To facilitate your discussion and sharing, please click "Follow" before reading. I will update the quality content daily, thank you for your support~
Wen 丨Sun Hu's study
Editor丨Sun Hu's study
Research Today, the Internet, artificial intelligence and big data technology are in full swing.
In order to quickly and efficiently complete the production of disseminated content, many media professionals have adopted armed high-tech intelligent technology to further optimize the quality of service in the media industry.
In the media field, artificial intelligence will have great potential, and many technologies such as speech to text, robot writing, speech synthesis and artificial intelligence anchors can be widely used in the process of interview and manuscript writing.
This kind of integration and innovation of media technology not only improves the professional level and intelligence of the media, but also makes a significant difference in people's lives.
Although artificial intelligence has been widely used in many aspects such as news media and editing, data analysis and audience needs, it also faces the problem of irregular use.
First, technological innovation
In terms of technological innovation, with the continuous upgrading of artificial intelligence technology, the speed of updating artificial intelligence anchors is relatively fast, and the technical deficiencies are constantly updated, which can meet the technical requirements as much as possible.
In addition, due to the support of AI technology, the development of artificial intelligence anchors will become the culmination of AI technology, integrating the cutting-edge scientific research results of AI technology for services in the radio and television industry, and reducing the technical pressure of traditional hosts as much as possible.
Although AI has made news media reporting more efficient and more effective, replacing human journalists is an almost impossible task.
Because in the news reporting process, artificial intelligence can give help and support, because robot-made news can only be used for simple news reports, can not produce more in-depth news.
Relying on artificial intelligence technology alone cannot make a reasonable explanation for complex and changeable situations, and it is also difficult to effectively grasp and discover the original face of things hidden behind the news.
This is the most important responsibility of media practitioners, and it is also something that cannot be achieved by current artificial intelligence technology.
Traditional hosts are slow to develop in terms of technological innovation, and they still bear technical pressure in the face of program hosting and news broadcasting, and need to constantly meet the needs of programs and audiences, but traditional hosts also have an irreplaceable role.
Although with the continuous development of artificial intelligence technology, huge amounts of data can be deeply mined and collected, the news content generated according to fixed algorithms can be consistent with the creative experience of media professionals.
And in the process of dealing with various social issues, for the current robot writing, the new news media reporting methods are not flexible enough, and the information lacks in-depth mining and processing.
In terms of technological innovation, artificial intelligence anchors and traditional anchors have their own advantages and shortcomings, and only by combining traditional media with artificial intelligence technology to achieve the perfect combination of "people + technology" can we produce news reports that meet the needs of the public.
Second, the aspect of abstract reasoning
Artificial intelligence anchors supported by artificial intelligence technology can analyze and summarize according to big data well in abstract reasoning, and then form a certain news material chain, and form interest and orientation push.
In addition, for some competition or puzzle TV programs, artificial intelligence anchors can predict the results of the competition in advance according to the number of votes cast by the audience or the performance of the contestants, and form the influencing factors of program ratings through the comparative analysis of big data.
In addition, for consulting or information TV programs, artificial intelligence anchors can fully assume the effect of program hosting.
Using the setting of AI technology to import relevant information into its intelligent system, the AI anchor will fully assume the host function of communication and communication.
However, the inability of AI anchors to analyze human minds is the biggest flaw found so far.
This flaw leads to many shortcomings in the process of in-depth thematic reporting or interaction with humans.
For traditional hosts, the ability to deal with emergencies is the most basic requirement, and announcers and hosts are constantly improving their ability to deal with emergencies.
The situation reproduction skills used in broadcast programs require the host to have good cross-domain reasoning ability, who can reproduce the scene in the brain according to the person, event and environment in the case, and express the corresponding feelings and attitudes based on this.
This skill requires the host to have good associative ability in order to substitute himself into the event and spread the true feeling to the audience.
In order to improve their cross-domain capabilities, human beings have found a variety of ways to think, including: (1) sorting out the information in the event according to the direction of thinking; (2) Carry out logical reasoning according to the basic situation of the event, and find corresponding evidence to verify it; (3) Understand the event from the perspective of the protagonist of the event, etc.
However, this does not mean that traditional moderators have these abilities to do all the hosting work.
Due to the rapid update of TV program types and the increasing requirements for program hosts, cross-border hosts or science literacy program hosts require hosts to have more professional knowledge, which is far less than the big data skills of artificial intelligence anchors in the process of abstract reasoning.
In addition, the summary in the process of abstract reasoning by the host will also have a certain orientation, which in turn affects the audience.
Therefore, deviation in abstract reasoning is irresponsible for the audience and unqualified for traditional moderators.
It can be seen from the above that consulting programs can give full play to the advantages of artificial intelligence, but there are many shortcomings in the programs that need to be analyzed.
Therefore, how to take advantage of the strengths and avoid the weaknesses is the key to the current research on artificial intelligence.
Third, the aesthetic aspect of art
Human aesthetic abilities are unique and at the same time cannot be replicated by artificial intelligence, quantifiable or explained in words.
But from an aesthetic point of view, you can clearly judge the ugliness and beauty of art.
Everyone's aesthetics are different, and there are certain differences in the criteria for judging beauty, but it can be described in words, and people tend to share each other's feelings and experiences of beauty.
Similarly, artificial intelligence does not have such capabilities, but aesthetics is a basic ability that hosts and anchors must have.
Aesthetics is a kind of inheritance of beauty, a part of language art, only with sufficient aesthetic ability can effectively convey this art.
At present, artificial intelligence anchors still cannot feel the existence of beauty, and cannot explain and express it.
However, the host and announcer are an artistic expression in themselves, and the core content of understanding the manuscript cannot be expressed without aesthetic ability.
Scene reproduction, inner language, and object sense are the three basic techniques for broadcast creation, but they are not an actual carrier, but an abstract concept.
The microphone placed in front of the announcer is the bearer of the skill, and it is used as a tool for imagining, and the facts of the manuscript mobilize one's own feelings to express the sense of the object.
The lack of aesthetic ability and abstract ability determines the fact that artificial intelligence cannot understand the manuscript and mobilize its own emotions.
If the effect conveyed in this situation is bound to make the listener feel impersonal, neither ear nor heart.
Fourth, sentiment analysis
At present, the hottest focus on artificial intelligence is the emotion analysis technology, as long as it is provided with enough training samples, artificial intelligence can learn this technology, which can analyze the basic emotions of human beings according to human facial expressions and body movements.
In this regard, Mr. Lee Kai-fu once said that emotional analysis is the weakness of artificial intelligence, and machines cannot have rich emotions like humans.
But from the perspective of technical research, machines can understand and express emotions, but they cannot feel emotions.
When faced with the same manuscript, different announcers will generate different scene reproductions and express different feelings.
The classic saying "a thousand Hamlets in a thousand eyes", but what the machine learned from previous training samples is still temperatureless and emotionless.
The artificial intelligence anchor Xiaoice entered the anchor room of "Look East" in 2015, and held an appointment ceremony before that, this format itself does not have much practical significance, mainly to achieve the purpose of program innovation and artificial intelligence promotion.
When the artificial intelligence anchor Xiaoice appeared in "Look East", many scholars studied it as a case.
Based on the basic situation of the Xiaoice, SMG Media Center created an unprecedented full-course artificial intelligence anchor program "Xiaoice Shake It", which was broadcast from March to September 2017.
In the program, Xiaoice shows the lively and interesting attitude of a host in both tone and content, and the effect is better.
However, this situation does not appear in other programs, and artificial intelligence anchors often show dullness and boredom in these programs, showing a sense of strangeness in the face of emergencies.
From a technical point of view, there are often two completely different concepts of technology worship and fear, and the technical concept that the host has is often expressed through his tone.
No matter what type of variety show it is, the host will show the audience an aura that integrates public anchors and popular anchors, and the demonstration effect is obvious.
The dialogue between the host and artificial intelligence often conveys an attitude of human beings when facing new things, and does not convey strangeness but expresses "friendliness".
Like a friend from afar, it conveys kindness and freshness, and demonstrates the correct technical concept for people.
The difference between mass media and artificial intelligence is the embarrassment of Xiaoice, in the program "Xiaoice Shake It", Xiaoice can use big data to sort out the focus topics of the day, collect hot comments from netizens, and fully have the ability to interact with traditional hosts.
However, the shortcomings can be found in the hot topics and topics collected by the Xiaoice.
The first is the lack of timeliness, and the information obtained is often news that has been broadcast 2 to 3 days ago;
Second, the news it gets from the hot list has a orientation problem and cannot be broadcast in the mass media.
It is these two problems that make Xiaoice face embarrassment, just the tech show, people laugh that every sentence expressed is edited by humans in the background.
AI can only spread messages that can be easily automated such as weather.
Hosting and broadcasting are inseparable from people-to-people communication, so emotional communication is crucial, and the audience feels the feelings of the person in his speech.
Even if artificial intelligence can understand and express human emotions, but it cannot produce empathy, such expression must have no temperature, it is difficult to enter people's inner world.
As Go master Ke Jie said after the dueling with the robot: "Learn its moves, and then let's see what our weaknesses are." ”
Oxford and Harvard University concluded after interviewing 254 experts in the field of artificial intelligence that by 2026 AI will have a medium to high level of writing.
However, when it comes to "writing", it is less likely that AI will replace the host in the next 10 years.
As far as "broadcasting" is concerned, the speech synthesis technology of artificial intelligence has been continuously upgraded and innovated, and now it has basically realized pause, continuous reading and accent annotation, but the emotional description and prediction ability of emotional speech synthesis needs to be improved.
bibliography