AI Trainer

Main mission

Shapes the behavior, tone, and responses of models.

5 key responsibilities

• Define the tone, values, and expected behaviors of the model.
• Write and evaluate examples that shape responses.
• Identify undesirable behaviors and design corrections.
• Collaborate with RLHF teams on preference data.
• Document behavior policies and their arbitrations.

Key skills

Writing, editorial judgment, qualitative evaluation, understanding of LLM, applied ethics.

What's expected

A model with consistent and defensible behavior: a judgment role, accessible to profiles in humanities, law, and social sciences.

Career paths

Lead Model Behavior, AI Ethics Officer, AI Policy.