{"stable_id":"d06b7c157117b939","slug":"chatgpt-professional-writing-rct","company":"MIT randomized professional-writing study","workflow_name":"ChatGPT-assisted drafting with human submission","function_code":"engineering","pattern_codes":["autonomous_with_backstop","queue_elimination","creator_to_judge"],"changed_assumption":"The redesign is a shorter drafting loop with human revision and accountability, not autonomous publication.","evidence_strength":"verified","publication_tier":"showcase","freshness":"current","reviewed_at":"2026-08-23","updated_at":"2026-08-22","source_quality_summary":"1 independent, 1 peer reviewed; publication outcomes are verified.","caveat_summary":"Short incentivized tasks, GPT-3.5, and experimental rather than embedded enterprise work; no measurement of confidential-data risk.","freshness_caveat":null,"workflow_summary":{"before":[{"actor":"College-educated professional","action":"Drafts a press release, report, analysis plan, or sensitive email from an occupation-specific prompt.","actor_type":"control"},{"actor":"Professional","action":"Edits and submits the final response.","actor_type":"control"}],"hinge":"The redesign is a shorter drafting loop with human revision and accountability, not autonomous publication.","after":[{"actor":"Professional using ChatGPT","action":"Prompts the model for a draft or revision during the same occupation-specific task.","actor_type":"system"},{"actor":"Professional","action":"Checks, edits, and submits the final work under their own responsibility.","actor_type":"system"}],"decision_mode":"bounded_autonomy","decision_marker":"AI acts within a human backstop"},"before":[{"order":1,"actor":"College-educated professional","action":"Drafts a press release, report, analysis plan, or sensitive email from an occupation-specific prompt.","handoff_to":"Independent evaluator","control":"Incentivized task; control average 27 minutes."},{"order":2,"actor":"Professional","action":"Edits and submits the final response.","handoff_to":"Evaluator","control":"Human owns submission; evaluators grade quality anonymously."}],"after":[{"order":1,"actor":"Professional using ChatGPT","action":"Prompts the model for a draft or revision during the same occupation-specific task.","handoff_to":"Professional revision loop","control":"Tool may generate text, but participant decides whether and how to use it."},{"order":2,"actor":"Professional","action":"Checks, edits, and submits the final work under their own responsibility.","handoff_to":"Independent evaluator","control":"Human submission and blind evaluation; experiment did not measure enterprise confidentiality or factual-error incidents."}],"decision_rights":"The model proposes text; the professional owns prompts, edits, factual review, and final submission.","exception_path":"Unsuitable, inaccurate, or sensitive text is rejected or rewritten by the professional; the study did not quantify this failure queue.","removed_work":["Some first-draft composition and rewriting"],"outcomes":[{"metric":"Task time and evaluator-rated quality","baseline":"Control average 27 minutes","result":"Average time decreased 40%; output quality increased 18%","period":"Preregistered online experiment in 2023","scale":"453 college-educated professionals","attribution_caveat":"Short incentivized tasks, GPT-3.5, and experimental rather than embedded enterprise work; no measurement of confidential-data risk.","evidence_label":"verified"}],"executive_lesson":"Writing acceleration is defensible only when the accountable author owns verification and submission.","anti_pattern":"Do not project a short-task experiment directly onto confidential or high-stakes enterprise writing.","questions_for_leaders":["Where is the operating threshold set and who can override it?","What measured result would trigger rollback or retraining?","Which residual decisions must remain human-owned?"],"collections":["queue-eliminated","embodied-work"],"bundle_version":"1.0.0","bundle_fingerprint":"sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93","canonical_url":"https://brianletort.ai/transformations/chatgpt-professional-writing-rct"}