This page explores the design and development of software systems that can detect, predict, or emulate narcissistic traits.
It focuses on open-source tools, psychological datasets (such as from OpenPsychometrics.org), and agent-based models enhanced with generative AI. These systems can serve both as research tools and simulations of narcissistic behavior for educational, experimental, or policy use.
# Goals - Predict narcissistic traits using survey/interview data - Emulate narcissistic behavior in conversational agents or simulations - Explore emergent group behaviors in systems populated by narcissistic agents - Encourage ethical reflection on the creation and use of these systems
# Data Sources - **OpenPsychometrics.org**: Provides public access to raw test results from validated psychological instruments such as the Narcissistic Personality Inventory (NPI). - **Custom Surveys**: Design and distribute your own questionnaires using validated scales (e.g. NPI, NAR-Q, Five-Factor Narcissism Inventory). - **Structured Interviews**: Use semi-structured interview protocols with questions designed to elicit self-focused, grandiose, or empathy-related narratives.
# Open Source Tools and Models - **scikit-learn / PyCaret**: Use for baseline prediction models (e.g. logistic regression, SVM) on personality datasets. - **Transformers (Hugging Face)**: Fine-tune generative language models to emulate narcissistic communication styles or generate synthetic narcissistic dialogue for training/testing. - **LangChain / LlamaIndex**: Integrate language models into agent-based workflows and give them access to long-term memory and reasoning tools. - **Mesa (Python)**: Build agent-based models to simulate social systems with varying proportions of narcissistic agents. - **Rasa**: Create chatbots with modular dialogue policies that can simulate or react to narcissistic traits. - **GPT-based Classifiers**: Use prompt-based or fine-tuned generative models to classify text or responses for narcissistic indicators.
# Agent-Based Modeling Concepts - **Agent Personality Profiles**: Each agent can be assigned personality traits based on NPI-style scores or Big Five dimensions. - **Social Dynamics**: Agents interact through networks or environments that allow for status-seeking, manipulation, flattery, and conflict. - **Reinforcement Loops**: Agents with narcissistic traits may learn to reward behaviors that preserve self-image or elevate social rank. - **Emergent Behaviors**: Study how group outcomes change when narcissistic agents are added to collaborative, competitive, or cooperative systems.
# Use Cases - **Mental Health Research**: Simulate clinical scenarios or public health trends. - **Sociological Prototypes**: Model environments like startups, cults of personality, or political echo chambers. - **Conversational Testing**: Train agents that can roleplay narcissistic personalities for education or therapy simulation. - **Art & Satire**: Build fictional ecosystems or interactive art pieces exploring ego, identity, and self-promotion.
# Ethical Considerations - Ensure transparency in model goals and limitations. - Avoid reinforcing harmful stereotypes or using simulations for manipulation. - Provide disclaimers when using generated personas that mimic clinical traits. - Consider consent, anonymization, and data security in survey/interview data collection.
# References
- Raskin & Hall (1979). "The Narcissistic Personality Inventory"
- Back et al. (2013). "Narcissistic Admiration and Rivalry Concept (NARQ)"
- Grijalva et al. (2015). "Narcissism and leadership: A meta-analytic review"
- OpenPsychometrics.org ![]()
# See Also - Agent-Based Models and Personality Simulation - Ethical AI and Psychometrics - Open Psychology Engines - Guide to Dysfunctional Societies