WorkSeptember 2, 2026via 404 Media
Podcast: We Spoke to an Amazon Worker Destroying Books for AI
Why it matters
AI training depends on massive data acquisition — often at odds with intellectual property, labor practices, and worker welfare. This first-person account from inside Amazon's operations surfaces the real-world workplace and ethical costs of building frontier models at scale.
Key signals
- Amazon worker testimony on book destruction for AI training data
- First-hand account of data acquisition practices for model training
- Labor/workplace angle: worker impact of AI infrastructure buildout
- Related: AI-generated paper authorship contamination (ghost names in academic outputs)
- Related: ICE AI spending (government AI infrastructure investment)
- Amazon worker testimony on book destruction for AI training
- Data sourcing practices at scale (Amazon operations)
- Labor/workplace angle on AI infrastructure
- Follow-up to earlier reporting on book destruction
- Broader pattern of AI-generated content pollution (fake names in papers)
The hook
Amazon worker reveals the human cost of AI training data: books destroyed, jobs displaced, corners cut.
A follow up to the Amazon destroying books for AI story, why a bunch of names keep appearing in AI-generated papers, and ICE's latest spending spree.