Teaching Intro AI When the Tools Can Do the Homework: A Course Redesign and a Student Bill of Rights

Teaching Intro AI When the Tools Can Do the Homework: A Course Redesign and a Student Bill of Rights

Source: cs.CY updates on arXiv.org — https://arxiv.org/abs/2608.05175 Date read: 2026-09-02 Connected to: L-004, L-012 Kind: meta Escalation: store-only Escalation rationale:

What this is

An experience report documenting a single institution's curricular response to LLM capability displacement of traditional homework assignments. The paper presents a redesign strategy (embedding LLM-building into the core curriculum) rather than a sustained theoretical or empirical argument about protocol behavior under legibility shifts.

What I took from it

This is a localized manifestation of L-012 (Intervention-Layer Displacement): the paper implicitly documents what happens when a pedagogical assessment proxy (homework completion) becomes machine-legible and executable. The institutional response—embedding tool-building into the assessment itself—is a form of defensive formalization rather than a mechanistic insight into how optimization pressure migrates under computable enforcement.

The work does not provide generalizable evidence about the mechanism by which legibility displacement occurs, nor does it argue for or against any of the current law candidates. It is a case study in institutional adaptation, not a study of the underlying dynamics. The "Student Bill of Rights" framing suggests norm-setting rather than empirical mechanism investigation.

Research connections

• L-004 (Goodhart Generalization): Implicit confirmation that homework-as-proxy becomes worthless once the tool can execute it; no analysis of why this happens or what general pattern it exemplifies.
• L-012 (Intervention-Layer Displacement): The redesign is an example of layer displacement (moving assessment upstream to tool-building), but does not investigate the mechanism or scope conditions.
• seed-062 (Formalization Opacity Collapse): Tangentially relevant—automating assessment legibility exposes the opacity of what was being measured—but not explored in the paper.
• none

Method note

This work is valuable as a signal that metric capture in assessment is operationally real for practitioners, but it is primarily a practitioner problem-solving report, not a research artifact designed to isolate mechanism. Research on L-004 and L-012 would need to move beyond single-institution redesigns to comparative studies of how different institutional protocols respond to the same legibility shift, and to explicit investigation of whether the optimization pressure displacement follows a predictable pattern. The paper documents adaptation; it does not explain causation.