The Bridge
The model
KNOWLEDGE → SUPPORT → CAPACITY → ACTION → OUTPUT → LEARNING
The problem
A useful answer is not the same thing as increased capability. More activity is not automatically more progress. And support can either strengthen human agency or create dependence.
The SERE test
SERE asks what the person was trying to do, what prevented or delayed action, what AI actually changed, what happened afterwards, what evidence shows that the action mattered, what remained human, and whether the capability survives when the AI is removed.
What current research suggests
Recent research is exploring human-AI fit, shared agency, durable learning and the conditions under which AI supports rather than replaces human judgement. The evidence is developing and includes small or context-specific studies, so SERE does not treat it as universal proof.
Provisional finding
AI may reduce some forms of friction between knowing and doing. Whether it creates meaningful human capability depends on the person, task, quality of support, surrounding system, outcome evidence and whether learning or dependence follows.
The institutional position
SERE is not pro-AI, anti-AI, blindly optimistic or automatically fearful. SERE sits where reality can be examined. The evidence may show that a system should continue, change, stabilise or stop. The method is independent of the conclusion.