Artificial intelligence prevalent, what should I do?

Home > Sci-Tech

Artificial intelligence prevalent, what should I do?

2021-11-26 00:12:53 4 ℃

With the continuous development of technology, AI artificial intelligence has been infiltrated into people's daily lives, and there has been a certain impact on all walks of life. In the 2018 World Economic Forum report, it has been mentioned that: "By 2025, the robot will undertake 52% of the work in the workplace." And the professor of Tsinghua University also has the same idea. He said that "artificial intelligence will be in the near future," Will replace a large number of people's work. "

However, in many industries, the most profound industries are mostly affected by the most profound industries. Because artificial intelligence and cultural industries have mutually promoted positive effects, therefore, various AI cultural products have emerged. They can simulate the sound of the live-person, or you can do some basic work. This makes them have a very high commercial value, and has become the popular choice of advertisers, and the perfect replacement of live anchors.

In addition, the dubbing industry has been artificially intelligent. In various short video, the frequency of use of AI dubbing is getting higher and higher. They use the painted sound of the mechanized sound line to vomiting the picture, thus achieving the formation of "one more fun" in single video, increasing video viewability. This video-style fire has greatly contributed to the raging industry in the dubbing industry, and even some dubbers will receive a low price of AI's voice.

More and more videos do not "vocal" challenge the nerves of language workers. Is AI's continuous development, would you really let the dubbing actor even have been replaced? Is the 52% of the work that Ai will bear?

First of all, it is undeniable that artificial intelligence does have many advantages that many of the traditional dubbing are not possible:

1, low cost, high efficiency

Because the traditional dubbing industry has always been manual dubbing, this must have time consumption and not enough accuracy. If you use the AI ​​artificial intelligence technology, you can effectively reduce human cost. Not only that, since Ai dubbing is training the unsorted data of the target sound, it can be simple and fast to generate a full audio of a smooth standard, avoiding the pronunciation errors such as pronunciation when manual dating, greatly improved Efficiency of dubbing.

2, rich in tone, high work

The sound of Ai dubbing is composed of "climbing data", so as long as there is enough acquisition sample, it can change the tone at will, combined with different scenarios to distribute different sounds. The British startup Sonantic has issued a video called "Faith_ First Ai That Can Cry".

On the basis of "climbing data", the voice of this video also traces the micro-gap of human sounds by training robots to distinguish between standard machine sounds with robotic sounds. Acoustic and expression emotions. And this emotional understanding is that the machine passes through the analysis of the text, and the adduct of the work is extremely high. It is also because it is not possible to completely express the core due to insufficient understanding of the work, not accurate enough.

In addition to these advantages, Ai dubbing is still unprinted, that is, it can rely on vocal synthesis techniques to restore the voice of the dead, let some people who have psychological trauma caused by painful people and lover Heal. Make a lot of congenital language barriers have their own voice through technology synthesis. It can be seen that AI dubbing not only has an impact on cultural industries, but also has a positive contribution to the medical industry.

But whether this represents the AI ​​can replace the dubbing actor, become the main force of the dubbing?

The result is not exhausted. Although now in technology, AI synthesis audio can't hear too much difference on the surface, but it still has many problems compared to vocals. For example, when AI dubbing is listening carefully, you can hear it sometimes the adhesion between the words and words is too mechanized, the stress is not clear, and the cavity is too flat.

Secondly, because each time the volume is maintained very stable, it violates the aspect of social communication to a certain extent, and it is easy to make the audience to have aesthetic fatigue. Not only that, but because Ai dubbing is actually "imitation", it is difficult to "innovate". It can't show emotions that have not been entered, so I said that in some emotional expressions, the true feeling of vocals is that AI cannot do. So, what are the advantages of voices?

1, feel more abundant

As we all know, in addition to the documentary, all other types of film and television works require dubbing actors to perform art interpretation with extremely high emotions and rich feelings. Many times, when the dubbing actor is in full care, the mood will reach a burning point. At this time, the energy in his voice is that Ai dubbing cannot be comparable.

Good artworks are often passionate. AI dubvical can be used according to people's speech tone and the law of syllable changes, but this technology is still not mature. Therefore, the emotion of Ai is "there is no emotional machine", how can it compare with the real people with explosive power?

2, more innovative, able to sublimate the work

Most of the voice in dubbing is a sham, it has both elaboration of the content, and has self-emotional and experience. In other words, dubbing is the secondary creation of the work, and is the active output and expression of the view. But this, AI is obviously can't do. It is true that AI can imitate the external skills of dubbing expressions, and do it, the rhythm is accurate, and the tone is slightly changed. But there is no subjective initiative, and it can only rely on the data algorithm to "the drum to be painted", and it is impossible to reflect the "charm" in the sound.

The only content itself in Ai dubbing is itself, and the true voice has already integrated these content into his own life experience, and it is sent. In this level, the advantages of vocals are self-evident.

3, style type more diverse

Everyone's voice has its own unique style label, or steady and temperament, or calm magnetic, or height, or let people listen to "hungry". The bubble is stronger, and it does not have any personal characteristics. It is always a plate. It is easy to lose its freshness based on the algorithm according to the algorithm.

In short, complex humans have thousands of things per minute, so people are also full of emotional, but Ai dubbing is always only a bunch of ice-cold data. Then you said, can the voice may be completely replaced by AI?

But even so, we can't relax your vigilance. Although people need more realistic realistic, more creative, making the listener in a comfortable and true state. But with the continuous improvement of this technology, maybe Ai artificial intelligence has also broken these 桎梏, threatening the voice of people. So, we must keep the intention of learning, constantly progressing, always motivating and spur yourself, let your voice "evolution" to unreplaceable.

Course recommendation cooperation plan

If you like our courses

And want to recommend it to the partner

Want to know / learn more dubbing dry goods

Winter holiday offline training dubbing class

Details reference poster

Line under training 丨 假 配

Line set training 丨 base & sound

From the mouthpiece / / from the heart

Let us feel the charm of sound

Sina Weibo: Nine-party name

Himalaya FM: 9 channel

Submission email: [email protected]