Summary
MIT IDE writeup based on a presentation by Siddharth Suri (co-author with Mary Gray of the book Ghost Work, 2019). Suri documents how on-demand workers globally perform the essential micro-tasks — web search verification, content moderation, image recognition, spam detection — that power modern AI. Median wage: $2/hour. McKinsey estimated on-demand work platforms could add $2.7T to global GDP by 2025, yet workers remain invisible to policy and public discourse.
Key Points
- Amazon Mechanical Turk and similar platforms employ millions of on-demand workers worldwide
- Median wage: $2/hour; workers motivated by flexibility and skill-building despite low pay
- Tasks performed: content moderation, data labeling, search quality rating, transcription, image tagging
- Suri’s core thesis: this work is not transitional — it’s structural. AI systems will always require human judgment for edge cases and quality control; ghost work won’t disappear when AI improves
- McKinsey projection: on-demand work platforms → $2.7T GDP contribution by 2025; yet zero policy attention to worker conditions
- Gray & Suri coined “ghost work” to name the invisible labor that makes AI function
Newsletter Angles
- Suri’s structural argument: The strongest version of the ghost work thesis isn’t “AI isn’t ready yet so humans fill the gaps.” It’s “AI will always require human judgment at the margins, and this labor is being structurally concealed and undercompensated.” If true, ghost work is a permanent feature of the AI economy, not a temporary condition.
- The $2.7T accountability gap: McKinsey tracks the GDP contribution of on-demand platforms while the workers generating that value earn $2/hour median. That gap is a policy story.
Entities Mentioned
- Amazon — Mechanical Turk as the dominant platform
- Mechanical Turk Pattern — Suri & Gray’s book is the foundational research document for this concept
Concepts Mentioned
- Mechanical Turk Pattern — “ghost work” is their term; this book is the academic foundation for the concept
- Leverage Erasure Through Automation — workers without leverage because their labor is abstracted, interchangeable, and invisible
Quotes
“Ghost work” — work delivered through APIs and platforms that makes humans look like function calls rather than workers
Notes
Primary research: Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass by Mary L. Gray and Siddharth Suri (2019) is the canonical academic treatment of this labor pattern. This source is a derivative presentation summary. The book is the citation to use in long-form newsletter pieces.