ChatGPT produced executed, grader-accepted submissions for all three fixed personalized Qiskit assignment instances in 150 of 150 sessions¶
Claim · All claims
Evidence · 1 study · designr2·q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q2 i? For each of three fixed assignment instances, 50 separate ChatGPT sessions each yielded an executed, grader-accepted final artifact, so observed ChatGPT-resiliency was zero under the study's operational definition. → Alexei Kaltchenko and Gurnivaj Tiwana 2026
Evidence¶
Alexei Kaltchenko and Gurnivaj Tiwana 2026¶
Alexei Kaltchenko and Gurnivaj Tiwana. (2026). ChatGPT Solves All Tested Qiskit Homework Assignments. arXiv. https://arxiv.org/abs/2608.19707 doi:10.48550/arxiv.2608.19707
q2 · i? · design · r2
Repeated fixed-instance trial series: for each of three AI-deterrence-modified Qiskit packages, the same student-visible instance was tested in 50 separate ChatGPT sessions under a minimally engaged student workflow, and "all 150 passed." The study reports Ca = Na = 50 and Ra = 0 per package.
"All 150 final artifacts were executed and evaluated with the corresponding grader, and all 150 passed."
Discussion¶
Related Claims¶
- In archived sessions, no operator code changes or quantum-logic corrections were needed; eight of nine notebook executions only required one permitted fix of a missing visualization dependency — related
- Deterrence layers failed because seeds changed parameters rather than task structure, scaffolding exposed solution steps, and hidden grading verified output consistency rather than authorship or understanding — a broader claim this one bears on
- Optional IBM Quantum hardware execution supplied extra evidence but did not remove substantial AI assistance, as ChatGPT generated the code and could interpret returned noisy data — related
- In the pilot workshop, the LLM unreliably produced fill-in-the-blank snippets, often returning complete code solutions instead — related
- Assessments requiring novelty and originality were more resilient to AI-generated output — related
- Visual media assessments (posters, presentations) were not passable because ChatGPT could not produce complete slides or poster content — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
chatgpt-completes-all-150-qiskit-homework-sessions |
title |
ChatGPT produced executed, grader-accepted submissions for all three fixed personalized Qiskit assignment instances in 150 of 150 sessions |
description |
ChatGPT produced executed, grader-accepted submissions for all three fixed personalized Qiskit assignment instances in 150 of 150 sessions |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2026 |
recheck_by |
2029 |
generated |
by process:wiki-ingest on 2026-10-10 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — alexei-kaltchenko-and-gurnivaj-tiwana-2026 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.