I started editing videos around the same time I started my writing career, just over ten years ago. My idea was that the videos could act as a kind of decoy to get people excited about the content of my first book. In the beginning I only edited shorter videos and then moved on to longer ones. About four years ago, I started using more AI to edit my videos, including the addition of AI-generated narration. Before that, the technology didn’t even allow it, or the synthetic readers available were still deafeningly metallic and monotone. My first such video with natural narration was my October 2020 Warning of the coming Antichrist and his global climate revolution!

Everything else about the editing of my videos, however, has been meticulous craftsmanship. That is, I write the narrator’s lines myself, find suitable background music to enliven the narration, add visual elements such as stock footage (video or still images such as news headlines), which should usually relate to what the narrator is saying, or be in some way in sync with background music (a common effect in movie trailers), and add other effects such as the often used gradual zoom. Images and videos should not change too jerkily like slides on a screen, and so the editor should also choose the transition technique by which one image or video changes to another (personally, I have generally preferred the least distracting fade effect). In simple and inexpensive editing software, those functions are usually automated, so there is no need to do the fading yourself.

The most time-consuming part of the whole process has usually been searching for background music and stock footage (images or videos) from paid services such as Storyblocks and the like. For a half-hour video, you have to add quite a few songs and video clips on the topics you are talking about. In addition, you often have to search for material from YouTube videos. For example, for a few seconds of footage, you sometimes have to download an hour-long YouTube video and soon your hard drive is so full that your computer overheats and “crashes” and nothing works properly anymore (luckily, work in progress is usually saved automatically even if the software crashes). And the time spent on all that can add up to over 100 hours for a video longer than 20 minutes.

So to save my time and nerves, I decided to try to do the whole job this time using AI. In services like Invideo.io you can, for a monthly fee, have an AI create videos based on nothing more than your dictation of what kind of video you want it to make for you. So you can generate the video’s text, narration and all other visual content with simple commands. And you can also make it generate content in any language. So what do you need a human being for anymore? As someone who has used the service, I have to say that it still needs people to fine-tune the work.

For example, AI seems to add new background music to the narrative at the start of each new sentence. This gives you a lot of options when it comes to choosing background music, but at the same time it makes it sound rather restless if the song changes every five seconds. So this is where a human can edit and fine-tune the AI’s work. Similarly, if I am not happy with the video image selected by the AI, the service allows you to find another one to replace it (as with music). Similarly, the narrator’s text can also be edited manually.

I was most impressed by the voices of the narrators, who are no worse than the presenters of big-budget TV documentaries. I used more voices for the project because the AI also seems to select them automatically based on the content of the text. I identified the first voice as that of the famous Sir David Attenborough from BBC nature documentaries. Attenborough is in fact a notorious neo-Malthusian (an advocate of human depopulation) who has had close relations with English royalty, who also share his anti-human philosophy. So apart from the Elon Musk interviews at the beginning, and other fine-tuning, the entire 38 minute video is entirely the work of AI.

It would probably take me 200 hours to manually assemble a video this long. With AI, I could do it in just over 20 hours, ten times faster. This is exactly why AI will lead to the “age of abundance” heralded by Elon Musk, because time is money and the more people and companies are able to grow products and services in the shorter time, the faster the overall wealth of society will grow (although I don’t earn anything from these video products). The content of the video is based on the same text I published in my previous blog post.

The idea that machines will replace human occupations and eventually make people redundant is not a new perspective, by the way. This was the thinking also 500 years ago when Gutenberg’s printing press replaced the craftsmen who copied books by hand. The only difference with the past is that today machines can also do jobs that require great cognitive intelligence and creativity, such as writing, composing or editing video (and often better and faster than humans do). It is therefore very uncertain what role humans will play in the age of artificial intelligence. Well perhaps my video below will provide some answers to that question.

---

---

---