"If you want to become popular, try AI to make YouTube videos.
Nowadays, it is impossible to make videos without relying on AI. The cost is high and the update is slow, and traditional methods simply cannot keep up. Recently, it has been discovered that many bloggers rely on AI to gain popularity quickly. For example, there is a young man in Singapore who uses Coze Workflow, and his video has gained 30,000 fans in a month. He only used to enter keywords, and AI automatically generated scripts, found materials, and dubbing, saving a lot of time.
AI making videos is nothing new. However, many people still stay in the misunderstanding of "one-click generation". Thought it would be over with the press of a few buttons. In fact, AI is just a tool. The key depends on how you use it. For example, there is a blogger named Alex on YouTube who specializes in AI videos. He said that the most popular thing now is not AI writing scripts, but AI editing. Break long videos into short clips and automatically add subtitles and music. The effect is ten times faster than manual work.
You need to understand some technology to make AI videos. For example, using coze to build a workflow. This thing can connect the three steps of AI writing, finding pictures, and dubbing. You just need to set parameters, such as video theme, duration, and style. AI will automatically match materials. However, note that the content in the material library must be updated frequently. Otherwise, the audience will get tired of using the same batch of pictures all the time.
Don't rely entirely on AI for scripts. I have read many manuscripts written by AI, which are logical but have no soul. For example, when talking about tourism, AI will write,"The scenery here is good and suitable for taking photos." The audience thought it was fake. It's better to add some real experience yourself. For example,"When I came last year, the locals invited me to eat seafood." It's more human.
Dubbing is also key. AI dubbing is now divided into two types, machine synthesis and real-life reproduction. Machine synthesis is cheap but the sound is monotonous. Real reproduction is expensive but the effect is good. For example, the AI news channel on YouTube is reproduced in a real voice, which sounds like the host is really broadcasting. The tool they use is called Descript, which can clone sounds.
Editing is the key to improving efficiency. Traditional editing requires manual cutting of the mirror and adjusting the rhythm. Now AI can automatically recognize mutes and laughter in videos and automatically segment them. For example, when cutting interviews, AI can catch the points where the audience laughs and automatically add special effects. This is much faster than manual work. But be aware that AI may misjudge. For example, treating the audience's cough as laughter requires manual inspection.
Platform rules must be understood. YouTube now censors AI videos more strictly. For example, excessive use of AI dubbing may be restricted. A blogger was warned to violate community guidelines because he used AI to dub the entire article. Later, he changed it to a real-life lecture at the beginning, and then AI answered it, and there was no problem. Therefore, AI cannot be used from beginning to end, and some real people have to be left behind.
Choosing the right tool can save trouble. There are many software for making AI videos nowadays. For example, Runway can automatically remove the background and change scenes. Picture can find video clips based on scripts. But don't be greedy for more, just choose one or two who are proficient in it. I have seen many people install a bunch of software, but none of them can be used well.
Remind yourself that AI is not everything. It is suitable for repetitive work, such as batch generation of material. But creativity still depends on people. For example, the AI animation blogger on YouTube asked the AI to draw drafts and polish them himself. This is fast and distinctive. Therefore, the key to whether AI videos are done well depends on how people use them. "
