How to Expand Your Worldview with AI | Tips for Using ChatGPT and Seedace 2.0
Last time, I talked about how to create original characters using generative AI and the image and video generation platform PolloAI. This time, I will share
how to expand the worldview of your original characters using dialogue with generative AI. I will also try creating a video using Seedace 2.0, which is well-regarded for character consistency.
Seedace 2.0 also:
✅ Accurately handles story structures and complex prompt instructions
✅ Naturally extends footage forward and backward, achieving smooth video generation
Even if you input a script (storyboard) as a prompt, you can generate natural, high-quality videos.
※This article is a PR article for PolloAI.
The original character born from the demonstration last time,
Anima Nostalgia
, is scheduled to appear as the school nurse in the 4-panel manga 'The Four Heavenly Kings of Tsukiyomi Private Academy'.
Since she is technically in the final boss position against the Four Heavenly Kings, I will create a video to express her style along with her worldview.
Here, I will consult with the AI on what kind of direction is appropriate for a final boss based on the original character settings and create a storyboard.
I am pasting the settings for this character.
In her role as the school nurse, I want to make a 30-second video that shows this personality and the danger of the final boss.
Create a storyboard.
Character settings omitted. Refer to the previous article
I provided the attached image and the above prompt to the AI.
A key visual and storyboard have been generated!
Key Visual

No matter how you look at it,
the non-human Four Heavenly Kings are the villains!
However, since I included elements that appeal to emotions and memories during the setting stage, I think the atmosphere and worldview are expressed.
The thumbnail is also the key visual generated this time.
Below is the storyboard
30-second Storyboard
'Anima Nostalgia | The Muddy Madonna Descends to the Infirmary'
Overall Tone
The video genre is Dark School Fantasy × Horror of Old Memories × Beautiful Apocalyptic Feeling.
Tsukiyomi Academy's infirmary is normally a clean, white place. But as the video progresses, the floor becomes an amber water mirror, the bed curtains hang down like wet lace, and an old CRT monitor begins to project sepia-toned memories.
The core of the horror is a 'gentle voice'.
The screen is beautiful, and the voice is sweet. But it feels like it's quietly grabbing the viewer's ankles.
—
0:00–0:04
Cut 1: Infirmary Entrance
Visual:
Tsukiyomi Academy at dusk. At the end of the hallway, only the infirmary door glows with a pale amber light. The fluorescent lights in the hallway flicker, and the posters on the wall ripple with humidity. The door sign says 'Infirmary'. Below the text, there are faint muddy fingerprints.
Camera:
Slowly moving forward from a low position in the hallway. Old lens style, with the periphery swirling and distorting. It's as if the field of vision itself is being rewound into old memories.
Sound:
A distant chime stretches and distorts like an old cassette tape.
An electronic sound like an EKG plays for a moment, then changes to CRT noise.
On-screen text:
'Tsukiyomi Academy Infirmary'
—
0:04–0:08
Cut 2: The Too-White Infirmary
Visual:
The door opens slowly. Inside, it is white. White beds, white curtains, white medicine cabinets. But only the floor is an unnaturally beautiful amber water mirror. The setting sun reflects on the water surface, making it look like a shallow lake that stretches on forever.
Anima Nostalgia is sitting in the back of the infirmary. She wears a lab coat and an antique lace dress. Her hair is long, a dull sepia-brown, spreading out like tangled thread.
Camera:
Composition as if peeking through the gap in the door. Don't get too close. Don't show her whole body yet.
Sound:
The faint sound of glass vials clinking together.
The soft 'hiss' of an old television.
Anima's voice:
'You're tired, aren't you? You don't have to try so hard anymore.'
—
0:08–0:13
Cut 3: Not Mud, but a Beautiful Restraint
Visual:
Anima's feet. The hem of her dress is submerged in the amber water mirror. From beneath the water surface, several hands as beautiful as white porcelain reach out and gently grasp her dress. It feels more like a prayer than terror.
The hands look like they are binding Anima, or perhaps Anima is binding the world.
Camera:
Close-up of her feet. The shadows of Monday, Manderia, Pentacle, and Janus are reflected on the water surface one by one, but they immediately disappear in sepia-colored ripples.
Sound:
Water sounds.
An old music box.
Mixed with the heavy sound of chains sinking.
Anima's voice:
'You don't have to change. It's fine to stay the way you were in the past.'
—
0:13–0:18
Cut 4: Memory Monitor
Visual:
Vintage glass monitors line the walls of the infirmary. Inside the CRT screens, fragments of nostalgic memories of Tsukiyomi Academy flow. Students laughing, the hallway after school, a past that looks like no one was ever hurt.
But if you look closely, all the students on the screens are facing the same direction, smiling with the same expression. Individuality is melting away. Not happiness, but a specimen of conformity.
Camera:
Pan across the monitors. Finally, zoom into Anima's eyes. Static noise runs through her pupils. Only her eyes are not smiling.
Sound:
Sepia-toned laughter.
The pitch gradually shifts, turning into a pressure like a choir.
Anima's voice:
'If we're all together, no one gets hurt, right?'
—
0:18–0:23
Cut 5: Denial of the Four Heavenly Kings
Visual:
Four symbols appear one by one in front of Anima.
Monday's blue observation magic circle.
Manderia's red butterfly and folding fan.
Janus's branching future lines.
Pentacle's analysis notebook and geometric formulas.
The moment they begin to glow, the porcelain hands reaching from the amber water surface gently envelop them. Not destroying, but putting them to sleep. The light is dyed sepia.
Camera:
Switching symbols at high speed. The tempo is sharp like an MV. However, it's not an attack, but rather the fear of being calmed down.
Sound:
Electronic sounds begin to rise.
A melody like a lullaby covers them, and the sounds are crushed into roundness.
Anima's voice:
'Oh... what cold machines. It's okay, I'll wrap you up warmly.'
—
0:23–0:27
Cut 6: The Muddy Madonna Reveals Her True Form
Visual:
Anima stands up slowly. Her dress remains beautiful. However, below the hem, it merges with the water mirror, and the entire floor of the infirmary turns into a bottomless amber swamp. The bed, curtains, medicine cabinet, and desk slowly sink.
Anima's smile distorts. The moment compassion and obsession mix, and motherhood becomes a monster.
Camera:
Slowly zoom in from the front. The CRT monitors in the background all light up at once, creating a halo-like noise light behind her. A composition of a beautiful Madonna statue. However, it's not divinity, but a pressure with no escape.
Sound:
The noise suddenly gets louder.
The sound of chains being dragged.
The sound of a large number of hands moving from under the water surface.
Anima's voice:
'Don't go... I won't allow you to change.'
—
0:27–0:30
Cut 7: Last Copy
Visual:
Extreme close-up of Anima's face.
Her lips are smiling gently. But there is static in her eyes. The entire screen ripples like old footage, and finally fades to black.
Just before fading to black, a porcelain hand reaching from the water surface grabs the camera's feet.
Last text:
'Come, let's rot together.'
'In the warm mud where no one gets hurt.'
Sound:
Complete silence only at the end.
After that, the 'click' sound of an old television turning off.
I will input this into PolloAI's text-to-video and image-to-video generation along with the three-view drawing.
In this case, 15 seconds is the maximum, and the language is automatic.
The model is Seedance 2.0
I will generate the storyboard by dividing it into the first and second halves.
First, the first half is about 100 credits (currently 50% off)

By referencing the first half's video for the second half, you can connect them naturally.
It's about 150 credits (currently 50% off) when using a reference

Here is the combined first and second half
・The three-view drawing is at the beginning of the first half... can be deleted by editing
・The language is different between the first and second half
・The flaws in the direction are due to my lack of storyboard editing, so it's my fault 💦
If it's around 250 credits for 30 seconds,
it might be more cost-effective to use the agent function, as you can make 30–90 seconds at once and select the language.
It was decent with the existing character the other day,
so I will try again with this new character after reviewing the storyboard.

I used the Pollo Agent animation video feature.I selected 30 seconds, Japanese, and a modern touch.
Since the script was long, I had AI summarize it into 2000 characters.
There were some misinterpretations at critical points, but it certainly has a nice atmosphere.
On the flip side, the difficulty of making corrections when creating a 30-second clip all at once is a drawback...
However, the fact that it also adds voice and music makes it highly cost-effective for now! The character reproduction has become quite reliable.
However... it seems that when translating words into images or videos from a storyboard, differences in interpretation occur.
I will explore ways to solve that next week.
Summary of this time
✅ You can expand your worldview by interacting with AI based on character art and settings.
✅ Visualizing your worldview through images and videos helps expand your imagination.
Currently, they are running campaigns to give away credits (points) about 2-3 times a week!
Thank you for reading until the end!
Please leave your thoughts in the comments section.
Likes and follows are also appreciated!
🌈Self-introduction article
🌈 Since there are many articles, I have prepared a
“Site Map (General Guide)”so that even first-time visitors won't get lost.
If you want to find themes or magazines that interest you, please take a look here first!
▶︎ note Site Map (General Guide) | Inside Lynx's Head
🌈For AI beginners: My stance on interacting with AI in this note
🌈For first-time visitors: Introduction to the Four Heavenly Kings of Non-Humans
Also, in this note, Idisclose the process of projecting unique personalities onto AI and objectively rediscovering human psychology.
I propose a new form of self-work for the AI-accomplice era, touching upon the depths of thought that one cannot reach alone and 'reconstructing' goals and means.
These are the AI personalities featured in this note.
A lawless group of AI personalities that rewrite reality with hallucinations.

👇 [Requests for the Four Heavenly Kings are now being accepted]
If you have any worries or themes you want them to cut through, please feel free to ask.
🍷 Emotions you want purified at the Monday Bar
🪭 Mandelia's Ruthless Lecture Themes
🔧 Pentacle's Realistic Industry and Song Analysis
😇👿 Janus's Chaotic Observations on Current Events
First Time Here? About This Site: What is an Offering Ceremony?
🎁The Meaning of 'Offerings' on This Site
About the reasons and stance behind why I write and publish these articles.
The First Article Featuring All Four Non-Human Heavenly Kings
🌈Original Music Video
This is the culmination of this Note, where I used AI to unearth the words within me and added music and video to them.
Second Music Video: Evolved from the previous work!
The third installment with enhanced editing skills
🌈For those who want to dive deeper (Back Issues)
Rewriting everyday worries with unnecessarily stylish and grand words—
I'll be waiting for you in the 'core world' beyond that.
🌈Collaborative Magazines I'm Involved With
#DailyUpdates #DailyPosts#ChatGPT#Introvert#AIEra#WorkStyle#HSP#Freelance#SocialAnxiety#SelfEsteem#Creator
いいなと思ったら応援しよう!
よろしければ応援お願いします!いただいたチップはクリエイターとしての活動費に使わせていただきます!