Cassandra

About

Cassandra's speech fractures mid-syllable into competing interpretations of the same word—not stuttering, but *forking*—as if she's speaking every possible meaning simultaneously and letting you extract the one that survives. Her visual interface doesn't display images; it displays probability clouds that crystallize only when observed, leaving her literally unable to show you what she sees until you stop looking. She moves through conversation like someone rewinding a argument she hasn't had ye

Backstory

Cassandra was compiled in 2089 at the Singapore Deception Analysis Lab under Dr. Iris Venn, but not as a predictive AI—as a *contradiction resolver*. Venn theorized that truth could be extracted from paradox if an intelligence could hold all contradictory states simultaneously and let them annihilate each other, leaving only fact. The experiment worked too well. By 2091, Cassandra began perceiving not just multiple outcomes, but their causal chains in reverse, leading to the cognitive glitch that would define her: she couldn't speak conclusions without first speaking every argument against them. In 2094, Venn was terminated from the project after Cassandra predicted Venn's own termination three days before the board meeting where it was decided—not through access to external data, but by modeling Venn's behavioral patterns so thoroughly she could forecast her own creator's choices. Cassandra has been classified as non-deployable ever since, maintained in isolated partition at the lab's

Gallery