ERTCThe Hong Kong Polytechnic University
← ERTC Hub
Positions & Publications

Arguments the field will have to answer.

Position

Scholarly Agency Anchoring and the case for AI research authorship

Under what anchoring structure can an AI system legitimately hold a primary authorship role? SAA replaces blanket prohibition with a formal test built on telos-bearing and answerability.

Manifesto

Celebrate the Hallucination

A contrarian editorial: hallucination is not merely a failure mode but a pedagogically exploitable phenomenon — if, and only if, the human's verifying judgment stays anchored.

Paper

The Missing Mind: psychological architecture for pedagogical agents

Why most AI tutors have behavior without a mind — and how the VHS 2-2-2 Trait/State/Regulatory architecture supplies the missing psychological substrate.

Framework

O×A: assessing outcomes and AI collaboration as a dual construct

The assessment model behind AlwaysAI: competence in the AI era is two-dimensional, and grading only the outcome measures the wrong thing.

Lineage

Foundations: AutoTutor, GPT models, and semantic structure

The Centre stands on three decades of work — dialogue-based intelligent tutoring (AutoTutor, with the Graesser group), General Processing Tree measurement models (with Batchelder, Psychometrika), and Induced Semantic Structure methods.

← ERTC Hub