According to the report, machine learning will help to popularize the generation of news headlines and content recommendation through algorithms. The User Generated Content (UGC) and automatic content creation technology will personalize and intelligentize future news content.
As noted by Jarno M. Koponen, who is working on intelligent systems and personalization technologies and the product lead at Yle, Finland’s national public broadcasting company, the content at the next stage will be customized based on consumers’ preference and emotions and automatically generated by the software. With the aid of AI technologies such as user digital footprint, personal preference and natural language understanding, the content, including news article, real-time video and streaming media service, will be automatically generated through the software.
According to Moivebook, a Chinese company devoted to intelligent image and video production technologies, image and video will be the main information carriers following text and picture. As video is becoming the major information carrier more and more quickly, the creation, production and broadcasting of image and video content will become more personalized and intelligent. Intelligent image and video will coexist with their predecessors.
From intelligent recommendation to more intelligent content
When users use YouTube, Facebook, Google, Amazon, Twitter, Netflix or Spotify, algorithms will recommend personalized content relevant to them. Nonetheless, as of now, there isn’t much difference in content experience among different users. For example, when the same news article, streaming video or TV series is recommended to different users, all of them will read, watch and experience the same content.
The advent of intelligent content will, however, change this soon. It will not be long before users could experience new forms of intelligent content, thanks to the seamless combination of users’ personalized need tracking, machine learning and content.
What is intelligent content (intelligent image and video)
Intelligent content means that content itself is influenced by the content for reading, watching or listening. With the aid of the technologies such as AI, users’ personalized need tracking and machine learning, new forms of content including audio, video, text or image, different from the physical world, will be created. The content itself will vary according to the number of users.
Pioneers in this area have emerged. The whole content experience of TikTok is driven by short video and audio-video content sequence. If users consent, they can be ordered and packed together through algorithms. Each user will be presented with different personalized content based on his/her view history and profile.
Recently, Netflix began to test new forms of interactive content (TV series such as Black Mirror: Bandersnatch). For instance, users’ choices will directly influence their content experience, including dialogues and story plots. The best is yet to come. Netflix is experimenting the episode sequence of the series Love Death＆Robots so that different users can enjoy episodes in different sequences.
Interactive video and audio content was employed by the streaming media of sports events in the early days. Users could decide which stream to follow and how to interact with the on-site content. For example, the user may select the repeated stream and find the critical moment based on his or her interest.
Moreover, machine learning is being used to create images and videos of imaginary people, creatures or scenes. The current system can rebuild and change the whole video, such as the video structure, scene, lighting, environment or leading character. MovieBook also pointed out that the solution of AI intelligent image and video would be able to produce different genres of music and videos.
The personal short videos of TikTok will be changed to the personalization mode automatically by the AI intelligent image and video system. The whole video can be customized or remade to be suited to the user. It is also possible that the option of interactive content of Netflix will affect the plot, dialogue, audio track or even video frame. All of these choices will be created automatically based on the preference of the users.
As a matter of fact, the full application of the intelligent image and video solution to the streaming media is turning personalized intelligent content. According to MovieBook, its automatic content production system adopts NLP technology and self-developed MAPE production engine and hence is able to produce a large quantity of visualized news content which is short, easy to understand and even innovative. At present, some media companies are capitalizing on their automatic content creation systems or robot reporters to create complete articles, video and audio editing as well as visualized news materials. Content atomization (breaking down content into smaller and modular information blocks) and machine learning can greatly increase content production to support the creation of intelligent content.
How to create intelligent content (intelligent image and video) with different experience
Content itself can be seen as an iterative and configurable commodity or process rather than a ready-made static whole waiting for release.
What is important is that the core of content experience has been altered. Intelligent content is comprised of modular elements which have been atomized. Those elements can be modified, updated, remixed, replaced, omitted or activated pursuant to different rules. In addition, content modules generated in the past can be reused if applicable. The design and development of content can be iterated just like software.
Currently, innumerable human and computing resources have been devoted to content distribution and content preparation for the recommendation system, from intelligent news application to service transmission on demand. It is not an isolated process of intelligent content, content creation and the preparation for release and distribution channels. By contrast, the original data used to describe and define content and other hidden functions have been an indispensable part of content creation from the very beginning.
Intelligent content, narrative image or video are part of the iterative feedback cycle. The action, emotion and other signals of the users can create the content and recommend content experience during the whole consumption cycle. By virtue of the intelligent content function, short news videos or streaming media can conduct iteration and management for different content intra-frame elements, which makes it extremely easy to fulfil functions such as star replacement in the entertainment streaming media and tool creation in the streaming media.
Creating intelligent content entails human planning and machine intelligence. Human beings concentrate on the things which need creativity and in-depth analysis while AI is responsible for automatically generating, assembling and iterating the dynamic and self-adapting content just like software.
Intelligent image and video can provide distinct configuration of user application, equipment, language and environment for different users. The same piece of content incorporates elements which can be accessed through voice user interface or displayed in the enhanced reality application; the whole content can also expand into completely immersive virtual reality experience.
Intelligent image and video represent the final combination of AI technology and story telling. News media should be among the first group of organizations that took advantage of intelligent content. As intelligent image and video begin to become the main information carrier, players who can master it earlier than others will emerge as the digital giants in the future. That is one of the major reasons why technology behemoths take intelligent content so seriously. The era of intelligent image and video is approaching.