Ghost in Your Computer — PhantomOS：互動式 AI 幽靈桌面體驗
👻 Ghost in Your Computer — PhantomOS
PhantomOS 是一款以 Bytegeist 為核心的互動式桌面體驗，讓 AI 幽靈聊天、切換四種情緒並觸發隨機數碼事件。項目採用 Meta Llama 3.1 8B Instruct 開放權重模型，透過 Backboard 提供託管推理；桌面體驗為安全模擬，不會存取真實系統檔案或命令，並設有 AI 無法使用時的 Demo Mode。
This is a submission for the Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the Machine
What I Built
I built Ghost in Your Computer — PhantomOS, an interactive desktop experience featuring Bytegeist, a mischievous AI ghost who lives inside a fictional operating system.
What if your computer had a ghost with its own personality?
Instead of creating another ordinary chatbot, I wanted to make the experience feel like interacting with a digital creature. Bytegeist can chat, react, change moods, create unexpected digital incidents, and even take temporary possession of the interface!
Users can explore four different moods: Playful, Mischievous, Dramatic, and Slightly Spooky.
Some key features:
- 👻 Interactive AI ghost with a unique personality
- 💬 AI-powered conversation with session memory
- 🎭 Four personality moods
- ⚡ Random digital incidents
- 🖥️ Possession Mode with CRT effects and screen animations
- 🔊 Optional sound effects
- 📊 Live personality telemetry
- 🛡️ Safe simulated desktop experience — no access to real system files or commands
I interpreted the theme as bringing a little supernatural personality into everyday technology — funny, unpredictable, and just a little spooky!
Demo
🌐 Live App: https://ghost-in-your-computer.onrender.com
🎥 Video Demo: https://youtu.be/ch6dU9f2LrM
Code
💻 GitHub Repository: https://github.com/kusumaranikusumarani3232-hub/Ghost-in-Your-Computer
How I Built It
I built PhantomOS using React, Vite, and FastAPI.
For the AI core, I used Meta Llama 3.1 8B Instruct, an open-weight model, through Backboard.
The model powers Bytegeist's conversations and personality-driven responses. Backboard provides the hosted inference connection, while the application handles the interface, mood system, digital incidents, and interactive effects.
The project also includes a clearly labeled Demo Mode as a fallback when live AI is unavailable.
Technology Stack:
- Frontend: React 19, Vite, CSS
- Backend: Python, FastAPI
- AI Model: Meta Llama 3.1 8B Instruct
- AI Integration: Backboard
- Deployment: Render
Why Does Open Innovation Matter?
Open innovation gives developers the freedom to experiment, build, and create experiences beyond traditional AI chat interfaces.
Using an open-weight model like Llama 3.1 allowed me to build Bytegeist around a model whose underlying weights are openly available, rather than depending entirely on a closed proprietary model.
It also gives developers more flexibility to explore different inference providers and deployment approaches as their projects grow.
For me, open innovation makes experimentation more accessible and gives independent developers the opportunity to build creative AI experiences.
The Spirit of It
This project started with a simple question:
What if your computer wasn't haunted by bugs, but by a ghost with an attitude? 👻
I wanted to make something that would make people smile, surprise them, and give them a different experience from a typical AI chatbot.
Meet Bytegeist — your computer isn't broken. It's just haunted!
Thanks to Hacktoberfest and the open-source AI community for inspiring this experiment.
來源：DEV Community · dev.to