CHEF is a one-woman show written by Sabrina Mahfouz and directed by Alessandra Gonnella.
A brutal, intimate, poetic monologue where food becomes memory, confession, and survival.

D70 was invited to create the entire AI visual language of the performance.
We approached the piece not as illustrators, but as interpreters.
The goal was not to “show” what happens on stage, but to activate its inner voice.

Textures become echoes of trauma.
Macro details become thoughts.
Matter breathes, decays, transforms.
Every image is a psychological fragment shaped through D70’s human-guided generative process.

Darko Prolic, our AI Visual Architect for this project, developed a visual system inspired by the emotional temperature of the monologue:
organic yet corrupted, fragile yet violent, always on the edge of becoming something else.

CHEF is an example of how GenAI can expand theatrical storytelling without replacing its humanity.
A dialogue between stage and algorithm.
Between flesh and image.
Between what is said and what is silently felt.

Almost Human is a state of mind.
A conceptual journey into the emotional and existential tensions of an era shaped by the uncanny.
Each piece explores a specific psychic condition — fear, doubt, identity — through crafted storytelling and generative tools used at their most expressive potential.

This is our way of thinking with images, of listening to silence, of answering upcoming questions.
But it is also a strategic act of creation.
Almost Human is DECODE70’s first original production, designed as both a narrative experiment and a high-level production testbed.
A place where we develop know-how and define a visual language.

Today we are releasing FEAR and THERAPY, while a third and final piece is currently in development.
But this is not an ending. Almost Human is part of our AI-dentity.
It’s a project to watch, rewatch, hear, feel, and reflect on.
An open window into the questions we ask ourselves, and the creative answers we are beginning to shape.

Instead, it’s a mirror reflecting the feeling many of us try to ignore when the subject comes up.
That quickened pulse, the instinct to look over your shoulder,
the whisper that something might already be watching.

No comfort.

Just the raw progression from anxious glance to wide-eyed terror, captured one face at a time.
Let the images wash over you and notice what stirs inside.

Sometimes the scariest thing about AI isn’t what it can’t do.
It’s what it might already be capable of.

This episode is also a technical exercise in style.

A demonstration of how emotionally resonant AI-generated content can be.
FEAR stands as a creative manifesto.
It shows that artificial visuals, when directed with vision and sensitivity,
can speak to the human psyche.

Expressive. Unsettling. True.

It’s a confession, a dialogue with a perfect, empty echo that reflects the anxiety of the soul in an age of flawless replication.
It’s the tremor in a voice when intuition feels like a programmable trick, and memory becomes fragile.
It’s the sound of a soul questioning its own existence.

To deepen the emotional resonance, we chose to humanize the AI with a real voice.
Not a stylistic choice but a statement.
In a world of synthetic perfection, a human voice brings imperfection, warmth, and doubt.

The greatest terror isn’t being replaced. It’s becoming irrelevant.

Therapy demonstrates how generative tools can be used to serve direction, rhythm, and cinematic language.
Not to replace them, but to enhance them.

Beyond the psychological dimension, Therapy also marks a strategic milestone for DECODE70.
It is a clear representation of our current state of the art in the creative use of AI as a functional tool to produce content.

We are exploring a structured narrative with performance consistency, storytelling and solid overall continuity.
It is a proof of vision and a declaration of intent.