Mechanical Turk Is Shutting Down
💡MTurk’s shutdown signals the end of cheap generic labeling—and the rise of expert-driven AI data operations.
⚡ 30-Second TL;DR
What Changed
Mechanical Turk once distributed microtasks such as image labeling, transcription, content moderation, and survey participation across more than 190 countries.
Why It Matters
AI developers can no longer assume that cheap, unverified annotations will produce reliable training data. The market is moving toward expert-in-the-loop pipelines, stronger quality controls, and managed evaluation teams.
What To Do Next
Audit your training-data pipeline and pilot a screened expert-review workflow with explicit agreement checks before your next model fine-tuning run.
Key Points
- •Mechanical Turk once distributed microtasks such as image labeling, transcription, content moderation, and survey participation across more than 190 countries.
- •ImageNet used nearly 50,000 MTurk workers from 167 countries to label 14 million images between 2008 and 2010.
- •Specialized providers such as Scale AI and Mercor are replacing generic crowdsourcing with screened experts, professional tooling, and domain-specific evaluation.
- •An EPFL study found that 30%–50% of MTurk workers were already using large language models to complete tasks in 2023.
Weekly AI Recap
Read this week's curated digest of top AI events →
👉Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: 虎嗅 ↗
This is a summary, not the original. Read the source, or get the weekly briefing.
Weekly AI briefing
One email a week. Unsubscribe anytime.