Why ciniq is built this way
Against convergence.
In 1951, Solomon Asch sat a person in a room of strangers and asked which of three lines matched a fourth. The strangers, planted, answered wrongly in unison — and a third of the time the person denied the evidence of their own eyes to agree with the room. In 2006, the MusicLab experiment gave thousands of people the same unknown songs in parallel artificial worlds; the only difference was a visible download count. Different songs became hits in each world. The count did the choosing. Quality barely constrained it.
Film apps run on that machinery. The score arrives before the judgment: you know a film is a 7.8 before you've watched a frame, and the 7.8 does some of your watching for you. Ratings drift toward what looks defensible in public. The feed rewards saying what the room already says, and people congregate in it to be reassured of what they already believed. The average is the product.
What ciniq refuses
No crowd scores anywhere — that is the point, not an omission. No critic quotes standing in for your verdict. No feed, no counts, no followers, no audience to perform taste for. And no analytics reading your judgments on some server of ours: a verdict formed in private stays private.
Commit, then compare
A card carries your reading of one film to one friend, and they score it blind — your numbers stay hidden until theirs are committed. Then both readings are revealed side by side, and you see exactly where you split. Influence arrives after judgment, as an argument — never before it, as gravity. Two people arguing over a revealed split is deliberation; two people watching a number converge is contamination.
And your friend's numbers never enter your engine. What the exchanges build is something better than a blended score: over time you learn whose taste actually tracks yours, dimension by dimension — this friend is reliable on direction and hopeless on endings. Trust between people, calibrated by evidence, instead of an average of strangers.
Juries, not averages
The canon is thousands of films chosen by festival and award juries — people who watched everything and signed their names to a choice. Juries curate; crowds average. And the engine that draws from it never pretends to predict your verdict: it retrieves films with real edges to what you've loved, tells you its because, and gives each one ten minutes. The audition is the classifier. Your verdict is yours to render — across ten dimensions, negatives allowed, with no number in the room but your own.
The cost, accepted
Built this way, ciniq declines the network effect that convergence powers. There is no feed to grow, no count to climb. It spreads the way an argument does — one card, one friend, one revealed split at a time — not the way a chant does.
That is slower. We'll take the spine.
— Labham LLC · labham.com