AI Can Coach a Movement; It Cannot Embody a Ministry

A learner stands before a camera and practises a movement. An artificial-intelligence system compares the learner's posture with an expert reference, identifies a joint that is out of alignment and supplies immediate corrective feedback. The learner adjusts, repeats and improves. Here AI is doing something educationally serious: it is shortening the distance between practice and feedback.

Now place the same logic inside teacher education, nursing, public administration or theological formation. Can a system infer from a visible performance that the learner will exercise judgement under pressure, keep confidence, recognise a vulnerable person, receive correction, act justly or serve a community faithfully? Can a generated score become a witness to character or calling?

The answer requires more precision than either technological enthusiasm or blanket rejection. Artificial intelligence can coach bounded aspects of embodied performance. It cannot embody a vocation, bear responsibility for a learner or certify the whole competence that another person or community must later trust.

A peer-reviewed study published on 29 August 2026 tested an AI-supported Tai Chi tutor in a compulsory university physical-education course. One hundred and twenty undergraduates were randomly allocated either to the AI-supported condition or to video-based instruction using the same curriculum and practice dose.

The system used human-pose estimation to provide joint-specific corrective cues and practice summaries. Expert instructors rated a standardised routine at pre-test, post-test and a two-week…