When a student pauses for thirty seconds before a math problem, the system records the pause and moves on. It cannot tell confusion from careful thinking, anxiety from strategic deliberation.
AI tracks what. It misses why.
Chatbots handle 85% of customer interactions at leading companies. Recommendation engines shape what 2 billion people see daily. Adaptive learning platforms guide millions of students. The systems that understand users deeply will outperform those that only track them.
Clicks and timestamps capture automatically. Emotions and motivation do not show up in server logs — they must be inferred from patterns most AI development skips.
Psychologists study motivation but lack computational tools. Computer scientists build pattern recognition but lack theoretical grounding. Cross-citation is minimal.
When engagement metrics become KPIs, systems optimize for time-on-platform rather than user goal achievement. Personalisation stays on the surface.
Each layer is generated by the one beneath it from the user's perspective. AI must work in reverse — inferring upward from observable Actions.
A 30-second pause before a complex financial trade is prudent deliberation. The same pause before a simple checkout is probably confusion.
Signal-to-state mapping is probabilistic and context-weighted — never deterministic.
Industry, task type, interaction norms. Hesitation in healthcare ≠ hesitation in checkout.
Individual baseline, history, expertise. A consistently methodical user isn't hesitating.
Time of day, journey stage, recent UI changes. Spike after redesign signals disorientation, not indecision.
This is a conceptual framework study — validation focuses on conceptual coherence, theoretical grounding, and practitioner utility.
Integrative review across SDT, affective computing, JITAI, learning analytics. Substantially complete.
Semi-structured, 15–20 min. CX primary, EdTech secondary. Thematic analysis (Braun & Clarke).
Descriptive breadth on signal recognition, state inference, framework utility.
One CX, one EdTech. Illustrative applications using public documentation.
Activities run in parallel. Live progress is at predictwhy.com.
Finalise framework, interview protocol, survey. Pilot-test. Outreach to 10–12 interviewees.
5 interviews, one per day. Survey deployed across LinkedIn, mentor network, EMBA alumni.
Two AI systems analysed through the ABSD lens using public documentation.
Thematic analysis. Cross-method refinement of the framework.
Mentor review cycles. Final revisions. Submission 14 June 2026.
A synthesised conceptual framework connecting four siloed research traditions. Shared vocabulary across layers of Action, Behaviour, State, Drive. Platform for future empirical testing.
Design heuristics for state-aware AI systems. A diagnostic and implementation guide for organisations moving beyond event counting toward context-aware user intelligence.
Anchored empirically in CX and EdTech. Transferable to finance, healthcare, and any domain where AI mediates human decision-making.
Simplicity. Universality. Specificity. Groundedness. Actionability. Validated through practitioner feedback.