Knowledge, Context and Agentic Engineering for Knowledge Work
05

Module 05 · Critical Perspectives and Governance

Critical Perspectives and Governance examines evidence, authority, rights, infrastructure and power across AI-supported knowledge work.

Learning question. Which claims, decisions and delegated actions require technical verification, domain validation, responsibility and governance?

Scope

Model output is a candidate representation whose adequacy must be established for a defined purpose. Technical evaluation, source verification, domain validation and acceptance answer different questions. Governance assigns each question to a responsible role and preserves provenance, rights, disclosure and intervention points.

  • Separate formal evaluation, source verification, domain validation and authorised acceptance.
  • Trace how rights, privacy, infrastructure and institutional authority constrain an AI-supported workflow.
  • Analyse automation claims through asymmetric amplification, political economy and cybernetic control.

Concepts

  1. Candidate representationis the model-generated proposal under examination.
  2. Evaluation and verificationtest formal properties, observable behaviour and correspondence with sources.
  3. Domain validationassesses scholarly meaning, relevance and acceptable uncertainty.
  4. Authority and acceptanceassign consequential decisions to an accountable role.
  5. Rights and provenancegovern source use, privacy, attribution, disclosure and redistribution.
  6. Infrastructure and powershape access, dependency, labour, environmental cost and the capacity to intervene.

Starter sequence

  1. 1. Begin with a candidate representation

    A generated transcription, classification or interpretation remains a proposal. Its status follows from recorded sources, transformations and checks rather than fluent presentation.

  2. 2. Distinguish evidence stages

    Evaluation measures specified behaviour. Verification compares claims with sources and requirements. Domain validation applies specialist judgement. Acceptance records that an authorised role permits the result to progress.

  3. 3. Locate critical expertise

    Critical expertise identifies where automated assistance reaches a scientific, legal or institutional decision boundary. Those boundaries determine intervention, escalation and sign-off.

  4. 4. Govern provenance, rights and disclosure

    A defensible workflow records source origin, transformations, licence scope, personal-data constraints and the role of AI assistance. Publication follows the rights attached to every component.

  5. 5. Examine infrastructure and deployment

    Openness has separate layers: weights, code, data, documentation and deployability. Hosting and tool dependencies affect access, auditability and institutional control.

  6. 6. Test automation narratives

    AI often amplifies capability unevenly across actors and tasks. Political economy and cybernetics make ownership, feedback, optimisation targets and control relations visible.

  7. 7. Close with canonical write-back

    Accepted findings, limitations and decisions enter the maintained knowledge base. This creates a recoverable record for later work and revision.

Learning routes

  • In preparation · web-native deep dive

    Evaluation and Governance

    A structured route through evaluation, verification, domain validation, authority, acceptance and disclosure.

  • In preparation · domain route

    Research Data and Knowledge Work

    Applies provenance, rights, privacy and critical expertise to source-bound research workflows.

  • Available · external method

    Research Mission Control

    Defines authority, evidence stages and release boundaries for coordinated research and implementation.

  • Available · workshop case

    KUG Web Module

    Shows source preparation, transcription, structured extraction, verification and use in a small research application.

Teaching material

The Full Slide Deck provides the reference structure. The Full Lecture Notes expand the distinctions among evidence, authority, rights and governance.

Videos

The maintained playlist contains seven Module 05 videos in the order accepted by the Research Director. This starter list presents the first five; the complete library preserves all seven.

  1. Matrix 2025: Gefangen in der Kalifornischen Ideologie – scobel
  2. Douglas Rushkoff – das Mindset der Tech-Milliardäre | Sternstunde Philosophie | SRF Kultur
  3. Die geheime Steuerung, die niemand wählte – scobel im Gespräch: Mit Anna-Verena Nosthoff
  4. Kybernetik kurz erklärt | 5 Fragen an Anna-Verena Nosthoff
  5. What the hell happened with AGI timelines in 2026?

Open all seven Module 05 videos in the Video Library

Workshops

Six registered workshops include Module 05. These examples connect governance with research data, historical sources and situated professional practice.