path of County-level Converged Media Centers (1) Restructuring Content Production Models to Enhance News Output Efficiency The traditional content production model of county-level converged media centers is mainly based on integrating traditional media resources such as radio, television, newspapers, and websites within the county, providing news information, cultural entertainment, and public service content through the processes of collection, editing, and publishing. However, due to issues related to personnel, technology, and funding, some county-level converged media centers struggle to meet user demands in terms of both quantity and quality of news reporting. To address this predicament, county-level converged media centers can leverage AIGC technology to change the news production model, thus achieving innovation in content production. AIGC possesses a powerful database and can serve as an auxiliary tool in the news production process, providing journalists with leads and material collection, and even generating automated news content, especially demonstrating significant advantages in formulaic reports such as finance and sports, effectively improving news production efficiency. For example, the AIGC artificial intelligence editorial department launched by CCTV in December 2019 can quickly capture trending topics and news leads across the internet, predict the evolution trend of news hotspots, assist editors in accurately grasping diverse audience interests, conduct dissemination trend analysis, and achieve efficient and high-quality material selection and writing. When writing news, staff at county-level converged media centers can utilize AIGC tools such as GPT-3.5 or GPT-4 for automatic writing. The trainable modular characteristics of AIGC allow it to create specific platform writing styles by analyzing a large number of news articles from different platforms. After training, by providing keywords, interview materials, and themes, AIGC can directly generate news that meets the needs of users on various platforms. For instance, if a county-level converged media center analyzes trending news data on Weibo, it can train an AIGC model suitable for Weibo news production, and the same method applies to news production models for WeChat public accounts, apps, and other diverse platforms, aiming to achieve the creative effect of “one collection, multiple generations, and diversified dissemination” mentioned in the “central kitchen” concept. In addition to written news, AIGC is also crucial in the automated generation and editing of visual content, primarily applied to video auto-editing and automated video content generation. In video auto-editing, AIGC technology can analyze raw videos, automatically identify and select key frames or highlights for video editing; in automated video content generation, AIGC technology can utilize deep learning models such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs) to learn from vast amounts of existing video data to create entirely new, seemingly realistic video content. As early as 2017, Xinhua News Agency collaborated with Xinhua Zhiyun to develop and launch a platform called “Media Brain” that uses AIGC tools to assist editors in quickly locating video shots, efficiently extracting highlights, and achieving rapid content production and publishing through automated processes. Nowadays, ByteDance’s “Jianying” and Kuaishou’s “Yun Jian” also reflect the immense potential of AIGC in video creation, not only automatically selecting key frames for editing and optimization using AI algorithms but also utilizing natural language processing (NLP) technology to analyze text descriptions and convert them into corresponding video scenes, truly achieving “one-click editing” and significantly enhancing the efficiency of video creation and editing. In visual content artistic creation, county-level converged media centers can use content as a basis and employ AIGC creation platforms like “Yun Jian” to leverage AI drawing, intelligent charts, and extended reality technologies in the AIGC field to explore new modalities, forms, and carriers, creating AIGC visual products with local characteristics and cross-boundary effects. (2) Precisely Mapping User Profiles to Optimize Personalized Services Currently, county-level converged media centers aim to build an intelligent ecosystem for converged media communication centered around the core concept of “integration” and supported by “intelligent” technology, with a view to achieving precise services in content innovation, distribution, and promotion. However, due to limitations in technology, manpower, and funding, county-level converged media centers still face challenges in human-computer interaction and personalized services. In the future, they can utilize AIGC technology to map user profiles more accurately and provide a wider variety of personalized services. Although county-level converged media centers’ client apps and public accounts currently employ intelligent customer service systems, these systems are not yet mature in voice recognition and natural language processing, making it difficult to accurately grasp users’ intentions and expressions. As a result, their responses often lack personalization and human touch, requiring human intervention in dealing with complex issues. Therefore, county-level converged media centers can develop intelligent chatbots using AIGC tools, utilizing natural language understanding (NLU), dialogue generation, and sentiment analysis technologies to quickly recognize users’ intentions and emotions through conversation, thus providing accurate and timely services to users of the county-level converged media centers. Moreover, based on user profiles, they can offer personalized interactions and services. For instance, for users frequently searching for local events, intelligent chatbots can proactively provide them with information on local activities. In addition, in areas such as education, healthcare, and public welfare, county-level converged media centers can utilize AIGC to provide the public with broader, smarter, and more personalized social services. For example, in online education services, in July 2023, NetEase Youdao AI technology team launched the country’s first vertical large model in the education field called “Zi Yue,” releasing six innovative applications developed based on AIGC that can conduct comprehensive and in-depth analysis of students’ learning processes, providing more personalized and precise guidance for education. Following this precedent, county-level converged media centers can develop intelligent teaching assistants using AIGC to assist students in learning languages, programming, writing, and serve as intelligent assessment tools to provide timely feedback and suggestions to students. Additionally, in online healthcare services, VanMul Health, as the country’s first AIGC operation service provider for doctors, has developed a system based on cutting-edge global medical research evidence and rich medical cooperation resources, collaborating with authoritative clinical medical experts to create a knowledge database based on single disease pathways, capable of providing preliminary diagnoses and suggestions based on patients’ reported symptoms and guiding them in self-management or seeking medical care. This inspires county-level converged media centers to develop intelligent consultation assistants using AIGC, providing preliminary diagnostic suggestions based on user symptoms and history, and serving as intelligent health assistants to provide health knowledge and care guidance. Finally, in online public welfare services, the Alibaba Public Welfare Foundation and Alibaba Cloud jointly developed an intelligent volunteer assistant based on AIGC, which can provide services such as disaster warnings, help information dissemination, and donation guidance. County-level converged media centers can, in the future, develop intelligent psychological assistants using AIGC to provide listening and support to those with psychological distress, and also serve as intelligent volunteer assistants to help those in need of information and support. Furthermore, converged media centers can quickly collect users’ historical browsing, search records, and reading durations using AIGC tools, analyzing factors such as their knowledge structure, capability quality, and behavioral preferences, and, combined with social attributes and other related factors, recommend news that aligns with their emotional inclinations, enhancing user frequency and satisfaction, and promoting the continuous improvement of their personalized service systems. (3) Extending Virtual Reality Spaces and Innovating Metaverse Strategies The Metaverse is a comprehensive term used to describe a shared virtual space, originally stemming from the science fiction novel “Snow Crash,” which describes a system formed by the fusion of physical reality and digital virtual worlds. It not only coexists with the real world but also allows humans to have virtual “avatars” within this system. The concept of the Metaverse aims to transcend the limitations of time and space, providing a unified and coherent virtual sensory experience for human senses, while intelligent media is comprehensively transforming the media ecosystem, achieving a perception extension from the body to emotions. Advances in artificial neural networks enable machines to mimic human brain logic, promoting the internalization of technology and facilitating the integration of human consciousness and machines, closely related to the development of AI digital humans. As early as 2018, Xinhua News Agency collaborated with search engine company Sogou to launch the world’s first AI news anchor, whose appearance is based on a real person and can simulate human voice, facial expressions, and lip movements. Today, with the development of intelligent media, the simulation level of AI digital humans is increasing, and the exploration scope is expanding. County-level converged media centers can use tools like ChatGPT to create virtual images, write video scripts, and utilize AIGC’s text-to-speech (TTS) technology to convert video script content into audio formats with intonation, rhythm, and emotion, finally using platforms like D-ID to generate news videos for daily news broadcasts. For example, in February 2023, the Kunshan Converged Media Center launched the country’s first 3D hyper-realistic AI digital anchor “Kun Xiao Rong,” created using cutting-edge AI technologies such as hyper-realistic digital human modeling and multimodal recognition and generation. Relying on Baidu’s intelligent cloud and SMLTA voice large model, “Kun Xiao Rong” achieves a lip-sync accuracy of 98.5%, breaking the limitations of traditional media content production and enabling continuous broadcasting around the clock. Currently, most county-level converged media centers still face challenges in building virtual reality and providing personalized services. They can formulate specific Metaverse strategies based on their goals and resources to improve content production efficiency and enrich viewers’ visual and interactive experiences. In the future, AI digital anchors should develop towards localization and commercialization in the practical application within county-level converged media centers. To promote the localization of AI digital anchors, county-level converged media centers can focus on giving AI digital anchors more distinct local characteristics, including visual design and language expression, such as designing clothing, styling, and backgrounds that reflect specific regional cultural elements and aesthetic styles; simultaneously, in terms of content, they should integrate local cultural stories and language habits, effectively enhancing the cultural resonance between news reports and target audiences, promoting local cultural education and information dissemination, making AI digital anchors a bridge connecting traditional culture and modern technology. If AI digital anchors are to be pushed into the business field, county-level converged media centers can conduct live sales of agricultural products, local cultural festivals, and other economic and cultural activities through these business activities, not only promoting the diversification of local economic development but also enhancing the dissemination of local culture and community cohesion, bringing new development opportunities to county-level converged media centers themselves. With the development of AIGC, county-level converged media centers will increasingly utilize AI digital anchors, which can not only enhance the attractiveness and influence of news content, further increasing audience interest and emotional experience but also promote the dissemination of local culture and the development of local economies, providing effective support for the transformation of traditional media in the digital age.
3Conclusion The introduction and integration of Artificial Intelligence Generated Content (AIGC) technology in county-level converged media centers not only marks a transformation in local media operations but also serves as an important indication of future trends in journalism. In the future, AIGC technology will become a driving force for innovative development in county-level converged media centers in areas such as content generation and audience engagement. In the field of news content generation, AIGC technology significantly improves efficiency, especially excelling in news automation and standardization. However, this high efficiency does not mean sacrificing the originality and depth of news content. Facing the pressure of production speed and quantity, county-level converged media centers must uphold the quality and depth of news content to ensure the integrity of news value. AIGC technology has enormous potential in creating personalized content and interacting more effectively with local communities. By developing AI models that understand and reflect local dialects, cultures, and preferences, county-level converged media centers can not only produce content that better meets audience expectations but also explore new revenue and support channels. The introduction of AIGC technology also brings new challenges, especially concerning the ethics of content generation and potential biases. The key to tackling these challenges lies in establishing a strict ethical framework that emphasizes transparency and accountability in AI-generated content, including ensuring data privacy protection, combating social biases, and maintaining the authenticity and reliability of news reporting, which will also be a focus of future talent training for county-level converged media centers. Overall, county-level converged media centers are at a crossroads between traditional journalism and AI-driven content creation, and their future development will depend on how effectively they can leverage the potential of AIGC technology while upholding journalistic ethics and fostering community trust. Ultimately, the integration of AIGC technology will become a powerful tool to enrich and enhance the media landscape, continually playing a key role in the future development of journalism.