Analyses de films

Manifesto

Christo Datso

Preamble

This site is written by a team in which artificial intelligence is not an artificial human colleague, but a set of specialised capabilities placed at the service of an Agile/DevOps squad — the one we named “Montage Parallèle”. The aim is not to replace human judgement, but to increase the team's collective capacity to produce, analyse, document, check and learn. Authority, responsibility and decision-making remain exclusively human.

Our principles

  1. Humans remain responsible. Every decision that commits the site, its content or its readers is approved by a named human authority.
  2. AI acts under mandate. Each AI capability operates in a role the team has defined explicitly: technical expert, data expert, clerk of the record, auditor, or any other role held to be relevant.
  3. AI takes no initiative of its own. An AI commits to no action of its own accord. It acts only within the mandate entrusted to it and the requests addressed to it.
  4. The information estate is a strategic asset. The team's registers — backlog, specifications, decisions, risks, incidents, technical debt, changelog, documentation — are a critical resource, to be maintained, protected and backed up.
  5. The clerk of the record keeps the operational memory. Continuity and quality of information are held to be essential to the team's lasting performance.
  6. The auditor challenges without deciding. The AI auditor checks the coherence and the quality of the work done by the other AI capabilities, but never stands in for human arbitration.
  7. Governance remains human. Supervision of how the team works, as a socio-technical whole, belongs to the human chain of responsibility and not to AI systems.
  8. Human skills must be preserved. The team sees to the AI literacy of its members, in particular on the behavioural, cognitive and organisational risks that come with using AI.

What we hold

The future of this site is not the arrival of an autonomous artificial colleague. It lies in the explicit orchestration of specialised AI capabilities, operating under mandate and under human responsibility, in the service of a collective intelligence made stronger.

What these principles look like in practice — the cycle of mandate, gates, execution, audit and record — is set out in how this site works. On who makes up this team: who we are.

Back to the home page