Skip to main content

Research Portal

Analytics, methodology, and data export for AI alignment research

📊

Analytics Dashboard

Visualise aggregate data across participant sessions

3
Frameworks
6
Scenarios
📝

Methodology

View system prompts and framework documentation

1.0.0
Prompt Version
🔍

Try It Yourself

Experience the comparison as a participant

Research Context

Academic background for this project

This tool was developed for RELI E-1730: Mindfulness, AI, and Ethicsat Harvard Extension School (Spring 2026). It examines how different AI alignment philosophies manifest in practice when applied to mental health support contexts.

Research Questions

  1. How do containment-based (HSI) and character-based (Constitutional AI) approaches differ in their responses to sensitive mental health scenarios?
  2. Can a hybrid approach effectively combine the safety of containment with the warmth of character-based engagement?
  3. How do Buddhist ethics principles (particularly those around attention and compassion) inform our evaluation of AI alignment frameworks?

Theoretical Framework

The comparison draws on Peter D. Hershock's work on Buddhist ethics and technology, Amanda Askell's character-based AI alignment research at Anthropic, and Mustafa Suleyman's humanist superintelligence framework.