The Australian Financial Review
Details
- Date Published
- 5 Apr 2026
- Priority Score
- 2
- Australian
- Yes
- Created
- 4 Apr 2026, 10:00 pm
Authors (1)
Description
They vent, confess and role-play with strangers, all to help machines learn how to sound human.
Summary
This report examines the human labor market behind Reinforcement Learning from Human Feedback (RLHF), where professionals like lawyers and comedians simulate complex emotional scenarios to refine AI communication. While detailing the gig economy's role in making frontier models sound more empathetic and human-like, it highlights the increasing sophistication of AI social manipulation capabilities. These advancements in human-centric interaction are critical to understanding how future AI systems might successfully bypass human barriers through social engineering or persuasive deception. The article provides context for the labor-intensive processes used by leading AI labs to reduce superficial errors while potentially introducing new nuances in how chatbots role-play sensitive situations.