R0007/2026-03-20/C011 — Assessment

Research	R0007 — AI Cognitive Amplifier
Run	2026-03-20
Claim	C011

BLUF

Confirmed. Noy and Zhang (2023) found that ChatGPT compressed the productivity distribution by benefiting low-ability workers more. Average time decreased 40%, quality rose 18%. Inequality between workers decreased. Published in Science (July 2023).

Probability

Rating: Almost certain (95-99%) Confidence in assessment: High Confidence rationale: Based on web-accessible evidence.

Reasoning Chain

1. Noy and Zhang (2023) assigned writing tasks to 444 college-educated professionals. ChatGPT reduced time by 40% and increased quality by 18%. Inequality between workers decreased as ChatGPT compressed the productivity distribution by benefiting low-ability workers more. Published in Science. [SRC01-E01, High reliability, High relevance]

Evidence Base Summary

Source	Description	Reliability	Relevance	Key Finding
SRC01	Experimental Evidence on the Productivity Effects of Generative AI	High	High	See BLUF
## Collection Synthesis
Dimension	Assessment
-----------	------------
Evidence quality	Medium to Robust
Source agreement	High
Source independence	Assessed per claim
Outliers	None
### Detail
Noy and Zhang (2023) assigned writing tasks to 444 college-educated professionals. ChatGPT reduced time by 40% and increased quality by 18%. Inequality between workers decreased as ChatGPT compressed the productivity distribution by benefiting low-ability workers more. Published in Science.
## Gaps
Missing Evidence	Impact on Assessment
-----------------	---------------------
Full-text access	Low to Moderate
## Researcher Bias Check
Declared biases: None.
Influence assessment: Standard verification.
## Cross-References
Entity	ID	File
--------	----	------
Hypotheses	H1, H2, H3	hypotheses/
Sources	SRC01	sources/
ACH Matrix	—	ach-matrix.md
Self-Audit	—	self-audit.md