README.md

---
title: Build a Bot
emoji: 🤖
colorFrom: blue
colorTo: green
sdk: streamlit
sdk_version: 1.32.0
app_file: app.py
pinned: false
license: mit
short_description: "AI literacy through experimentation"
---

# Build a Bot

**AI Literacy Through Experimentation**

Understand how LLMs actually work by experimenting with different AI personalities. Build discernment, not fear.

## What This Does

Switch between different AI "personalities" and notice how the same question gets different responses:

- **Neutral** — Standard assistant
- **Warm & Supportive** — High synthetic intimacy
- **Clinical & Detached** — Low warmth, facts only
- **Boundaried (GSPT style)** — Warm without performing relationship
- **Sycophantic** — Over-validation pattern

## What You'll Learn

- How prompts create the illusion of personality
- What synthetic intimacy looks and feels like
- Why AI often agrees too much (sycophancy)
- How to use AI tools with healthy boundaries

## Credits

Created by [Jocelyn Skillman, LMHC](http://www.jocelynskillman.com)