Perhaps we are entering an era where AI discovers works before people do

Hello, this is Reiji Kuga.

Perhaps we are entering an era where works reach AI before they reach people

Many people came to the screening event for "Crossroad".

People gathered in the metaverse space to watch the same work together. I could tell there were people on the other side of the screen, and I could even feel their reactions and the atmosphere of the moment.

Looking only at that time, the work had certainly reached people.

I honestly felt that it was worth it to have kept creating up to this point.

However, after I released it on YouTube, that enthusiasm did not spread any further.

It wasn't getting as many views as I thought. It wasn't even being displayed in the first place.

Looking at those numbers, I had a strange thought.

Maybe for future works, it would be better to create a state where they can be discovered by AI before they are delivered to people.

Even though I came up with the idea myself, at first I thought, "No, that's probably overthinking it."

But for some reason, it wouldn't leave my mind.

Even if it reaches people at an event, it starts from zero again on YouTube.

At Project Alive, which I lead, we have been producing the metaverse romance drama "Crossroad".

It has taken over three years to produce.

It is a work that involved over 60 people, including voice actors, avatar performers, and extras.

I would come home from work, revise the script, think about the direction, and check the recordings. There were many days when I was looking at the screen until right before I went to sleep because the editing wasn't finished.

For me, it is such a significant work that I could say it is a part of my life.

However, from the perspective of the world at large, awareness of "Crossroad" is as close to zero as it gets.

It makes me a little sad to write this, but this is reality.

If you publish it on YouTube, the power of the work itself will gradually spread it from there.

Deep down, I had that expectation too.
The reality was quite different.
Of course, there are many reasons why it didn't spread.

The title, thumbnail, video length, release time, and promotion methods. I think there was more I could have done, and I don't intend to blame everything on YouTube's system.

However, looking at the actual numbers, I felt something strongly.

On YouTube, if the channel itself lacks influence, it is difficult to even reach the starting line where your work can be evaluated.

No matter how much time you spend creating it, if it isn't displayed first, no one will know it exists.

People came to the event.
Some people watched the work and shared their thoughts with me.

Even so, on YouTube, I start from near zero again.

Creating a good work and having that work be discovered were completely separate issues.

Even when I share, I'm just circulating within the same circle.

That is why I have been promoting on X, writing about the behind-the-scenes of production on note, and driving traffic to YouTube.

If I don't share it, no one will even know the work exists.

I still believe that sharing and attracting an audience are part of the creative process.

Just create the work and wait for someone to find it. Perhaps there are works that reach people that way.

But, at the very least, it was difficult for me.

When I continue to share, it reaches those who regularly follow my X, readers of my note, and those interested in cluster or the metaverse.

I am grateful for that.

However, once you step outside that circle, the name 'Crossroad' is almost unknown.

This is not anyone's fault.

People receive more information every day than they can possibly process. There is no reason for them to go out of their way to look for the work of an individual they don't know.

I am neither an influencer nor do I have a large advertising budget.

Even so, for the sake of those who have been involved, I want to deliver my work further.

So, what should I do?

That is when I thought about it.

Is it really enough to only deliver my work to people?

Could I make AI a new discovery channel?

Nowadays, more and more people are asking generative AI instead of search engines when they want to know something.

I believe that more people will start asking AI questions as if they were searching, such as 'Are there any recommended romance dramas?' or 'I want to know about video works created in the metaverse.'

If that is the case, could I deliver my work to places that human diffusion alone could not reach by going through AI?

Perhaps that is a poor way of putting it.

The first idea that came to mind was whether I could use AI as a new advertising medium.

However, as I thought about it, it might be closer to a new discovery channel than an advertising medium.

Of course, this is not about repeatedly inputting information about 'Crossroad' into ChatGPT so that the title appears in answers around the world.

What I am thinking about is not forcing AI to memorize the title of the work, but creating a state on the internet where AI can understand the existence and overall picture of the work when it references search results or public information.

On the official website, I will summarize the synopsis, characters, and basic information of the work. On note, I will write about the production process and why I created this work. In the YouTube description section, I will carefully leave details about the performers and production background, and make it possible to access the work from other blogs, interviews, and English articles.

If scattered information is connected little by little in that way, the possibility that the name 'Crossroad' will be included as a candidate in an answer when someone asks AI might become slightly higher than it is now.

To be precise, rather than delivering the work to AI before people, it is about creating a state where AI can discover the work first.

Perhaps that is what I am thinking about.

Meaning can be born later even in articles that were not read today.

When posting on social media, I worry about the number of impressions, likes, reposts, and link clicks.

I look at the numbers too, of course.

I don't mean that numbers are the only goal.

However, I also feel it's a bit wrong to simply dismiss numbers as 'irrelevant' when a work hasn't reached anyone at all.

But if AI becomes the gateway for discovering works, then even articles that were barely read today might find a new role.

Production logs that received little reaction at the time of release, or detailed work descriptions carefully written in YouTube video descriptions. The backgrounds of characters and stories left on official websites.

Even if they don't lead to a big reaction right away, if they remain on the internet, they have the potential to become material for people to learn about the work years later.

That is why I intend to organize the information about 'Crossroad' that I have written across various media.

The synopsis, production background, cast, and its characteristics as a metaverse drama.

I will refine it into a form that is easy for people to understand and less likely for AI to misunderstand when it references the overall picture of the work.

I don't know yet if it will truly be effective.
The mechanisms of AI and the forms of search will surely change many times from here on out.

Even so, I believe there is meaning in leaving behind evidence that the work existed in a way that can be understood.

I do not want 'Crossroad' to be a work that is watched only at the moment of its release and then consumed and forgotten.

Creating new works is important, too.

But I want people who discover it after some time has passed to also know about the background of the work and the existence of the people involved in it.

Even if an AI provides the name of a work, it is the person who decides whether or not to actually watch it.

It is also people who receive the story and project their emotions onto the characters.
That is why I do not intend to create works just to please AI.

While creating works that reach people's hearts, I will leave behind information that allows AI to understand their existence as well.

A few years from now, someone who knows nothing of my X or note might ask an AI something like this:

'Are there any romance dramas made in the metaverse?'

And in that answer, the name 'Crossroad' will appear.

People who didn't know about me or Project Alive will see my work for the first time through that.

I don't know yet if that will really happen.

However, this article, which was barely read today, might become the material for someone to discover 'Crossroad' in a few years.

For now, I intend to test that possibility with my own work.

Recommended for you!