KERA's Think The man (or woman) behind the curtain of A.I.
Nov 17, 2025
Varsha Bansal, a tech reporter for The Guardian focused on AI and labor, reveals the surprising reality behind AI systems. She discusses the critical, yet often unseen, role of human raters who train these systems under intense pressure and low pay. Listeners will learn about the emotional toll of moderating content, the complexities of AI guidelines, and the ethical dilemmas faced by those in the field. Varsha also highlights the urgent need for better working conditions and mental health support for these essential contributors.
AI Snips
Chapters
Transcript
Episode notes
Unexpected Trust-and-Safety Role
- Rachel Sawyer accepted a remote writing analyst role expecting technical writing but found herself rating and moderating AI-generated content.
- She was later assigned exclusively to distressing trust-and-safety content without prior warning or training.
Don't Rush—Flag Technical Uncertainty
- Expect rating tasks to be time-limited and avoid superficial judgments under pressure.
- Take notes when unsure and use team consensus mechanisms to flag ambiguous or technical items.
Outsourcing Keeps Liability At Arm's Length
- Big tech uses contractors and subcontractors to outsource AI rating, keeping liability and costs at arm's length.
- This creates fragmented workforces who perform crucial tasks without stable employment or benefits.