"Your video suddenly exploded.
Many people want to use AI to make YouTube videos, but they don't know where to start. It's actually not difficult. There are many AI tools nowadays, so you just need to know how to choose and use them.
Let's talk about selecting tools first. To make YouTube videos, you need three AI tools: one to write manuscripts, one to dub, and one to make pictures. You can use GPT for writing, ElevenLabs for dubbing, and Pika for screen generation. These tools are quite mature and easy to use.
Pay attention when writing. The script of YouTube videos should be as natural as chatting. Don't make up with those words of "empowering" and "reshaping". The audience will be tired. When writing, use multiple paragraphs, and each paragraph should talk about a key point. For example, if you want to talk about "How to use AI to make videos", you can divide them into four paragraphs: "selecting tools","writing manuscripts","dubbing" and "pictures". This way, the audience can hear clearly and you can write quickly.
Dubbing requires selecting the right voice. ElevenLabs has dozens of sounds to choose from. Try it first and choose one that sounds comfortable. Be careful not to speak too fast or too slowly, but it must be natural. For example, if you choose a male voice, the speed of speech can be more stable, and the female voice can be more lively. Try it a few times and feel it.
Picture generation requires some thought. Pika can generate videos, but the pictures generated by AI are sometimes unreliable. For example, if you say,"A person is running alone in the park," it may give you a space run. So you need to be specific when writing prompts. For example, it writes,"An Asian man is running in a Tokyo park, the sun is shining, and the camera is following people." This way AI can understand what you want.
Editing is also important. Video clips generated by AI may vary in length. You have to cut it with Climbing or CapCut to make the picture coherent. For example, if there are 10 sentences in the dubbing, you divide the picture into 10 paragraphs, one sentence matched by another. This way, when the audience watches, they won't feel like they are jumping around.
It's release. YouTube loves viewing completion rates. So make sure you catch the person in the first few seconds. For example, use a suspense at the beginning: "Your video suddenly exploded. Do you know why?" When the audience is curious, they will look down. Put more dry goods in the middle, such as "writing with AI is three times faster" and "choosing this tool for dubbing saves money." You can leave a hook at the end, such as "Next issue will teach you how to increase AI fans." This way the audience will want to watch the next episode.
There are many people making AI videos nowadays, but few are actually doing them well. The reason is that many people only use tools and ignore the content. AI is to help you save trouble, not to help you lazy. You must also be able to tell stories, catch hot spots, and interact with the audience. None of these AI can help you.
So, learn how to use AI tools first, and then learn how to do content well. For example, you can try using AI to make two videos a week, one generated purely by AI, and the other with AI and a real person appearing on the scene. See which data is good. If you have good data, practice more.
Also, watch more what others do. Some bloggers on YouTube specialize in AI and making videos. For example, Marques Brownlee talked about AI editing, and Deepak Singh talked about AI writing. Watch their videos and learn their skills.
In short, it is not difficult for AI to make YouTube videos, but it takes time to do it well. Start with tools first, and then break through with content. Step by step, your video will explode. "
