In this article, we want to discuss the most disruptive new media in the age of AI – creating a consistent character. A consistent character is a model you can place in various forms of media. It can operate as a company mascot, your virtual avatar, or a fictional person.

One example of these web ads popping up is where they show an old person claiming to retire and selling off their handcrafted leatherwork inventory for cheap. This old person was created by AI and it’s a mail order shop in China selling Temu-grade men dress shoes. This is an example of a virtual representation in marketing.

You can create a model to perform various roles, such as an AI music artist, a VTuber, a webtoon character, or a chillhop study character. This is a whole new paradigm shift, especially in Black media. What’s even awesome is content creation can be automated and there are both enterprise and entrepreneurial opportunities never seen and can operate as a digital nomad or abroad or in the hood.

You are the new Black media. We already discuss other forms of AI media and the opportunities but mostly all of them come from creating a consistent character. Let’s discuss how to start creating your own media and marketing characters using this technique.

We will discuss how to create a consistent character, techniques for video, comic strip, streaming, and tips on automation and running a business. However, there is a bonus – I have a referral link to NightCafe AI platform that will give you 20 credits from the link below:

https://nightcafe.art/ru/TheManufacturedSolution?refsrc=share

The free 20 credits is enough for you to get started on creating your consistent character, and they have a collection of the best genAI models on the market, including Nano Banana for editing. The nice thing and important thing – log in every day and get your free 5 points because they add up. Every 5 days in a row, you get additional 25 bonus points. I still have my thousands of credits because I only need the free points.

Creating Character Sheet

This is all you going to need to get started and have your character – you just create one image. In the prompt, write something like “create a model character sheet with front view, side view, and rear view of [description of your character]” and it should create a character sheet. I used Flux Krea for this generation, but you can use Flux Pro and HiDream models.

Seriously, this is all you need to get started – now think about what I can do with her. I can create a comic strip with this character, I can have her feature in PSA ads telling young sistas to stop dating knuckleheads and thugs, I can use her as the brand ambassador for a E-course line of self-learning charging $700/course. Or she can narrate like Dora the Explorer on a kid animated video stream.

See the opportunity and new paradigm of Black media? Yeah, I thought you did.

Nano Banana for Editing

This is a screenshot of our PZR009 “Minato Lady” demo – this is not the final version. But noticed what I have done.

First, I created a character sheet using a head and shoulder portrait. At first, the character sheet was only a black and white line hand drawing, and I ask Nano Banana to colorize the model with a pastel palette.

Notice in my prompt, I typed “Show me the model of the woman in the middle shopping at a luxury handbag store using the same anime style of the original drawing” and this created the image below.

This is the secret sauce because this allows you to create scenes from the character sheet. Noticed I said select the middle woman, but you can also prompt “a composition of the character sheet” which will combine the front, side, and rear view to make one character in the final scene.

Now, this is also you need from the character sheet – you create your script with scenes and put your characters in these scenes using Nano Banana. In NightCafe, after the image is generated, you will see a “Prompt to Edit” below the image, click on this and you can select Nano Banana.

These two things are all you have to do to build everything else in this article. There is an additional feature on NightCafe where you can combine multiple images like a café background and the character and additional items and generate a new scene with those multiple images into one. A good example is you create a storefront background image then you add new characters on each new advertisement doing something different in the store, from different angles. Or an e-course with different presenters in the same classroom and different text on the whiteboard for each course.

Create Video

This is an image that will be used for a chillhop stream. It will showcase a sista that moved to Southeast Asia and she planning her next moves. Her movement is supposed to be loopable and here is the technique you want to do, similar to the DJ video I forgot to mention.

Create the base image. Notice she got one hand on her chin and the other handwriting. For the loop, an additional photo will show her putting the writing hand on her chin, the end photo.

Create the video with the start and end image. The video platform will ask for the start and end image which will be the handwriting as the start, then both hands on her chin as the end. It will create the video between the start and end.

Reverse the start and end to loop it. Now what you do is use the original end video as the start video, then the original end video as the start video. What you just did was the ability to splice the original end video to the new start video of the same image of her with both her hands on her chin, now you have a longer loop.

You can use free video software like Blender, but I use Adobe Premiere, and you can layer the music tracks underneath. There are plenty of free music sources to get chillhop tracks and you just add them to the audio track of the video software and copy/paste the video looping sequence.

You can run this with a donation link on streaming platforms for monetization. You also can sneak advertisements in this between loops, and you can license this video to be streamed at a chillhop café.

Create Comic Strip

With a comic strip, you create the same character sheet and just use Nano Banana for each scene. This is really straightforward. Be mindful that a lot of prompts need to fine-tune and it will be some imperfect generations. People type in additional prompts such as no extra hands or heads and specify the number of people in the prompt.

Here I’m using bubble text in the prompt after I create the scene I want to move forward on. So create the scene, then when you happy, create the bubble text or you can do both in the same prompt to save tokens and credits.

Create VTuber Stream

To create a Vtuber character, just go ahead and make your character sheet. Then what you need to do is sound to video. But here is what is important – you want to use a voice genAI platform like Wondercraft or ElevenLabs. I believe Microsoft CoPilot has voices as well as Gemini but not sure.

It is also possible to download a text-to-speech engine to your desktop using ComfyUI and create your own dialog and script there. Or you can use your voice and record it then run it through the sound to video as an audio file.

I do not recommend the live voice format unless you are running a live stream show that accepts donations. It is a lot easier to run pre-programmed shows but there is another reason. You want global syndication – you want your VTuber talking in different languages all around the world, at the same damn time, generating as much revenue as possible.

So what I did here is I did not use Nano Banana, I used SeedDream image editor and asked it to a head and shoulder portrait shot against a white background. The reason why is two reasons. The first reason is I can put any kind of background behind her to replace the white background in Nano Banana. The second reason is Wan 2.5 sound to video need a portrait face and it is hard to detect a person within an image like the image below.

There are some editors that can do sound to voice in the image above and I haven’t tested them yet. This was done using Gemini Veo but the thing is you cannot control the voice selected and does not support sound to video. What I recommend is use these videos for the intro video showing her in action or walking into the studio with theme music.

Using N8N and APIs

For the upcoming newscast where we can have a real reporting show like Bloomberg or CNN, we are looking at automation to pull and summarize the latest news, then use AI to format the text to be read on a teleprompter then the AI characters will speak using sound to video.

Please note that I will be using headshots to speak on each segment like the real news. There is technology that will allow both to speak to each other but I’m still testing the technology at the time of this writing.

We are planning to use n8n for the workflow and it can communicate with AI through APIs and return the audio file. Then we call the sound to video API that generates the news/story segment, all automated.

Noticed we literally built a newsroom, weather reporting, sports reporting on a cheaper budget than a legacy news reporting station using AI and consistent characters. You do know this is going to be displayed on our Toshikiso Hub jumbotron.

Yeah, you sure you figured out who is going to be running Black media and tech mogul stuff? Yep - you and I, not these legacy Black media – we the Black media now.

Creating a Media Company

You can create your own media firm as a startup idea. The key is to just outsource the AI stuff to these platforms and with the NightCafe link, you can get started right now with free credits if you don’t have a platform. You can always check in every day and get 5 free credits. Now imagine everybody on your team doing that and you collaborating on what image/scene/video to create.

You can set this up as a student club or off campus as an incubator startup. You can do this from your kitchen table as a side hustle with a day job. Or you can move to Thailand and use one of the common meeting rooms in the hi-rise apartment complex and hookup with some chooms and collaborate. You got serious options to create the new paradigm of media.

You can create marketing campaigns, commercials, advertisement, TV shows, VTubers, comic strips in Korean, Japanese, Mandarin, Spanish, Italian, German, English and so on. This is the next generation, and I hope yall realize the new paradigm we are in right now.

You got everything to get up and running – enough credits to create a character sheet, edit scenes in Nano Banana, convert to video scenes or sound to video scenes, or create a comic strip, or chillhop video stream. The rest is up to your self-ecosystem and your unique individualism to bring to the table for the world to see.