Personality & Mental Health Specialist
AI Safety • Full-time • Berkeley, CA / Remote
Psychology
AI Safety
Research
Overview
We are seeking a Personality & Mental Health Specialist to ensure our AI systems interact with users in psychologically safe and beneficial ways. You will help design AI personalities that promote positive mental health outcomes and prevent potential harms.
Key Responsibilities
- Design and evaluate AI personality frameworks that promote positive user interactions
- Develop guidelines for AI responses to sensitive topics and vulnerable users
- Collaborate with engineering teams to implement safeguards for mental health considerations
- Conduct research on the psychological effects of human-AI interaction
- Create evaluation metrics for measuring the psychological impact of AI systems
- Develop protocols for identifying and responding to users in crisis
- Stay current with research in digital psychology and mental health
Requirements
- Advanced degree in Psychology, Cognitive Science, Human-Computer Interaction, or related field
- Understanding of personality psychology and mental health principles
- Experience applying psychological principles to technology design
- Knowledge of ethical considerations in AI-human interaction
- Strong research skills and familiarity with qualitative and quantitative methods
- Excellent communication and collaboration abilities
- Empathy and sensitivity to diverse user needs and experiences
Preferred Qualifications
- Clinical psychology background or experience in mental health services
- Experience designing or evaluating conversational AI systems
- Knowledge of trauma-informed design principles
- Familiarity with AI alignment techniques and challenges
- Publications or research in digital wellbeing or AI psychology
- Experience with user research and testing methodologies
How to Apply
If you're excited about this opportunity and believe you're a good fit, we'd love to hear from you. Please submit your application through our online portal or contact us directly.