In seiner Funktionalität auf die Lehre in gestalterischen Studiengängen zugeschnitten... Schnittstelle für die moderne Lehre
In seiner Funktionalität auf die Lehre in gestalterischen Studiengängen zugeschnitten... Schnittstelle für die moderne Lehre
A short AI-animated film about memory, blame, and perspective.
A couple finds a fallen plant in their apartment. Each of them remembers the night differently. Between their two versions, the film explores how memory bends, how blame begins, and how care can return when certainty disappears.
The starting point was the question:
Can the same small domestic accident become two different stories depending on who remembers it?
The project uses a simple everyday situation: a plant on the floor, soil spilled, two people looking for an explanation. Instead of solving the mystery, the film stays in the space between both versions.
The film is structured through conflicting memories.
The present
The couple returns to the apartment and discovers the fallen plant. This is the shared moment where the conflict begins.
Lily’s perspective
Lily remembers Leo coming home at night. In her version, he is tired, distracted, and careless. His bag accidentally hits the plant and knocks it down.
Leo’s perspective
Leo remembers Lily taking care of the plant too intensely. She waters it, adjusts it, and handles it repeatedly. In his version, care becomes control, and the plant falls because of over-care.
The plant
The plant works almost like a silent third character. It does not explain what happened, but it holds the damage. It becomes the object everyone looks at, blames around, and eventually returns to.
The project is not about proving who is right. It is about how memory becomes unstable when guilt and blame are involved.
The story was organised into four main parts:
Reveal / present moment
The couple enters the apartment and sees the fallen plant.
Lily’s version
Leo comes home at night and accidentally knocks the plant down.
Leo’s version
Lily over-cares for the plant until her attention becomes harmful.
Fix / ending
The couple returns to the plant together. The truth remains unclear, but the act of care becomes more important than being right.
The Space
The apartment was designed as a small, lived-in room with a consistent layout: window, couch, bookshelf, side table, lamp, coffee table, rug, and plant corner.
The room had to work in two different moods:
Morning: warm, soft, domestic, clear
Night: darker, blue, uncertain, memory-like
Maintaining the same space across different AI generations became one of the main technical challenges.
The Characters
Leo
Leo was designed as relaxed, sleepy, and slightly careless. His morning outfit is a white T-shirt and plaid pajama pants. His night memory version includes darker clothing and a bag, making the accidental plant fall believable.
Lily
Lily was designed as anxious, caring, and emotionally attached to the plant. Her grey hoodie and dark sweatpants make her feel domestic, soft, and familiar. Her night version shows a quieter, more vulnerable side.
The Plant
The plant was treated as the third main character. It was developed in multiple states:
healthy plant, plant on the table, fallen plant, damaged plant, repaired plant
The fallen plant had to remain visually consistent: terracotta pot on its side, soil spilled, broad monstera leaves spread across the floor.
Grok was used to generate and animate the film shot by shot. The process involved many tests because the AI often changed faces, shifted furniture, altered scale, or moved the plant between shots.
These problems helped shape the workflow: wide shots were used for body movement, close-ups for emotion, and each prompt became shorter, clearer, and more controlled.
The generated shots were edited in DaVinci Resolve.
The editing helped connect separate AI clips into one continuous story. It also helped control the rhythm between:
present moment, memory transition, flashback, argument, return to present
Sound and timing were important because the film depends on subtle emotional shifts rather than big action.
The project changed from a broader idea about conflicting perspectives into a simple domestic story about a fallen plant. This made the film clearer and easier to produce, while keeping the main question about memory and blame.
The strongest change in meaning came from the two memories. In Lily’s version, the plant falls because Leo is careless. In Leo’s version, the plant falls because Lily cares too much. The same object becomes evidence for two different emotional truths.
What worked well was the simplicity of the story. One room, two characters, and one plant were enough to create conflict.
The most difficult part was consistency. AI often changed faces, room layout, scale, or object positions. The project became a lesson in directing AI carefully: using references, controlling each shot, and keeping each prompt focused on one action.
The main discovery was that AI animation is not automatic filmmaking. It still needs planning, blocking, editing, and strong visual continuity. The clearer the references and the smaller the action, the better the result.
Some AI tests produced unexpected results: a third character appeared, the pot moved, the colour grading changed, or the plant stood at an unnatural angle. These mistakes became part of the process and helped refine the final workflow.
The biggest challenge was keeping the AI consistent. Faces changed, furniture shifted, colours varied, and the plant sometimes moved or appeared differently between shots.
Through this, I learned that AI animation needs very clear direction. Simple actions, strong reference images, and controlled prompts worked best. The process became less about generating everything at once and more about directing each shot carefully, then shaping the story through editing.
Light and Perspective
This exercise explored how light and camera angle can completely change the mood of a scene.
At first, a girl opens the door and warm light fills the room. The space feels safe, calm, and welcoming. Then the mood suddenly shifts with thunder. The camera moves to a top-angle view, making her look smaller and more unstable.
The bluish fridge light changes the feeling again. What first looked like a normal room starts to feel strange, cold, and tense.
The exercise showed that the story does not only come from action or dialogue. Light, perspective, colour, and camera movement can change how we understand the same space and the same character.
This was an earlier idea using the Rashomon effect in an exam setting. Two students are trapped in a time loop, and each believes the other caused it by cheating.
One remembers a printed cheat sheet falling on the floor.
The other remembers answers written on an arm.
Each loop gives a different version, but the truth is never fully confirmed.
The idea explored how pressure, guilt, and fear of failure can distort memory. Under stress, blame becomes a way to survive uncertainty