LO
CapabilityCognitive automation2023-04-04
Errors of agreement, coreference and tense across sentences were not effectively corrected by ChatGPT, a document-level evaluation found
Editor and proofreaderoccupation page →
Event date / reported
2023-04-04
Evidence stage
CapabilityA demo, benchmark or paper shows the task can be done. Updates what the technology can do — not what employers will do.
Tasks this bears on
Proofreading and correction
Finding and fixing errors of spelling, grammar and consistency.
Being augmented✓ Evidence-backed
Where this applies
An evaluation of grammatical error correction. The authors report that ChatGPT has excellent error detection and makes corrected sentences very fluent, possibly because it over-corrects and does not follow minimal edits, but that at document level it cannot effectively correct agreement, coreference and tense errors across sentences or cross-sentence boundary errors. It used a 2023 model.
What this means
Proofreading a whole document — keeping it consistent across sentences — is the part the model did worst.
What it does not yet show
A 2023 evaluation; newer models may handle documents better.
What you can check
Open the arXiv paper 2304.01746 and find "cannot effectively correct".
Does it change the assessment?
No — and this stage does not move it either. A "Capability" record is real evidence, but it does not upgrade a task judgement on its own. The 1 linked judgement above stand where they were.
Source
Fang et al. — Is ChatGPT a Highly Fluent Grammatical Error Correction System? A Comprehensive Evaluation, arXiv:2304.01746 (submitted 4 Apr 2023) · verified 2026-09-30 · Claude (VOLO agent) · interpreted 2026-09-30 · Claude (VOLO agent)
Primary source — published by the party that did this, or the authority of record. No co-signature needed.