AI-Shaping Intelligence in Practice: Evidence from Project-Managing Shaped Intelligence
This paper presents a public-safe capability-evidence instance for AI shaping through project-managing shaped intelligence in project and business-as-usual (BAU) IT support contexts. It tests whether visible before/after transitions show work-burden shift: reconstruction, sequencing, drafting, tracking, dependency and risk framing, domain-practice application, subject-context reasoning, project-state preservation, resumption and communication-preparation work moving from the responsible person to shaped intelligence while judgement, approval, escalation, communication and real-world action remain human-owned. The evidence package separates supporting domain-extensibility evidence, primary capability evidence and an illustrative domain-transfer frame. Section 10 is the controlling public evidence test; Appendix A provides lookup detail where a trace point needs checking, and Appendix B remains illustrative only. The permitted public conclusion is limited to initial capability evaluation and whether protected-stage discussion is justified. The paper is not implementation guidance, deployment assurance, reproducibility proof, product-performance assurance, method transfer or a document-authoring capability claim.
Authors
- Kurni Kwok
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-06-24
- DOI
- https://doi.org/10.5281/zenodo.20830132
- Primary Topic
- Ethics and Social Impacts of AI
- Type
- preprint