Successful Storytelling: Hollywood vs Anime

I must admit, I watch more anime than Hollywood movies. I came across an interesting YouTube video which made me think more about why. But then it made my think about implications for my hobby project of trying to tell deeper stories using AI video generation.

The Video

The video presents a point of view of why they believe Anime is now doing better than Hollywood. I have not verified the details behind the claims, but it presented a framework that I found interesting.

The financial model the video put forward on the success of Anime is as follows:

• Stories start as Manga - printed stories, fairly cheap to produce

• Stories with traction get Anime adaptions - animated, colorful, a few million dollars to produce

• Stories that survive multiple years get movie adaptions - higher quality, released in cinemas, larger budgets and have generally made solid profits

There are multiple levels of gates to reduce risk.

“With so many Marvel backstories and TV shows it now feels more like homework to keep up!” That line resonated! However does it also undermine the claim of why Anime is doing better? Would not cheaper TV production help identify stories that resonate making movies more popular? So the framework is interesting, but not sure I am convinced its why Hollywood has had failures.

Where I also hesitate however is Manga and Anime also have a bad reputation of terrible working conditions and low pay for many of the animators. So I am not going to sit here and say they are perfect role models.

A different point made is Hollywood got influenced by funding sources to create movies. It got tempting to tell stories ticking boxes for the investors instead of focusing on the story itself. The story had to compete with other obligations. With Manga/Anime, lack of public interest killed stories early. The result has been there are stories with deep and complex themes (e.g., Attack of the Titans). That makes more sense to me - if you want success, focus on what the audience wants and forget everything else.

So that leaves me with:

• Good stories trump big budgets

• Trying and testing multiple stories is a lower risk path to success than all-in bets

Storytelling and AI Video Generation

I have been working on a hobby project to me create stories - series - over a long period of years, part time. I have been working on variations of this for years frankly, but with AI coding tools it has really sped up recently.

What are the challenges?

• I think the story matters. I want to tell stories that resonate emotionally.

• I want to tell long-term series that span years. Keeping track of the details over a long period can be hard, especially if thinking about multiple at once.

• I don’t have much spare time. I want a tool that helps me be efficient as I want to tell a story.

• I have no interest in sharing AI generated stories - I want to be the storyteller.

• When I learn a new technique, I want help applying to all my projects. AI has been great here, capture rules in AI review agents that can critique what I create.

Ego check and disclaimer. I say the above, but I have not created ANY long term series to date. The above are desires that may never be achieved, but that is almost secondary to me. I want to create because I want to create.

And there is another distraction in the mix. I am a programmer. I also enjoy creating code! I find it an artistic outlet in itself, so I often get more distracted by creating tools than creating content. Oh well! ;-)

Bringing things back to the above video, what feels “right” to me in what I am currently building (whether I finish it or not) is it should make it easier for me to manage multiple stories in parallel. I have been building in AI assisted tools to spot errors and inconsistencies, so I don’t have to remember every detail myself. While I am coding, I still daydream mapping out scenes with interesting moral dilemmas. AI is making it easier to create content, easier as an individual to get reviews on demand, and easier to then create final video clips.

I will say the last step, the AI video generation, is both improving rapidly and the biggest weak point so far. Control and emotional depth are still hard. The best tools I have seen involve a human actor and then replacing them with another character. But maybe that is a good approach. The platform does all the boring legwork, humans create the ideas and movements that convey emotion.

The plan is then to publish a series of short episodes on YouTube, and let viewer demand influence which ones I put more time into.

Wrapping up

After watching the above video, the ability to create multiple series in parallel has become more important to me. That means I need even less overheads to create and even more assistance to remember all the details on the project. The good news is I was worried I was asking users to enter too much information into the system in order for AI to help. Now I worry if I am asking for too little!

Now all I have to do is finish the code.