See part 1 here:https://justabout.com/content-creators/my-ai-driven-content-creation-journey-from-idea-to-reality-part1

After setting up my channel, I dove into creating my first video. I wanted it to be a trip through each decade of the 20th century, crafted entirely using AI tools. Here’s how I went about it.

Conceptualizing the Content: For my inaugural video, I decided on a series of 10 four-second clips, each representing a different decade of the 20th century. My first stop, unsurprisingly, was ChatGPT Plus/4. I asked it to generate an image that encapsulated the essence of the 1900s decade.

From ChatGPT to Leonardo.ai: Once I received the initial image from ChatGPT, I inquired about the specific prompt used to create it. The quality of the images from DALL-E isnt what I am looking for (although it is the best at interpreting your natural language and generation the image prompt) so I took the detailed prompt and fed it into Leonardo.ai, my chosen service for this paticular video.

Refining the Vision: This process was repeated for each decade. Over a few days, I generated and sifted through hundreds of images, tweaking settings and the prompts, refining my approach until I had the set of images for each decade that I was after.

Bringing Images to Life: Next, I ventured into animating these static images. I used a mix of motion, animation, and movie mode features from various AI image providers. Due to the unique nature of AI generation, each run produced a different animation, influenced by the seed number and settings used. This phase was all about experimentation until each image was animated successfully. AI video is still very new and experimental, so it is very much a case of trail and error over and over (and over!) again until I got something acceptable. This took a long long time!

First Steps with CapCut:

This was uncharted territory for me, as I had never edited a video or engaged in similar creative work before. Using the CapCut app on my phone, I found it very intuitive. I quickly learned how to add cool transitions and text overlays to each of the four-second clips. This was the only part of the process where I was on my own with no AI help to make up for my lack of knowledge and skills.

Avoiding Music Licensing Issues: While CapCut offers music options approved for use on TikTok, I needed something suitable for YouTube. I headed over to the YouTube Studio audio library on my desktop, sorted by runtime, and listened to the selection of royalty-free YouTube-screened and approved tracks. I selected a one with a Creative Commons license, which required me to include the license details in my video description. Integrating this track into my CapCut edit on my desktop rather than my phone was a little more complicated but not really too difficult.

Final Touches with Topaz Video AI: After exporting the video from CapCut, I ran it through a local installation of Topaz Video AI. Although an expensive purchase, it was on sale, another product I probably shouldn't really have bought but I liked the idea of for future use (it's a one time purchase). Instead of upscaling to 4k for sharpness, I left it at 1080p (I don't think youtube shorts run at a very high res anyway), instead using Topaz's interpolator to increase the video to 60fps. This added extra frames and smoothness, mitigating some of the choppiness inherent in early AI animation tech. I avoided resolution upscale as given the nature of the images a higher resolution brings out even more of the imperfections and flaws rather than making it look better.

Uploading to YouTube: The next step was to create an engaging title and description for the YouTube short. Once again, I turned to ChatGPT for assistance. It initially came up with some SEO-heavy suggestions, but I tweaked these to make them more natural rather than focussing on SEO.

With the title and description set, I uploaded the video to YouTube. It was a straightforward process, but as a first-timer, every step felt new and exciting!

My first video is live:

Finally my first ever content was live on Youtube! Even though I used AI as much as possible in reality this took the best part of my time off work over the xmas break and took a lot of manual effort and learning. Hopefully I can create my next one much faster now I know what I'm doing.

I hope this run through of the steps I went through serves as a useful guide and motivation for people like me who have never done anything like this before and it proves useful for someone to get over that big barrier of getting up and running and making your very first ever video! Please like share and subscribe on my video below as they say! lol