news

Huang Chenxia:Hello everyone, welcome to "Six Days in Midsummer". The theme of this roundtable is "Generation and Truth: The Media Industry in the Age of Big Models". At the end of 2022, OpenAI will launchChatGPTOvernight, everyone started a heated discussion about the new round of changes in the application of artificial intelligence technology in various industries triggered by ChatGPT, which brought anxiety to the entire industry. Many people worried about missing the next round of technological iterations and updates. At the same time, many people believed that the technological changes caused by large models may trigger a new round of technological revolution in the next thirty to fifty years.

Therefore, the roundtable theme of this "Six Days in Midsummer" is "What kind of changes will happen to the media industry in the era of big models?" This "Six Days in Midsummer" is very honored to invite several heavyweight guests: Mr. Hu Yanping, chief expert of FutureLabs, Mr. Wen Da, a pioneer in artificial intelligence in the media industry and chairman of "Daily Economic News", and Mr. Yang Jian, vice president of Tencent Group and chief consultant of Tencent Research Institute, all of whom are senior practitioners in the field of media content.

Recently, everyone has been paying special attention to Apple's latest press conference, WWDC. The reason why everyone is paying attention to Apple's WWDC press conference is that they are paying attention to the AI ​​technology brought by this round of latest big models, and how Apple will apply new big models in the mobile Internet, especially in new content production. Mr. Hu has always been a senior observer and researcher in the field of Internet and technology, because around 1995, when the Internet industry was just emerging, Mr. Hu began to pay long-term attention to the impact of new technologies and the Internet on thousands of industries. Regarding the changes in the AI ​​field caused by this round of new big models, from the end of 2022 to now, what are the points that you are most concerned about in this round of new trend developments and changes?

Hu Yanping:Judging from the new wave of big models released at WWDC24 and GPT4.0, the most important points are four points, and these four points are precisely the distribution channels of media development, the content form of media, the scenarios of media, and the relationship between the future of media and intelligence. These four aspects are highly closely related.

First point: Strong intelligence is forming a new entry point, which changes the form of interaction with users. For example, when Siri cooperates with OpenAI and is based on edge intelligence, it will become a new entry point. This new entry point is natural interaction. This is why after the release of GPT4.0, I used it on the mobile APP for a whole night, walking on the street while debugging ChatGPT4.0, using the voice chat mode to debug it, its sense of delay, its content response, and its self-processing behind the chat, including what content it can deliver to you. So this is a very important aspect, that is, the new entry point, the new interaction, especially this natural interaction, will greatly change the form and channel of information distribution.

Second point: This is the direction that ChatGPT4.0 and Apple Intelligence are exploring. For example, Google, Meta, and xAI are all exploring multimodality. In a multimodal scenario, content is no longer text or news. It also changes the way content is generated, the way content creators generate content, and the way users obtain content, because multimodality is not just audio and video.

Third Point: The extension and extension of multimodal development, that is, spatial intelligence. Obviously, it is not only AIGC that can do spatial intelligence. In other words, when it is presented in scenarios such as embodied intelligence, spatial form, and spatial intelligence, it redefines information, news, and data to a certain extent. As for redefining the media and the audience, this goes without saying.

Fourth point: A mix of different forms of intelligence, on the end side and on the cloud side. Especially on the end side, such as small-scale models of 8-bit, 10-bit, or more than 10-bit. The second is on the cloud side, such as a hybrid fusion like GPT, including the combination with end-side capabilities, such as Apple Intelligence. In fact, before Apple, many other mobile phone companies, such as Samsung and Honor, as well as AIPC companies, have also explored this, and it has been proven that users actually need it very much, because for each individual, the first problem to be solved is the management of all local content data information, and based on private domain scenarios, to do various creation and productivity-related work on the basis of data information privacy protection. But at the same time, because the capabilities of the end side and the cloud side are different, it requires not only the integration of capabilities, but also the combination of information fields. When doing some creative work, you may need to call on more than just local content. In this case, cloud capabilities are necessary because the cloud will be combined with various methods such as search, and such integration has a very promising direction. In fact, it is also the ongoing prototype of digital intelligence and the emergence of "super individuals". Because the agent on the end is a foundation, allowing everyone to become a "super individual", a digital person with computing power and intelligence, the concept of the audience has actually changed.

From the past to the present, media has been redefined three times, from the stage of media, self-media to "super individuals", everyone is both a generator and an audience. As for the content itself, it has also been redefined three times, from news information to UGC to AIGC, and AIGC is not the end. So when we look at the new directions of these technological developments, we see the infinite possibilities of media in the future, the infinite possibilities of individuals in the future, and each media is already a completely new form, including the channel of information dissemination. On this basis, there was a term in the past that is still valid today. In the past, it was about the masses, and later it was about the segmented masses. In fact, there is a concept called "micro-crowd". Micro-crowd is the union of every tiny individual. The tiny individuals gather in an instant and disperse in an instant. At the same time, they gather on every topic and disperse on every topic. However, micro-crowd is still accepted by everyone in the industry within a certain range. Some people even registered a trademark with it and established a company.

At the same time, some academic researchers are also proposing the concept of micro-crowd, including our SICAS model. Therefore, this discussion not only opens up a new dimension of thinking, but also explores new paradigms, new scenarios, new distribution forms of the future of media, and a new possibility between the media and individuals. So I am very happy to participate in such a discussion today and share with you.

Huang Chenxia:Mr. Hu shared a few points. The first point is that the AI ​​revolution caused by the new round of big models will lead to great changes in human-computer interaction, because the most important focus of each new technology is that human-computer interaction has changed, which has changed or subverted a lot. The second point mentioned is that the capabilities of AIGC or big models will further impact the original model in the content field, such as the production of distributed content.

The third point explains that in the new era of artificial intelligence, "super individuals" will appear, which needs to be discussed in detail later. In the future, what role will "super individuals" play in the entire media field? What impact will it have on the original production chain and distribution chain? What kind of impact will the audience have? From the initial so-called mass media, from mass to segmented to micro-segmented, many social media now have a wave of "gathering". Everyone gathers in a certain topic and forms tribes in a certain place. It is not so stable and will soon disperse. Before talking about this topic in detail, you can first listen to Mr. Wen Da's views, because in the past few decades, Mr. Wen Da has always been a very determined practitioner in this industry. Just now, Mr. Hu also talked about the different changes brought about by the technology he has seen. In Mr. Wen Da's work in the entire media industry, especially in recent years, how do you apply these new technologies in your management of the entire "Daily Economic News" business? Because "Daily Economic News" is also a very advanced entity in the media field. What do you think the AIGC capabilities brought by the new round of big models will have on the media industry? What specific practices are you doing now to deal with it?

Wenda:Chenxia just mentioned a few points. First of all, I think there is a natural gap between the current mainstream media and artificial intelligence, and this gap is basically insurmountable. Because writers and artificial intelligence developers are completely two different groups of people. From the perspective of their academic background, writers are mainly liberal arts. New technology is a different way of thinking. Especially in recent years, if the media simply works in the traditional text field, its future prospects will become smaller and narrower.

Around 2020, I defined 2020 as the first year of technological transformation for the Daily Economic News. This is different from the viewpoint just raised by Mr. Hu, because Mr. Hu looks at this issue more from a big picture and theoretical perspective, while we apply it in actual combat. The Daily Economic News mainly targets listed companies. There are more than 4,000 A-share companies. But in 2019 and 2020, the number of listed companies that our reporters and editors can cover does not exceed 1,000. If you want the media to have enough influence and want to cover the entire market, but you don’t have enough manpower, you have to use technology to solve it. In order to seek new technologies, we talked about cooperation with Alibaba, Baidu, and Tencent, but traditional media were short of funds, so there was no way to increase costs to introduce new technologies.

Because it is possible that costs have been increased and new technologies have been introduced, but the rate of return is very low and the income is not enough to cover the expenses. Later, by chance, we reached a consensus with Hongmian Xiaobing and started to work on AI strategy. In May 2020, we launched the express system, which basically used AI to cover all 4,000 listed companies at the time and generate express news. The "AI Strategic Project" of the Economic Daily won the 2020 China Newspaper Deep Integration Development Innovation Case by the State Press and Publication Administration. Looking back now, it was very correct for us to launch the AI ​​strategic thinking in 2020. With such an action, in 2021, we defined "Daily Economic News" as the first year of video transformation. Because at that time, we saw a trend that young people were sharing and empathizing with videos more and more. However, how to convert text into video is also an extremely brain-burning process, and you need to carefully consider whether to increase cost investment.

Based on the text reports and news of listed companies, our second step is to think about how to convert text into video, audio, and overseas dissemination, which is the concept of "one chicken, many ways to eat". At the same time, we must always consider the issue of energy efficiency. So on December 20, 2021, we launched the world's first unattended AI TV for the Daily Economic News, which is available 24 hours a day, 7 days a week. This year in Shenzhen, we released the 3.0 version of AI TV. The 3.0 version of AI TV is basically our independent intellectual property, which solves the overall transformation of traditional media from text to video. Here, we have to mention the impact of the big model on us. Because what I was thinking about at the time was how to convert a large amount of text into the concise text required for a short video. We thought of many ways, such as various tags. We also found some partner companies in Shanghai to label the text. As a result, when the big model was launched on the market last year, I immediately realized that the best solution to this problem already existed, which made the entire "Daily Economic News" take a big step towards technological transformation. This big step can be successfully taken because our previous efforts have accumulated to a certain extent, and naturally we can see that this technical means can solve our current problems.

In July 2022, after the launch of AI TV, Daily Economic News has always been considering how to convert text into video? Although video is a trend, the labor cost is too high. At this time, we still use machines and computing power to solve the matching of text and media resources, and then generate videos. Considering how to distribute a large number of videos generated, we need to achieve one-click publishing to various platforms. But there is a trick here-there are significant differences between short videos of finance and short videos of cultural tourism and catering. Financial short videos pay more attention to conveying professional views and in-depth analysis. It does not require a strong on-site atmosphere or on-site recording. The audience is attracted by your views, that is, the sound of the content and the impact of the text. In the content matching stage, we need to have a relatively fuzzy processing in the matching process. As long as the matching degree of text content and media resources basically reaches the normal level, the output newsletter can meet the preset standards. Of course, everyone hopes that each short video can reach the standard of 100 points, but in order to achieve this goal, it may take 5 hours to produce. But now, it only takes 3 seconds to produce a short video with a standard of 80 to 85 points. I think this greatly improves the efficiency of our content production.

However, our media resources are relatively insufficient. For mainstream media like us, we must ensure that the media resources used are legally authorized, which is completely different from self-media. In order to deal with these problems, we have adopted two methods: First, we adopted the spirit of "Yugong moving mountains" and gradually accumulated media resources through the "ant moving" method. We have conducted extensive filming of listed companies to continuously enrich our media resource library. Another method is to use AIGC to do the work. Therefore, the whole process is very natural and smooth. You will naturally feel that a certain point should be handled in this way and a certain link should achieve such an effect. In addition, we also use big models to rewrite video texts. Because, although we can distribute videos to various platforms with one click, major platforms will check for duplicates. If the video, content, and pictures are exactly the same, then the platform will reduce the dimension of your content or block it in the algorithm. In order to solve the problem, it is too costly to invest more manpower costs. With the emergence of big models last year, especially with the increasing number of big models in China, we use big models to perform various text extractions on the same content to form different texts. By customizing a unique template for each video, even if the subject matter is the same, one-click operation can generate ten different videos around the event, and they are completely non-repetitive. Therefore, AIGC provides the highest efficiency for our traditional media. We now see that some self-media may publish content on Douyin or video accounts, and it is difficult to take into account all platforms, but through AIGC we can basically take into account all different platforms.

Chenxia just asked, how should a traditional media clarify its development direction? This is indeed a very thorny question. First of all, the key lies in top-level design optimization. It is indeed a top-level project, but this top-level project also faces huge challenges. The entire team may find it difficult to understand or agree with the reasons behind this decision. For example, the revenue and profit of the Daily Economic News have remained relatively stable in the past few years. Team members will wonder why reform and innovation are still needed? I know that I can't simply promise the team to sacrifice current interests in exchange for more generous returns in the future. Because I am not sure of this greater benefit, I must try it while taking into account the interests of the existing team. Especially as a state-owned enterprise, it is also faced with various rules and regulations. It is necessary to be extra cautious in capital operation and resource allocation. In this case, the most appropriate way is to reduce investment as much as possible, carefully plan, constantly create, and optimize details. Looking back now, I will find that we have spent a lot of energy in the past four years.

At the same time, in this process, traditional editing and traditional business operations cannot be left behind, so the entire transformation process requires a lot of energy and brainpower. We cannot let colleagues in business or editing participate in such future-creating retreats every day, because they still need to deal with daily practical work. So at that time, I and some young backbones and small teams were specifically exploring new directions. For example, my executive editor-in-chief, he still continued his daily work content. My executive general manager still focused on his traditional business operations. These two aspects cannot be ignored. We held at least 1,000 theoretical meetings to create the future, and even more than 10 meetings every day. Whenever new inspiration appeared, we would immediately put it into practice. After achieving certain results, we began to guide the team to feel the actual benefits brought by this transformation. For example, our publication volume increased 10 times, but there was almost no increase in labor costs. During the entire transformation process, we generated certain benefits. However, the overall traditional market is shrinking. It is these achievements that have made team members taste the sweetness, so that they are more aware of the importance of transformation, and everyone will realize that such a transformation is necessary.

Huang Chenxia:Mr. Wen Da's statement just now is very attractive and very practical. After Mr. Hu discussed in depth the changes in future macro trends, Mr. Wen Da used the "Daily Economic News" as an example to explain his views from a micro level. Although we may think that some trends will not change until the next 50 or even 100 years, there are many concrete problems in the rhythm and many practical operations at the micro level. For example, how to change the original muscle memory of the organization so that it can truly adapt to new changes. In any transformation process, there are challenges to maintain the stability of existing businesses and develop emerging businesses, unless there is sufficient capital to support continuous investment. Therefore, it is very important to stabilize the original business while extending new businesses, and use new businesses to drive and upgrade the capabilities of the old business in the past, so that it can move towards a new business model. In this process, artistic decisions at the micro level such as rhythm control, organizational adjustment and management optimization are crucial. Later, we can further discuss this topic with Mr. Wen Da. In fact, the changes in efficiency brought about by AIGC will eventually reach the other side of benefits in the next 3 to 5 years. How is efficiency and benefits balanced? This is also a topic we will discuss later.

Next, we invite Professor Yang to talk about this, because Tencent Research Institute has spent a long time publishing a report on the development and changes of the media industry under the new wave of artificial intelligence. So I would like to ask Professor Yang, you used to be in the media industry, and you are very familiar with the media industry. How do you view the impact of this round of new artificial intelligence technology on the news industry and the challenges that may be faced in future development and changes?

Yang Jian:Professor Hu is more of an observer in the media industry, while Professor Wen Da is more of a practical one, expressing their different views and opinions on the new big model. From my personal perspective, I have worked in the news industry for more than 20 years, but now I think I basically look at things from the perspective of an audience. Of course, the positioning of Tencent Research Institute is not exactly the same, but in general, I think media research is not the track that Tencent Research Institute is best at. We just tried to make some observations.

In fact, the depth of our observations is still somewhat different from what the two teachers just talked about. In general, our most intuitive feelings about the impact of news media are:

First point: The core competitiveness of the big model so far is its strong reading comprehension ability and cross-language understanding and communication ability. In fact, we were discussing this some time ago. After reaching the big model stage, artificial intelligence has broken the boundaries of all languages ​​at the bottom level, not just human natural languages, such as Hunan dialect and Beijing dialect, and English and Chinese. It is not limited to these. It also includes many programming languages ​​and machine languages. So it does have a particularly strong reading comprehension ability. In the past, when I first started doing journalism, the first report I did was about the failure of Beijing's bid for the Olympic Games. Many reporters were interviewing the specific application process, the mood of failure, and what to do next?

So my job is to make a comprehensive data, such as how many Olympic bids there have been, how many countries have applied, how many times have countries applied, and how many times have they failed. It is actually a very simple thing to do. But in every news event, the audience has this demand, because there is nothing new under the sun, and many times the stories we tell are repeated. When a new and similar event occurs, how to look at this event and interpret it often has to be linked to the past. So a large part of the work in news production is actually doing this. According to our survey, so far, about 80% of domestic news media practitioners have tried to use AIGC tools, and 60% of them said that they would use AIGC tools in their work in the future. The situation we have learned in foreign environments is similar. The Associated Press conducted a survey, and about 70% of people want to use or say that AIGC will change the form of news. 49% of people believe that the entire news industry will undergo tremendous changes due to the development of artificial intelligence. At that time, the Associated Press had a very interesting data. One is data collection, 54.8% of people used it, and translation, 44% of people used it. Why do I remember these two things in particular? As I mentioned before, the background reporting work of the Olympics will no longer require people. This is a disruptive technological change. Why do I remember 54.8? It means "I will die"(homophonic), I definitely have no chance, there is another 44%, "I'm doomed"(homophonic)This is the first point. Everyone uses AIGC's tools to understand, collect, and assist in reporting. The second aspect is that AIGC actually has a great ability. It not only understands these materials, but also understands the internal logical relationship between various things and some existing materials.

AIGC also has a strong ability. It has a better perception of the audience's needs and can help people do their work better. In addition, the audience has preferences. Professor Hu just mentioned that the phenomenon of micro-gathering is very prominent now. But on the other hand, in addition to everyone's instant gathering, there are also preferences and preferences. Why are some people used to reading the content of the New York Times, and another group of people are willing to read the content of the People's Daily? At what level do these media affect individuals, and have such a deep connection with the needs of these different individuals. In the era of traditional media, the audience and the media themselves may have had a preliminary understanding of certain relationships between each other. But now that artificial intelligence has emerged, it can help you analyze these relationships. Artificial intelligence not only helps us deeply understand the complex relationship between audience needs and media content, but also tells us how to use these relationships more effectively to optimize content. Take our short video section as an example. Although it is not completely equivalent to news media, we are now using such laws to make more innovations in future news production.

At present, the algorithm itself is not so stable. Many things are achieved through black boxes, and it is impossible to completely control from the source and exert influence on certain things. However, with the continuous development of artificial intelligence in the future, I believe that these problems can be reasonably solved. The third major impact is the generation of some multimodal news that was not previously able to be generated, so the entire news style will change greatly. In the era of traditional media, there are differences between print media and television media. Print media is limited to text expression, while television media can make videos and attract a wider audience. There are also newspapers and news agencies. Newspapers can only be published once a day, while news agencies can publish in real time. When the Internet appeared, the relationship between these three was balanced. After AIGC is market-oriented, it will even blur the boundaries between professional and amateur, and the competition between various media will become more intense. Therefore, there is a relatively large impact in terms of assistance, recognition and the generation of new forms. This is the main finding of our report.

Huang Chenxia:Professor Yang just gave a general summary of the changes in the industry from the perspective of the research institute. I would also like to discuss a specific question. As the big model is introduced to the market, all walks of life are facing a problem. What kind of changes and what kind of unchanged will this round of technological changes bring? Today, we put this question in the field of media. From the perspective of each teacher, what kind of changes will the technological changes brought by AIGC bring to the media industry? And what will not change?

Hu Yanping:The law of information value must remain unchanged, but the form of media or communication channels, including the relationship between the style of content and the audience, will definitely change. Chenxia just talked about an issue that I think is very important, that is, efficiency and effectiveness, which are actually two different aspects. I call it the difference between efficiency and effect. Whether it is in the computer stage, the Internet stage, the mobile Internet stage, or the current intelligent stage. Facing the information explosion and content explosion, that is, the efficiency of generating content is increasing dramatically. For example, now there are 100 digital people, there can be 1,000 live broadcast rooms, and one manuscript can be written into 100 articles, changed into 10,000 articles, and distributed to various platforms. This is an improvement in efficiency. But whether it is the audience or the micro-audience, the bandwidth and cost of each person's time are limited.

As efficiency rises, the marginal effect in the sense of generation is increasing, but the marginal effect on the output of unit content is decreasing, that is, the effect of creation is decreasing. So there are two very interesting phenomena. The first is that in the current communication environment, there will be a lot of seemingly plausible content. This is caused by the defects of intelligence itself. This problem will be solved to a certain extent in the next stage. But what is another more interesting phenomenon? It is the current hollowing out of the media and the individual's heart-to-heart, which have become two extremes. One is large-scale production, and the other is individual creators relying on their own content creation to attract a large number of fans. These two form the two extremes of today's world.

Why are there two poles in this world? Chenxia just touched on a very fundamental question, which I call the first principle or the underlying logic. So what is the underlying logic? Media is actually part of information, and information is part of the information flow. The information flow is part of the operation of this world. In general, it is the flow of information, logistics, and people. The style of media is just a container, so in this situation, only when we think about the difference between the present and the future, can we improve efficiency and effect at the same time. But how to improve the effect? ​​After reading the relevant report of Tencent Research Institute, I think there are four preliminary answers to this part.

The first aspect: In my opinion, no matter how intelligent the content generation is, people, humanity, humanities and human subjectivity are always irreplaceable. In fact, these elements go beyond the capabilities of artificial intelligence to a large extent. We should integrate humanistic elements into AI, rather than impose AI on people, because hundreds of millions, or even billions of users, each of whom is a unique soul, can capture the difference between the meaningful and meaningless information at every moment, which is a very subtle thing, and this subtle perception is difficult to measure by many technical means. Furthermore, humans themselves have a characteristic of resisting algorithms and intelligence, which stems from our spiritual pursuits and autonomous choices as independent individuals. So this is the first very important thing I think. I think the underlying logic level is to have both efficiency and effect, both diffusion and scale, and at the same time to be sincere. At least at this stage, sincerity is something that only people can do well.

The second aspect: We need to pay attention to a very fundamental trend. After the stage from block to fragmentation to atomization, a new situation has emerged. After continuous reconstruction and creation of various content styles, as efficiency has decreased, every user now has a more important influence on the content.

With the development of the times, people's way of obtaining information has changed. In the past, when we browsed websites, we might spend the whole night but could not get some effective information. Subsequently, the rise of mobile apps divided the Internet into information fields, and the mobility of information was significantly improved. At the same time, search capabilities declined in the mobile Internet stage. However, at the current stage, we have witnessed the rise of intelligent entities, such as Tencent's Hunyuan, Yuanbao, Qianmeng, Hongxin, and ChatGPT, which give users the ability to surpass algorithms and realize the so-called super algorithms. So what is an algorithm? It is like the distribution of news information in the past that created these algorithm-based platforms, but such an algorithm, no matter how it is combined with AI, is fundamentally not user-centric or individual-centric.

We are now in the super algorithm stage. For example, if we ask questions based on Siri or GPT4.O, it will complete a series of operations from search to screening and personalized recommendations in an instant. Take Tencent's Hunyuan as an example. It predicts the content that users may be interested in and displays it. So now we have reached a new stage. This new stage has changed the relationship between the media and the public, allowing each individual to regain an unprecedented initiative. This initiative means that the efficiency of many media forms has decreased, and that many media have been placed in the back end. For example, you have completed a distribution, but it is selected again in an instant, so this is a very important part.

The third aspect：We must fully consider scenario factors and market environment. Each user's time is limited. In fact, various apps and future intelligent entities are competing for users' time. When competing for users' time, we need to think about how to use information, scenarios or intelligent capabilities to improve user experience. The key is how to help users improve efficiency, not how to help the media improve efficiency. So I am more fond of this kind of reverse platform based on intelligence, content and information, because it can stand from the user's perspective and help users improve efficiency, so I think this is the third answer.

The fourth aspect：We see that from the stage of block, fragmentation, atomization to reconstruction, the content information obtained by users will roughly go through three stages. Starting from AIGC, the first stage is called "seemingly true but not true", and users will encounter a lot of ambiguous and difficult to distinguish true and false information. The second stage is "seeking truth from facts", which is actually both the evolution of AI capabilities and the improvement of users' selection capabilities. This is a process of two-way rush and two-way screening. The third stage is "digital-real integration", a fusion of the intelligent world and the physical world, but this integration must be based on the solution of the problems in the first stage. If the problems in the first stage have not been solved, then the environment we are in will be like a fantasy or a spiritual world. Because whether intelligence uses sensing or multimodal methods, in fact, intelligent agents can already begin to generate news and information. This is the case in many fields. There is no need for manual collection. We call it manual journalism, because manual journalism, from paper media to online media, and now to self-media that interacts with tens of thousands of people every day. Therefore, in the fourth stage, we need to fundamentally solve the problem of seeming true but not true, and face the world of digital-real integration in a realistic way.

In the world of digital-real integration, we receive a lot of news and information not only from humans themselves, but more from embodied intelligence, spatial intelligence, and intelligent bodies. So humans have really become an audience to a certain extent. But becoming an audience does not mean that we are completely dependent on intelligence, which is another question, because it is always people and things. At the level of things, this can be done in other fields such as news information, but hundreds of millions of people and billions of people actually have a very important field, which is the spiritual life of the entire human race. Spiritual life is not limited to reading literary works, but also the field where we interact with everyone anytime and anywhere. This is the spiritual world of people. So I look at the present and the future from the perspective of first principles or underlying logic, and analyze the efficiency and benefits that Chenxia talked about from this perspective. Only in this way can we grasp which are the key issues and which are the problems we will encounter in the process of development. For example, what we expected did not happen in the process of development. Why did it not happen? Because its efficiency is increasing, but the effect is decreasing. So I think it is also important to look at the present and the future from the first principles and underlying logic.

Huang Chenxia:During the conversation with Mr. Hu, I particularly agree with several points. One is how to think about the relationship between efficiency and benefit. The second is the relationship between intelligence and people's spiritual pursuits. How can we balance the two in the long years to come? I would like to discuss this issue with Mr. Wen Da, because Mr. Wen Da mentioned some related points in our previous conversation, which I think are very pertinent. At the beginning of the birth of the news industry, the news field was idealistic. What kind of changes will it undergo? In the future development, should such idealism become stronger or weaker?

Wenda:This question seems simple, but it actually contains profound connotations. In the current social context where human nature and popularity are generally impetuous, what remains unchanged for media professionals? This actually involves our spiritual pursuit and yearning. Humans themselves have an instinctive impulse to obtain information.

First point：This impulse stems from curiosity. Today, with the development of self-media, information dissemination has entered a new stage. At this stage, people's pursuit of authenticity is becoming stronger. Under the trend of traffic being king, there will definitely be differentiated and mixed content. In order to pursue traffic, some self-media do not hesitate to fabricate facts, distort the truth, or even make up stories out of thin air. I call this phenomenon "noise". Looking back, in the era without the Internet and smartphones, this noise was mainly limited to gossip among neighbors, and it would not constitute a major trend relatively speaking.

But nowadays, the popularity of the Internet allows everyone to speak out. Once a certain view or idea gets the attention of a large number of audiences, it is likely to spread quickly. In fact, its orientation is biased, but people pursue it out of curiosity, and the authenticity of the information is often drowned in the tide of the Internet. This is an aspect that we as traditional media people are very worried about. So I think the more in this situation, the principle of authenticity is a principle that traditional mainstream media must adhere to. Because people are still eager to know the truth of facts, the truth of policies, and the truth of the events themselves. Moreover, I think as media people, we ourselves should have a higher sense of morality than ordinary people.

Media professionals should first look at the world at eye level and then report. We usually come into contact with many successful people, but we cannot create content based on emotion just because we admire someone. Such reports cannot meet the needs of the audience. When dealing with interviewees, media professionals should not take a condescending attitude and be too contemptuous of people and events. Therefore, as a professional media professional, we need to think deeply about this issue. We should always treat every interviewee at eye level, and reflect and report the facts truthfully and objectively. Only in this way can we maintain our composure in our hearts and provide high-quality news reports to the audience.

Second point: We always adhere to in-depth reporting, on-site reporting and professional understanding of events. The National Business Daily has always been adhering to the banner of "news determines influence". I often say that "news determines influence" actually not only reflects our feelings as media people, but also our mission and responsibility as social watchers. In actual work, I also put forward the eight-character editing and editing principles of "professionalism, depth, story, and communication". As a financial media, we must be professional, otherwise we can't produce in-depth articles, but now it is not enough to have professionalism and depth, we must also learn to tell stories.

Because the media is generally still a fast food culture, whether it is a daily newspaper or a magazine, its core form is still fast food, so we need to attract audiences and expand the scope of dissemination through storytelling. Just like the Wall Street Journal in the past, before the 1930s, its circulation was only 50,000 copies. It was very professional, but it couldn't tell stories. When his editor-in-chief Baron was promoted to president, he suddenly realized that the reason why Homer was so popular with readers was because of his storytelling ability. Inspired by Homer's epic poems and realizing the importance of storytelling, Baron looked for inspiration in literature departments across the United States and incorporated it into media content to create the "Writing Line."（Write Line）". This move strengthened the storytelling of The Wall Street Journal, thereby attracting more readers. By the 1980s, the circulation was approximately 1.8 million copies.

In the past, traditional media did not need to actively spread information. After subscribers subscribed, information was spread through word of mouth. Mr. Yang also transitioned from the mainstream media People's Daily. He knew that media people were not good at self-propagation. When promoting information for other cities or companies, they had many different means, but they were often at a loss when it came to promoting themselves.

So today I am very fortunate to participate in Tencent Research Institute's "Six Days in Midsummer", which is actually a kind of communication for "Daily Economic News". Modern media must be both professional and in-depth, and at the same time be good at telling stories and learning to spread themselves, so that mainstream media can occupy the main platform. Then let's talk about the changes. With the impact of AI technology, whether it is UGC or AIGC, the core lies in the word AI. The development of AI technology has covered all other content. I think AI will have extremely high requirements for the quality of media people.

Different from previous technologies, AI technology requires a comprehensive ability. It can even be said that AI technology has broken through the boundaries of professions. I even think that there will be no so-called differences between science and engineering and liberal arts in the future. Take Mr. Hu as an example. Mr. Hu is very fluent in technology and has a deep insight into the media industry. In the future, we will need talents like Mr. Hu who have high comprehensive qualities and professional abilities.

In addition, the form of media itself is also changing. For example, in the past, only text and newspapers were used as information carriers, and other means of communication such as the Internet and video were not considered. However, newspapers are now a thing of the past. Looking back, Kodak film was a product that we thought was indestructible, but it eventually disappeared. So we can completely associate it with the form of newspapers. Its disappearance is an inevitable trend.

Therefore, under this circumstance, the media of the future will present a stronger richness, and the dissemination of information will be more timely and extensive. What remains unchanged is the anxiety of media people. In the traditional era, media people have anxiety. In today's era of change, this anxiety has escalated into panic. But in fact, I think technology is equal rights, it gives everyone equal principles, so under such an impact, what remains unchanged is our pursuit of new things. Media people need to keep learning and keep up with the pace of technological updates. It is never too late to keep up. Technological equality means that no matter how old you are, as long as you have passion and determination, a person can still make full use of technical means, cross the early stage of knowledge accumulation, quickly reach a certain level of knowledge, and combine past life experiences to create moving works. So I think change and constancy, that is, you have me and I have you. I think, as media people, we are still full of confidence in the future. In this world, there are always people who want to reveal the truth, and the audience's demand for real information is endless. Therefore, the role of media people as social watchers cannot be changed, and their enthusiasm for the news industry and professional qualities cannot be changed. With such a foundation, we will be able to cope with the ever-changing world, including the impact of technology, changes in rules and innovations in communication methods.

Huang Chenxia:I agree with what Professor Wen Da said. For the media industry, its core values ​​and fundamental purpose have not changed with the development of the times. In every era, there must always be some night watchers and lighthouses of the times. For a long time, the symmetry of news information has played the role of the night watchers or lighthouses of the times. Because of the continuous symmetry of information, it ultimately solved social equality, or became a very important driving force for the continuous realization of social equality, so the root of this kind of Taoism and idealism has not changed. As for the technical level, how to tell a better story, generate a graphic report, generate an audio and video, it will change due to changes in technology and media, so Professor Yang, what do you think are the directions of change and invariance?

Yang Jian:I particularly agree with what the two previous teachers said. I pay more attention to the changes in this industry from some micro and intuitive perspectives. After all, I come from this industry. First of all, I have observed that the mental state of my peers in the past is generally on a downward trend. Secondly, I found that the number of students applying for media majors is also decreasing year by year. So from this perspective, traditional media organizations and forms are indeed facing a huge impact. But from another perspective, if we extend our perspective, I think the entire industry is actually experiencing a process of breathing in and out.

In fact, the work done by the media industry can be essentially attributed to two aspects: one is content creation, and the other is the way of communication. In content creation, we need to take into account the communication effect and adjust the content, including the improvement of storytelling ability mentioned by Professor Wen Da just now. In terms of form, with the emergence of new distribution methods and channels, we also need to make corresponding adjustments. I think the objects of media service and the objects of service of the entire content product have a "breathing rhythm" in their needs. For example, in the rising period, the media may serve more social elite groups, and then, in order to expand its influence and coverage, the media will move towards popularization. At this time, it is necessary to consider the cognitive ability and habits of a wider audience. It feels like a breathing process, that is, your entropy is increasing, and after the increase, take a breath, and then reduce this entropy, and then increase it again after a period of time. Entropy increase and entropy decrease is a breathing process. I think we are in the breathing stage now, because we used to work in the traditional media field, and we may feel a little pessimistic. Like what Professor Wen Da said, this emotion has even escalated to anxiety and panic. But I believe that this stage will eventually pass. From the macro perspective of human history, we are always constantly overcoming entropy increase. The ideal of journalism actually means the same thing: overcoming the negative side of human nature and stimulating the instinct of life to go against the current. In fact, both artificial intelligence technology and the impact of digital technology on the traditional media industry have gone through such a process.

Around the end of last month, an old leader in his 80s sent me an article that said "In the past, it was said that institutional media had taken the wrong path and missed opportunities in the past 10 years." He said with emotion that it is now difficult to find people who care about this aspect. After reading the old leader's article, I expressed my optimism. I think, as the previous two teachers mentioned, even in the current development of general artificial intelligence, they still rely on second-hand information and are far from the stage of originality. And originality, at least for now, is a unique ability of humans that cannot be replaced by machines. People are more concerned about the physical and spiritual interaction of billions of people on the earth in physical space. Therefore, I think that it may not take 20 years, or even just 10 years, for the media industry to experience a change in breathing rhythm and return to an upward trend. Of course, this may be my overly optimistic estimate. I think Chenxia, ​​as the host today, has lost a lot of opportunities to share. In fact, she herself is a witness to the entire industry. I believe you have a lot to say, and I also want to hear about the changes and constants you have felt over the years.

Huang Chenxia:Teacher Yang's views triggered some thoughts in me. First of all, Teacher Wenda mentioned anxiety and curiosity. I think moderate anxiety and curiosity are indeed two key factors in promoting the progress of human society. If a person has no anxiety at all, then he may fall into a state of lying flat, and it is difficult to promote new changes in this state. So I think anxiety is normal. Curiosity and moderate anxiety are two factors that make the news information industry last forever. People always want to know what is happening around them or farther away. This demand always exists.

At the same time, Mr. Yang mentioned "exhalation and exhalation", "entropy increase and entropy decrease". He is very optimistic and sees this change or cycle will make the industry recover in the near future. I also feel that in fact, from 2012 to 2022, until now, it is the era of algorithms, because it is a very important technical capability in information flow distribution.

In foreign countries, during 2016 and 2017, when recommendation algorithms were extremely popular on major platforms, especially Facebook, many emerging media were born, such as Buzzfeed and Vice. At that time, their valuations were very high. They were very good at telling stories with new narrative capabilities and gained huge traffic on the platform. These are often short-lived. By 2022, many have begun to decline or even go bankrupt. Their problem is that they rely too much on generating traffic, but ignore the essence of the news or information industry.

But another example is The New York Times, The Wall Street Journal, and The Economist. In the past, they were considered very traditional and professional information producers. They used a more clumsy approach, and in their explorations of the past, they were very committed to advertising in their business models. Whether it was traffic or space in the past, we regarded it as an attention economy.

On this basis, they extended the payment, and the key point of payment is that when the content and information are scarce enough, there will be users willing to pay for it, just like the reverse screening mentioned by Professor Hu just now. I think whether it is the New York Times or the Wall Street Journal, they are very important filters for American audiences. Of course, there are also some media in China that pay for content, such as Caixin. I think such a business model will definitely survive in the long run. If all business models are bet on the attention economy, there will be some problems, because the attention economy will change due to human-computer interaction. In the past portal era, portals hijacked traffic, and later in the era of recommendation algorithms, the way of obtaining traffic has undergone significant changes. As a tool-based platform, social media has always maintained a unique position and will always exist.

When artificial intelligence appears, human-computer interaction and intersections change, causing traffic to change again. If we are always busy chasing these changes, we will indeed feel very tired. My experience at Tencent News made me realize that in the era of recommendation algorithms, we did fight a hard battle. But in retrospect, we found that there are some things that "others are hell". If your competitors redefine a new business model, but just copy the competitor's business model, it is difficult to achieve real success. In the era of the Internet and new media, economies of scale exist. Leaders will continue to expand their advantages. If you compete in the same way, you can only barely survive and cannot stand out because you don't have your own characteristics.

We gradually realize that recommendation algorithms and other natural language processing technologies, such as NLP technology before the big model, have indeed improved production efficiency, but its production is like what Professor Yang said. The content they generate is often a reprocessing of existing information. This content is often to cater to traffic, so the content itself has certain flaws, which can be said to be mixed with mud and sand. From 2012 to 2020, the entire content supply may have increased dozens of times, but the content that really caused information controversy is relatively small.

In the era of short videos, emotional content is predominant. Although such content is emotional, the information is not controversial. However, because its consumption threshold is extremely low, it can quickly attract users' attention and cause excessive consumption of time. However, unless such content is released on a large platform, it is difficult for it to have a profound mental impact on users.

For example, on different platforms, it is difficult for users to distinguish different news sources. For example, it is difficult for users to identify "Daily Economic News", "Caixin", "China Business News", and "21st Century Business Herald". For example, if "Daily Economic News" can continue to release exclusive information at the first time, as well as provide in-depth opinions and factual reports, it will attract a part of users, especially considering China's huge population base, who are willing to pay for it. This will not only generate revenue, but also enhance users' mental cognition, making them aware of the difference between "Daily Economic News" and other news sources. And the user's mental cognition is crucial. It not only affects whether users are willing to pay for content, but also includes the revenue generated by users' continued support for a brand because they like it, so that it will not be hijacked by the so-called platform.

In the era of recommendation algorithms, from 2017 to 2019, we experienced a typical bubble period. At that time, Douyin was the largest growth machine in the market. But later we found that Douyin had reached a certain stage of growth. Its content was actually circulating. Some users began to realize that reduction of complexity was everywhere, and recommendation algorithms could not completely solve this problem. Just like some of the views expressed by Professor Hu just now, people’s exploration of the diversity and spirituality of the world is very important. Some personal self-expression must be reflected in the content. If we rely entirely on recommendation algorithms and push content according to the choices of the public, then diversity cannot be achieved. So I think the pendulum will swing back a little from the current state of over-reliance on recommendation algorithms, but it will not return to the era without recommendation algorithms at all, but will swing to the stage where people have to stand on the recommendation algorithms. So I think the current AIGC may be in the period when the recommendation algorithm is just beginning.

At this point in the discussion, a new question came to my mind. Any technological development is accompanied by a bubble period. In the past, Gartner had a so-called technology development curve, with a bubble period in the middle. So I would like to ask the three teachers, do you think we are in a bubble period now? If we are in a bubble period, which bubbles should we filter out in this process so that we can usher in the more lasting and stable productivity improvement brought by AIGC and enter a new stage of development?

Hu Yanping:First of all, I will try to define a key question that Chenxia just mentioned, that is, what comes after the algorithm? There is still an algorithm after the algorithm, but now it is obviously in a stage beyond the algorithm. This stage is developing towards a human-centered intelligent coupling stage. However, at this stage of intelligent coupling, both poles of the world are rising, intelligence is rising, computing power is rising, and algorithms are rising, but people's universal rights and equal rights, people's pursuit of intelligence, and people's interference with algorithms and anti-algorithm capabilities are also increasing.

In this form, if we analyze it from the underlying logic, the atomization of information is creating a new order. One of the atomizations is the original source, and the other source is the source in the sense of TOKEN. For example, Huang Renxun calls it the source of new intelligence, such as TOKEN and Sora's Patch. The world is being deconstructed into the smallest units and will be reorganized in the future. We must highly recognize the importance of people and their central position.

Because Agent is actually people-centered, I agree with Chenxia's point of view. The current situation has returned to a certain extent, and more emphasis is placed on people-centered development. Why is it more people-centered? From the underlying logic, technology is the primary productive force, and AI is the primary productive force of technology, but people are the primary productive force of AI. It may be a different situation when it comes to the stage of endogenous intelligence and autonomous intelligence, but at least it is the case at the stage of AI. So when people are the primary productive force of AI, we will see three situations.

The first situation:To a large extent, human spiritual life requires human attention, care, thinking, and watching. This is a process between people. At least in the AI ​​stage, the value and meaning of this world still focus on the relationship between people. Things are dependent on people, just as Harari said, "This society and many productivity processes that are happening in this society are actually based on the common imagination and mutual communication of human beings."

Second situation:Whether it is AIGC, digital people, live broadcasts based on intelligent avatars, or a number of creator groups, we generate code, audio and video, etc. In this process, I think there is a very important principle, which is human breath, human spirit, human thinking, human soul, and human interest. Only with these things can there be a stronger spread and the value density of the information density per unit content. The laws of intelligence are very strange, at least the first two are human laws. I believe that defining the future media form, content form, distribution channel, and the relationship between people and intelligence, and between people and the content generated by intelligence, is the third key factor in determining the future development of media.

The third situation:After diffusion, it has to go through a convergence process, which is the return after the bubble stage and this is also a process accompanied by entropy increase and entropy decrease. From the perspective of the new entrance, it is such a process that will improve efficiency again, solve people's bandwidth problems, and relieve people's anxiety about information fragmentation and information redundancy. I agree with Chenxia's point of view just now. Anxiety will always exist, just like information asymmetry will always exist, and the gap in information will cause people to have a kind of information pressure, which is the source of anxiety. At present, technology is solving this problem, which is why I agree with Mr. Yang Jian's optimism. Fundamentally speaking, it solves the problem, although it will be accompanied by some problems at different stages. But in the most fundamental sense, it helps everyone. Under the condition of limited bandwidth, because we have physiological bandwidth, time bandwidth, and people as carbon-based intelligence, the bandwidth of our intelligence itself. Under the condition of limited bandwidth, our efficiency and effectiveness are improved. Efficiency improvement is a quantitative process, and effectiveness improvement is a qualitative process. For example, when everyone has the functions of the end side, and everyone has a leap from 100tops computing power to 1000tops computing power, our capabilities are different and completely become two states. For example, our computing power, end-side intelligent bodies, hybrid intelligence, and cloud intelligence give each of us this ability, so that personal capabilities will leap to the state of digital people or even "super individuals."

Although we already have a cloud-based model with trillions of parameters, our end-side has the ability to run intelligence offline. We can perceive, think, act, create, and interact with everyone at any time based on the entire human knowledge base and the sensing and perception capabilities of the entire global network. However, the most important thing at this moment is still everyone's own thinking and judgment. This is also the topic I discussed with the leader of another technology media two days ago. If they want to go further, what do they need to do? Then the feedback I gave is three points. The first is the source, the second is judgment, and the third is people. What do you expect from you?

The source means that we cannot run after the traffic, but should explore the origin and the primary things. The second judgment is that people do not lack information, but what they really lack is effective judgment, especially personalized judgment. For example, in today's discussion, I especially want to hear what the teachers' opinions are, rather than the information I saw on a certain website, because judgment has a strong personality attribute. We did a survey many years ago. For example, when people want to buy something, what kind of media do you prefer to get information from? Unexpectedly, the first medium is people, not outdoor billboards, elevator ads or TV stations. Among the different crowd media, relatives and friends account for the highest proportion. This is an interesting discovery. We have developed to the era of AIGC, but humans still rely heavily on the opinions of relatives and friends when making decisions. This is determined by human nature. The second is judgment. What is the third point? What do people expect from you? You can't say how much traffic there is and how much value there is based on traffic. This approach is a bit off the direction.

Including platforms, content, and other aspects, we need to do what we should do and not do what we should not do, and maintain restraint. We must always reflect on one question: what do people expect from us? This is more important than paying attention to parameters and indicators. Behind growth, it may deviate from people's hearts. On the Internet, many MCNs and algorithm recommendation algorithms are criticized because they create traffic but deviate from people's hearts. This is a very strange relationship. I think it is also a problem to think about in the next stage of the industry. Because if you want to promote technology, there are too many things to talk about. Regarding the information explosion or AIGC, in fact, I should be more of an advocate of technology, but it may be the opposite. I think we need a certain convergence.

So today is particularly rare. I feel that many of the teachers' thoughts are on the same dimension. This also reflects the two poles of this world and is also the most important relationship in the future, that is, the relationship between man and intelligence. How the relationship between man and intelligence will develop in the future actually depends to a large extent on how people act.

Huang Chenxia:So Mr. Hu’s point of view is that we should go back to the source, back to judgment, and back to people’s expectations to get through the bubble period. I would like to ask Mr. Wen Da, whether it is the era of recommendation algorithms in the past or the AIGC era that has already arrived, what do you think should be used to get through the bubble period?

Wenda:Just now Chenxia mentioned the topic of traversing the bubble period, which reminded me of several major revolutions since the industrial age. From the steam engine era from the 17th to the 18th century, to the electrical era in the 18th, 19th and early 20th centuries, to the information revolution from the 20th to the 21st century, to our AIGC revolution today. These revolutions span a very long period of time, and I think they are all a process of accumulation and development.

AIGC has just started to develop, and ChatGPT will be launched in November 2022 and will have a wide impact in 2023. The situation in China is the same. In fact, the so-called "Hundred Model War" just started last year. I think this bubble period may last for five to ten years, and this process will cause great changes in many industries, because computing power requires a lot of energy consumption, and related sectors in the capital market have also changed a lot. Back to the level of our news, the super-individuals mentioned by Mr. Hu just now, the big names in various self-media, can they be regarded as super-individuals? The answer is yes, they output personal opinions, which are recognized by the audience, but this recognition is based on their personal charm.

First of all, the opinions of these self-media celebrities are based on their rich knowledge, sufficient knowledge reserves and strong logical thinking ability. They understand the needs of the audience, so they can attract a large number of fans, form unique individual phenomena, and even bring individual economy and individual wealth. However, as AI technology develops to the next stage, pioneers have begun to try further breakthroughs, such as Musk's brain-computer combination. I think in the next 5 to 10 years, or even shorter, brain-computer interface technology will be fully applied. In fact, Musk's original intention at the time was to treat paralyzed patients or patients with Alzheimer's disease. But let's imagine that when a biochip can eventually be implanted in the human body, as technology develops, I think this is inevitable, then everyone will become a super individual.

Sages and philosophers at home and abroad have actually foreseen this trend. For example, Nietzsche in the 19th century mentioned the concept of "superman" in Thus Spoke Zarathustra, and the declaration that "God is dead". In my opinion, it actually represents the disappearance of a current rule. It is possible that in the era of steam engines, the "God" of agricultural culture may have "died", and with it the formation of new social rules. Similarly, in the electrical age, new rules have emerged. Today, with the development of science and technology, we are gradually moving towards a super-individual era, where everyone has the potential to become a "superman", which will lead to changes in the rules of the world. For example, in the future, humans may transform from carbon-based humans to silicon-based humans, which I think is also an inevitable process. And once we reach that stage, the immortality we talk about may exist.

Perhaps at this time, the greatest luxury for the rich is death itself. Even death may become a privilege for a few, but spiritual worries and problems still exist. So in this context, it is really hard to predict what the news industry will become, but everyone still achieves the so-called technological equality, because after the chip is implanted, human beings themselves are an encyclopedia.

The next few generations will certainly face the above problems. Perhaps in the near future, we will see that "Six Days in Midsummer" may be shared by robots. Of course, for the current media, the first task is to face this reality and not avoid the changes around like an ostrich. Crisis and opportunity coexist. Only those who bravely face the crisis can seize the opportunity of transformation and be reborn. This is exactly the thinking that our media constantly instills in the team during the transformation process. Professor Yang just mentioned in the Tencent report that 80% of people have been exposed to related tools, and 60% of people hope to make such a transformation. But superficial learning and in-depth learning are two completely different concepts.

Now many people are still in the stage of just trying out new technologies, so I think that if the current media people, as well as those who want to work in the media or become self-media people, should embrace technology, face technology, learn technology, and simplify some of their daily tedious operations, and completely replace them with technology. Then through continuous knowledge accumulation, output your own unique views, because the formation of views requires deep knowledge reserves and broad insights, and use new technologies to speed up the learning process of knowledge reserves. Before we carbon-based people really transform into silicon-based people, we still need to work hard to survive in society and gain a place. The same is true for the media industry, so we must see future trends and face the current reality. So when the bubble comes, we should not avoid it, but accept it. For a small number of people, bubbles may mean risks, but for most people, especially the 60% to 80% of people mentioned just now, bubbles are not enough to pose a threat.

Huang Chenxia:Therefore, we must see both poetry and the distance, as well as the reality in front of us. I think "super individuals" may be an important topic in this time period. I would like to ask the three teachers, what do you think is the most essential difference between the current "super individuals" and the original self-media?

Hu Yanping:I think there are still differences between "super individuals" and self-media. "Super individuals" can be said to be some individuals who have become super first, while self-media is a form and structure of media centered on each individual at a specific stage of media development. The importance of individuals here is also relatively prominent, but it still has strong media attributes in essence. In many cases, people are still in the back and content is in the front. Of course, the most fundamental difference lies in the ability structure and ability range. The ability structure means that each "super individual" has its ability in the sense of computing power, and the intelligent body it uses gives it the ability in the sense of empowerment, thereby greatly expanding the scope of its creativity. In terms of the scope of ability, "super individuals" go beyond the scope of traditional media. Among the creator groups I studied, at least nine categories of groups have become highly active creator groups, and among the creator groups, 1/9 of them focus on traditional media forms such as text and graphic creation.

In addition to creators of traditional media forms such as text and graphics, the creator group also covers many scenarios such as music creation, film and television production, program development, model fine-tuning, and e-commerce. Therefore, creators are now a larger category than self-media, and of course there are larger categories than creators. In fact, intelligence allows everyone to become a new possibility, but the process of becoming a new possibility will have an intelligence gap at the beginning, including the "super individual" mentioned just now. In addition to its ability improvement, it actually also has the alienation of intelligence and the huge psychological pressure and anxiety brought about by the scenario of being a "super individual". But generally speaking, it will experience a diffusion of innovation, which is also a process.

At the beginning, only a few people can become "super individuals", but in the future everyone will be a "super individual". When everyone is a "super individual", it will reach a new balance. In this new state, there are still problems of human alienation, including the problem of insufficient universal equal rights and the problem of the gap. I think this is the problem that will be faced in the process of technological development of productivity. So from this perspective, I tend to be optimistic overall, but being optimistic does not mean that there are no problems. In addition to the equal rights and alienation issues we just discussed, there is another problem, which is the disturbance of the entire information field and public opinion field. Because in the stage without media, the whole society is actually like a Brownian motion, and each individual is an atom or molecule in it. The birth of the media has caused the society to begin to form waves, but in the era of traditional media, the nodes and locations of these waves are relatively certain, and the scope of individuals involved is also relatively clear, but the emergence of the Internet has magnified this phenomenon.

When it comes to the algorithm and intelligence stage, the waves may turn into stormy waves, or even a tsunami. To use a specific metaphor, in the Brownian motion stage, a media may only affect a few individuals. When it comes to the self-media stage, a self-media may be able to affect millions or even tens of millions of people.

However, when it comes to algorithm recommendation and intelligence, there may be 100 million people paying attention to the same thing in an instant, and 1 billion people discussing the same issue. Even across races, national boundaries, and geographical regions, everyone is concerned about the same point, and such an order is formed. Because order is actually a social container, it can accommodate some relationships, but when this energy field is so large, it is challenging to the existing order and container, because it will form a huge energy body in an instant, but this energy body is in a micro-aggregation state, it gathers in an instant and disperses in an instant, just like a flock of birds. This is an interesting state, but how can this state be guided by the situation and transformed into a driving force for social progress, a driving force for productivity innovation, and a driving force for progress, growth, and care in the humanistic sense, rather than an impact or destruction on the existing order. On the other hand, it is the same for individuals. How to maintain independence in the face of huge shocks and alienation, and at the same time maintain the independence of human beings themselves while facing the future, and at the same time maintain a state of positive energy between people and society, and between people, is also a very important issue. The form of the media is closely related to the form of the information field, the public opinion field, and the stability of the entire order container. So I think filing is a huge capability, and how to keep it in a state of continuous improvement in the future is also a problem. But overall, I am still optimistic, just like I am optimistic about "super individuals". At the beginning, it is a matter for a small number of people, but it will affect everyone in the future.

Huang Chenxia:So what does Mr. Wen Da think about the phenomenon of "super individuals"? Is the "Daily Economic News" currently making some new attempts to promote the development of this trend?

Wenda:Today we are mainly discussing our views from the perspective of the media, so we must first clarify what a "super individual" is. For example, the current self-media, such as Dong Yuhui and Guo Youcai, can they be regarded as "super individuals"? If we understand it literally, Dong Yuhui is an absolute "super individual", while Guo Youcai is a short-lived "super individual". We are not sure whether his popularity will continue, especially when talking about the impact of technology on the media. I think self-media people like Guo Youcai and Dong Yuhui have not completely deviated from the essence of traditional media. They rely more on personal charm, rich knowledge reserves and personal experience to form such a resonance on the Internet.

So I think a "super individual" must be a highly personalized individual armed with powerful AI technology. It must be the personal charm of the individual plus the blessing of technology. As for the extent of this technical support, it is difficult to predict at present. "Super individuals" may become more and more dependent on technology and eventually become alienated. Because Dong Yuhui is still a flesh-and-blood person in essence, who can communicate with you face to face, but when he is called a "super individual", he undoubtedly gets a special definition.

Secondly, I think that "super individuals" are still distributed in a pyramid form. There may be many "super individuals" at the bottom of the pyramid, but the one who stands out in the end is No. 1. In addition, there is a difference between becoming a "super individual" through the blessing of technology and some people's innate talents. For example, some people are born with a particularly good memory. In a specific era, they can accurately describe a certain event at a specific time in a specific era, while the rest of us cannot describe it very accurately. But if armed with technical means, this difference may be narrowed.

Therefore, as a traditional media with official characteristics, the National Business Daily should advocate a kind of collectivist behavior, encourage teamwork, and work together on news and programming. However, this era is forcing us to constantly transform. If there are no outstanding individuals or teams in the entire system of the National Business Daily, how can we talk about the influence of the media? Last year, we won the first prize in the public opinion supervision category of the China News Award. This is inseparable from the outstanding individuals behind the scenes, but it is very difficult to win more awards.

On May 9 this year, we released the AI ​​TV 3.0 version in Shenzhen and put forward a concept, "one person is a TV station". It is easy to put forward this concept, but it is a very challenging task to realize it. However, we at the Daily Economic News always uphold our commitment to our readers, and we are currently working hard to make our vision come true. The first two words of the eight-character policy of financial media we just mentioned are "professional". Therefore, we expect our reporters and editors to conduct in-depth research in the market segments of the financial field, just like Mr. Hu's in-depth research on a certain macro-level issue in the entire Internet technology field, and Mr. Hu has become an expert in this industry field, so he is a "super individual" in this industry.

Currently, the National Business Daily is committed to enabling every reporter to conduct in-depth research in the niche areas he is concerned about. In this way, it is possible for a reporter to eventually become the most authoritative expert in the media industry who understands the oncology industry. However, it is unrealistic to make reporters experts in the medical field because they have not received relevant professional training. But among the group of reporters, he can stand out.

Two years ago, we incubated a product like "N Xiaohei", which to some extent showed the prototype of a "super individual", but he was more like Dong Yuhui, who relied on personal charm and writing ability to gradually stand out and convert influence into income. However, the most worrying aspect of traditional media is that after you incubate such an individual, such a "super individual" may leave the original platform in order to seek greater benefits.

We had disagreements on the distribution of benefits, so we had to acknowledge the current reality and make concessions, so we finally reached a consensus that even though such a "super individual" left the team, he would still create content around "N Xiaohei". We also learned from this lesson, so in the "One Person is a TV Station" project of the "Daily Economic News", we decided to let all reporters try to break through themselves. Their professionalism depends on their personal interests and choices, and we use the existing artificial intelligence technology of the "Daily Economic News" to help them.

First:We assist journalists in building their personal knowledge base so that they can better accumulate and organize relevant information.

second:We build vertical models for journalists in their research fields to help them explore professional topics in more depth. These tasks cannot be completed by journalists alone, so our technical team intervenes to assist in building various customized models, so that each "super individual" has strong technical support as a backing, so that they can work and research more efficiently.

third:How to maximize the influence of "super individuals". We are now trying to reach cooperation with various platforms, such as our cooperation with Tencent to jointly develop vertical models in the financial field. Because we hope that the opinions output by each "super individual" can be quickly spread to major platforms in the shortest possible time, which is also impossible to achieve by relying on individual strength.

But with the emergence of big models and artificial intelligence technology, these tasks can be completed through the National Business Daily. This will make everyone's replication at the technical level exactly the same, whether it is for one person or a hundred people, the effect is the same. This year is also the 20th anniversary of the founding of the National Business Daily, so we plan to launch 20 so-called "super individuals" in December. Next year, we may further expand to 100. Over time, if some talents in the middle or top of the pyramid gradually emerge from these "super individuals", the transformation of the entire media will be successful in a sense. In the process of such a transformation, we will also clearly understand how to solve the relationship between the distribution of interests between individuals and teams. I think we need to use a contractual method to define it in advance.

Because the formation of a theory and its specific implementation in practice are different. It is a very good thing to have novel ideas, but in the end, the key lies in the various details of the actual operation. So I think that when we only have one "super individual" like "N Xiaohei", his departure may have an impact on our platform. However, once our platform successfully builds 100 or even 200 "super individuals", they all use our technology, and we fully consider how the "super individuals" can help them take technical risks when they fail, and meet their interests when they succeed, the market share of our platform will slowly expand. At the same time, we have many self-media partners. Although their communication channels are relatively single, these self-media do have unique advantages. In order to give full play to their potential, we will help these self-media with a certain amount of traffic to break through technical limitations and further enhance their value. Through our support and cooperation, these self-media will gradually develop into "super individuals" of our non-Daily Economic News reporter group, thereby jointly building a powerful "super individual" matrix.

We are currently promoting development according to this plan. But objectively speaking, it does take a long process to sort out ideas and finally achieve this goal. So we also extended an invitation to Mr. Hu. We are willing to use the technical strength of the "Daily Economic News" to first help Mr. Hu become a "super individual", at least so that Mr. Hu's content can be quickly spread on various platforms. If Mr. Hu does this himself, he may need to register an account on each platform separately, which will be cumbersome. And we are very familiar with the algorithms and rules of each platform. In this way, Mr. Hu's content is more likely to achieve good dissemination effects on various platforms. It is very difficult to create content that achieves full marks on various platforms. Such content needs to be finely polished, but if it is only done well on one platform, some platforms do not publish content, or the content on some platforms is close to full marks, then the overall efficiency and benefits are completely different.

Huang Chenxia:The relationship between "N Xiaohei" and the Daily Economic News mentioned by Professor Wen Da does remind people of the relationship between Dong Yuhui and Dongfang Selection in the financial field. The rise of "super individuals" will indeed bring challenges to the existing relationship between organizations and platforms, triggering the reconstruction of order and changes in the interest distribution mechanism. Professor Yang, I hope you can analyze it from the perspective of the research institute. From the newspaper stage to the television stage, there are actually "super individuals". We can see that the various hosts of the TV stations back then, from a certain perspective, are actually "super individuals" in the current sense. What do you think of the current phenomenon of "super individuals"?

Yang Jian:At the beginning, I didn't fully understand the concept of "super individual". It is still a relatively new concept for me. After listening to the sharing of several teachers just now, I feel that "super individual" is also hierarchical. We can subdivide "super individual" into different levels such as small superman, medium superman and big superman. Some people may like small superman, the one who attracts multiple groups may be the medium superman, and the one that all groups like may be the big superman, but the one who can maintain the audience's loyalty for a long time may be the super superman. It is not just a solution with many curves together, but a set of solutions interwoven by multiple factors, and even new solutions appear every day. In such a situation, it is quite difficult for a natural individual to have all these abilities at the same time. So when we encounter this situation, we will say that this "super individual" must be supported by a team, technology and capital, which is usually true. But for capital, technology, and team, it is not easy to find a suitable carrier. "Super individuals" need to adapt and integrate into the model that supports their development. This is not only a test of their personal abilities, but also a challenge to the coordination ability of the team and resources behind them.

Not everyone can easily become a "super individual". It is precisely because of this that the situation mentioned by Professor Wen Da and Chen Xia occurs. When the influence of an individual increases to a certain extent, the original team or platform may be difficult to bear. This is because the relationship between us and the supported individuals is a personalized relationship, and it has not yet achieved a universal relationship. For example, as Professor Wen Da said, when many super groups can be incubated within the "Daily Economic News", it is not just a personalized relationship, but a common relationship. I think that by building a common relationship, the contradiction between the "super individual" and the platform can be better resolved.

Huang Chenxia:The reason I just talked about this topic is that this year, I have frequently heard different people mention "super individuals". The expressions of the three teachers have also inspired me. So as to how to define "super individuals", I have two opinions:

Second point:At present, there is a huge centralized traffic, which may cause a certain individual to quickly receive widespread attention in a short period of time. When many viewers notice this individual, his personal charm is outstanding or a certain expression happens to touch the audience's emotions, he may transform into a "super individual" at a certain stage. But as Professor Yang said, "super individuals" have different levels, and the duration of influence is also different. This is one of the meanings of "super individuals" in my opinion, such as the impact of the "celebrity effect" on the general public. But there is also a kind of "super individual", in which individuals are empowered by many technologies. For example, individuals originally have limited knowledge of certain knowledge, or know little about certain fields, but now through the future brain-computer interface or existing intelligent bodies, they can quickly master and accurately express complex knowledge. In fact, technology is empowering many individuals and giving them abilities that they may not have had in the past. When many individuals acquire these new abilities through technology, they have the opportunity to transform into various "super individuals" and then become people at the top of the pyramid. In this process, the individual's personal charm and human characteristics are the key factors that determine whether a "super individual" can stand at the top of the pyramid. This is the inspiration I got from the views of several teachers just now.

In general, in our discussion today, first, we explored the development trends and future prospects of a new round of large-scale model technologies such as AIGC. Then, we deeply analyzed the actual application of these technologies in the current media industry, as well as the differences between macro trends and micro rhythms. In addition, we also discussed the bubble period that may occur in the process of technological development, how to cross this stage with the right mentality and method during the bubble period, and whether there is a bubble period at all.

Because the college entrance examination just ended a few days ago, many people are standing at a new stage of life choices. There is a hot topic that has attracted much attention and has been on the hot search, which is the point of view put forward by teacher Zhang Xuefeng. Teacher Zhang put forward his view on the "uselessness of news", which aroused widespread discussion at the time. Of course, I do not agree with this view because it narrows the news industry too much. Based on thinking about the future, I invited three teachers to make some suggestions. If you give some advice to students who are preparing or studying journalism and communication, what abilities do you think should be focused on in addition to focusing on yourself?

Yang Jian:First, let's talk about the choice of major. It is true that before and after the college entrance examination, many people would consult me ​​about whether their children should study abroad, which school and major to choose. After AIGC was launched, I think the professional field has been reshuffled. Some majors that were considered practical and popular in the past may not be as promising as expected in the future. Relatively speaking, those majors involving fields such as mathematics, philosophy, sociology and anthropology are more promising. Because these majors are more about helping humans survive in the world and giving individuals meaning in the world, these majors may be more valuable and popular than those practical majors. Secondly, let's go back to the issue of journalism. As I mentioned earlier, when artificial intelligence becomes more and more powerful in processing world materials and information, it needs more first-hand information for processing.

In a sense, the biggest task of journalism is to provide the truth of the facts, or at least to provide strictly screened information materials. From this perspective, I think journalism still has broad development prospects in the future. If these young people want to reserve skills for their future and pursuits, I think young people should learn to think about problems from the essence earlier, just like the first principles mentioned by Teacher Hu. Whether teachers or parents, they should guide young people to start from the first principles and have a deeper understanding of the world and the world's expectations of them. Only by thinking more about problems from this perspective can young people grasp the direction more accurately.

Wenda:I have been engaged in journalism for more than 20 years, almost 30 years, but I am not a journalism major, so I often say "journalism has no academic background". The journalism industry does not overemphasize professional background. I have seen many excellent journalists, and they are not all from journalism majors. Similarly, we rarely see famous writers in the world who graduated from Chinese departments. Not all writers who won the Nobel Prize in Literature came from the Faculty of Arts.

First point:I think there are some basic principles that should be followed when working in the journalism industry. First of all, you should be passionate about journalism. Journalism is not a job just for personal income. I think the journalism industry will continue to develop because humans, as carbon-based creatures, always have curiosity and pursuit of the truth. As long as this curiosity exists, the journalism industry will definitely exist. Therefore, I think that when choosing to work in the journalism industry, passion is the most important factor, rather than being overly concerned about your professional background.

Second point:As a journalist, we are not doing it for the so-called "face". Because journalists have more or less some party A colors, what kind of content we publish will determine what the audience will see. And journalists are generally reporting, not being reported, so it may give people a strong sense of host-guest relationship. But I think that a journalist should pay more attention to watching and observing the society, and convey these observations to our audience in a calm and humble manner, rather than mixing them with selfish desires. So I often say that the requirements for journalists on the moral level should be more stringent, because we cannot assert that the moral standards of journalists must be higher than those of other industries, but compared with other industries, the requirements for the news industry on the moral level should be higher. Because the speech of ordinary people may only affect individuals, but journalists represent institutional behavior, which will have some public instrument colors. Therefore, journalists must be in awe and must not abuse public instruments. If we can use public instruments reasonably, we can form a very positive impact and spread widely. If we use them improperly, it may cause a severe tearing and backlash effect.

Third point:Journalism is an endless learning process. This goes back to the curiosity I just mentioned. As long as everyone has such curiosity and is willing to continue to learn, I think journalism is still very challenging. Moreover, this challenge not only brings a high sense of personal achievement and pride, but also due to the penetration of technology, some people who did not simply work in writing in the past can find their value in the current journalism work, such as the technical department, video department or design department. Therefore, I think the journalism industry is still very meaningful. Finally, you can make a lot of friends, because after all, people must travel thousands of miles and meet a lot of people, so that you can have greater opportunities.

Hu Yanping:I think this question is both grand and subtle. We need to start from the big picture and start from the small details. This is also true for media talents. Therefore, I would like to make four suggestions: First, we must have a very clear sense of ecological orientation. Second, we must have a strong perception and judgment of the situation. Third, people’s spirit and personality charm. Fourth, to use AI well, we must clearly know what is happening in this era? In which direction is this era heading? What are the changes in the media industry?

First point:As mentioned above, from media to self-media and then to "super individuals", this is one dimension of change. From news information to UGC and then to AIGC, this is one dimension of change. From the masses, niches to micro-masses, this is another dimension of change at the individual level. At the same time, we must also focus on the future rather than stop at the present, so this is the first point, to establish a sense of ecological orientation.

Second point:The ability to perceive and judge the situation is very important. This ability enables us to have a deep insight into this era, accurately grasp the common concerns of the public, and at the same time, select truly valuable content from information consultation and pass it to the audience to meet the public's demand for news.

Third point:Human spirit, personality charm and care for people. As we mentioned before, this is still something that needs to be done by people, not by intelligence. The spiritual life field of human beings fundamentally depends on each person. Even if you become a "super individual", it does not mean surpassing humans or being above others. It may be that everyone cares about others, pays attention to and cares about others, which makes this group of people interact. And fundamentally, it is people who give AI spirit and soul, rather than AI alienating people, so AI is a tool to give people abilities.

Fourth point:We must make full use of the capabilities of AI technology, because it does improve efficiency and create new possibilities. It is not just about transforming news information from a handicraft to an intelligent one. In fact, it contains very profound changes. That is, the transformation process from algorithm recommendation to intelligent coupling reflects the return of the importance of human beings and re-establishes the value concept of people-centeredness, considering both efficiency issues and issues at the level of human spiritual life. Therefore, no matter what era the media is facing such a field, making good use of such capabilities is a great process of liberating productivity and people. Fundamentally speaking, the empowerment of AI not only expands the scope of what we can do, but also provides a historic opportunity.

This is the shackle that we humans have not been able to break free from for thousands of years. It is because of the relative backwardness of productivity that we have to make a living every day, and even hundreds of millions of people are still trapped in poverty and famine. The emergence of intelligent technology can be said to be the first time in history that humans have been provided with the key to break through this shackle. Intelligence not only optimizes production relations and distribution methods, but also enables the great changes in productivity to benefit all mankind. So we are at such a juncture to examine the relationship between AI and humans, the significance of AI as the first productive force to the liberation of mankind, and the role that media, information, and news can play in this process. So I think today's discussion is very meaningful. Although it may be a moment in the trend of this era, it is at a critical time point that connects the past and the future. Including our discussion of "super individuals", everyone will be a "super individual" in the future. "Super individuals" are not rigid and unchanging existences, but because of everyone's choices and orientations, the influence of "super individuals" can become very large or very small in an instant. The rigid order is obviously going against the trend. In today's discussion, the views of several teachers are very inspiring to me. I think this discussion has opened up a new perspective on how to face the future, even at the methodological level.

Huang Chenxia:Thank you very much for the sharing of all the teachers. Just now, Teacher Wenda mentioned that he has been working in the journalism industry for nearly 30 years, and I have also spent 21 years in the news media field. Although I am not a professional graduate of journalism, I am still willing to explore in depth the many different factors in the news industry at this point in time.

First point:Demand, the demand for the journalism industry is expanding, not shrinking.

Second point:As several teachers mentioned, the current development of technology has brought new empowerment to the news industry, which also brings some challenges to the education of future journalism schools, such as how schools can use interdisciplinary knowledge to cultivate students with more diverse skills.

Third point:Although many universities or colleges emphasize cultivating students' curiosity and questioning spirit, I think journalism majors play an important role in this field.

In the previous discussion, we explored the relationship between intelligence and human nature. So I think the core of human nature lies in having a strong curiosity and questioning spirit, which can constantly stimulate our deep thinking ability. I think the ability to think deeply determines how people and intelligence will be further integrated in the future. In other words, what kind of future will the integration of carbon-based and silicon-based lead to? Thank you again for the participation of all the teachers. I am very happy to share our views with the three teachers in such a beautiful place. Finally, I hope that we will always maintain our curiosity and questioning spirit, and continue to be the night watchmen of society under the empowerment of technology.