🧠 StillMe - Self-Evolving AI System

A revolutionary AI system that learns and evolves from the internet daily, becoming smarter with each interaction.

🌟 What is StillMe?

StillMe is a Self-Evolving AI System that continuously learns from the internet, adapts to new information, and evolves through different developmental stages - just like a growing organism. Unlike traditional AI systems that remain static, StillMe gets smarter every day.

🎯 Core Concept

• 🧬 Evolutionary Learning: AI progresses through stages (Infant → Child → Adolescent → Adult)
• 📚 Multi-Source Learning: RSS feeds + Public APIs integration
• 🌐 Real-time Data: Live data from multiple trusted sources with transparency
• 🛡️ Ethical Filtering: Comprehensive ethical content filtering with complete transparency
• 📊 Transparent Dashboard: Complete visibility into all learning sources and data
• 💬 Interactive Chat: Communicate with your evolving AI assistant

🛡️ Our Uncompromising Commitment

🌟 100% Transparency - Nothing to Hide

• Every line of code is public - no "black box", no proprietary algorithms
• Every API call is visible - see exactly what AI learns from and when
• Every decision is transparent - from ethical filtering to quality assessment
• Complete audit trail - full history of all learning decisions and violations

🎯 Ethical AI - Our Highest Priority

We believe that ethics isn't a feature - it's the foundation. StillMe is built with unwavering principles:

• Safety First: Harmful content filtered at the source
• Cultural Fairness: Respects global diversity and perspectives
• Full Accountability: Every mistake is public and corrected
• Community Control: You decide what's acceptable, not corporations

"We challenge the AI community to choose: Support transparency and ethics, or remain silent and admit they don't care."

🔒 Privacy & Data Protection

• No personal data collection - learns only from public sources
• Self-hosted codebase - you maintain complete control over your data
• Delete anytime - your data, your rules, your control

🛡️ Ethical AI Transparency

StillMe features the world's first completely transparent ethical filtering system:

• Complete Visibility: All ethical violations are logged and visible
• Open Source: Filtering rules and algorithms are publicly available
• Community Driven: Blacklist and rules can be managed by the community
• Audit Trail: Full history of all ethical decisions and violations
• Configurable: Ethics level can be adjusted based on community needs

This transparency ensures StillMe learns responsibly while maintaining community trust.

🚀 Quick Start

# Clone repository
git clone https://github.com/anhmtk/stillme_ai_ipc.git
cd stillme_ai_ipc

# Install dependencies
pip install -r requirements.txt

# Start backend
python start_backend.py

# Start frontend (new terminal)
python start_frontend.py

📊 Dashboard Features

• Evolution Panel: Real-time AI stage and progress tracking
• Ethical Filter: Complete transparency into ethical decisions
• Learning Analytics: Historical progress with flexible timeline analysis
• Community Controls: Manage ethical rules and blacklist
• Raw Data Access: View actual API responses for verification

🧬 AI Evolution Stages

StillMe progresses through distinct developmental stages:

🍼 Infant Stage (0-100 learning sessions)

• Basic pattern recognition
• Simple content categorization
• High safety focus
• Manual approval required

👶 Child Stage (100-500 sessions)

• Improved content understanding
• Basic reasoning capabilities
• Selective auto-approval
• Enhanced safety protocols

🧑 Adolescent Stage (500-1000 sessions)

• Advanced reasoning
• Context awareness
• Smart auto-approval
• Balanced learning approach

🧠 Adult Stage (1000+ sessions)

• Sophisticated understanding
• Complex reasoning
• Autonomous learning
• Expert-level knowledge

🚀 The Vision: Fully Autonomous AI Evolution

🧠 Self-Evolution Goal

StillMe aims to become a fully autonomous learning AI:

• Self-Assessment: Knows what it knows and what it doesn't
• Proactive Learning: Actively seeks new knowledge sources
• Self-Optimization: Adjusts learning process based on effectiveness
• Autonomous Review: Gradually reduces human dependency as trust builds

🔬 Future Evolution Pathways

We open these questions to the community:

• AI Self-Coding? - Should StillMe learn to debug and improve itself?
• Red Team vs Blue Team? - AI attacking and defending itself for enhanced security?
• Multi-Agent Collaboration? - Multiple StillMe instances collaborating on complex problems?
• Cross-Domain Learning? - Expanding from AI to medicine, science, and other fields?

"This isn't our roadmap - it's a community discussion. What direction do you want AI's future to take?"

🔧 Architecture

Backend (FastAPI)

• Learning Engine: Core evolutionary learning system
• RSS Pipeline: Multi-source content fetching
• Ethical Filter: Comprehensive safety system
• Memory Management: Advanced knowledge storage
• API Integration: Public APIs for diverse content

Frontend (Streamlit)

• Dashboard: Real-time monitoring and control
• Evolution Panel: AI stage visualization
• Ethical Controls: Community management tools
• Analytics: Historical learning data
• Chat Interface: Interactive AI communication

Database (SQLite)

• Learning Sessions: Track AI evolution progress
• Content Proposals: Store learning opportunities
• Memory Items: Advanced knowledge storage
• Ethical Violations: Complete audit trail

📚 Learning Sources

StillMe learns from diverse, trusted sources:

RSS Feeds

• Hacker News, Reddit, GitHub
• TechCrunch, ArXiv, Stack Overflow
• Medium, Academic sources
• News outlets, Subreddits

Public APIs

• NewsAPI, GNews
• Weather, Finance data
• Translation services
• Image understanding APIs

🛡️ Ethical Safety Filter

StillMe features a comprehensive ethical content filtering system that ensures responsible AI learning:

Core Principles

• Beneficence: Content must benefit learning and users
• Non-Maleficence: Blocks harmful, toxic, or dangerous content
• Autonomy: Protects privacy and personal information
• Justice: Prevents biased or discriminatory content
• Transparency: Complete visibility into all filtering decisions
• Accountability: Full audit trail of ethical violations

Filtering Capabilities

• Input Filtering: Blocks harmful content at the source (RSS/API)
• Content Analysis: Detects toxicity, bias, and sensitive topics
• PII Protection: Automatically identifies and blocks personal information
• Source Validation: Flags unreliable or suspicious sources
• Real-time Monitoring: Continuous ethical compliance checking

Transparency Features

• Violation Logging: Complete history of all ethical violations
• Dashboard Integration: Real-time ethical metrics and statistics
• Community Management: Blacklist keywords and rules can be managed
• Audit Trail: Full transparency into all ethical decisions
• API Access: Programmatic access to ethical statistics and controls

🔧 Configuration

Environment Setup

# Copy environment template
cp env.example .env

# Edit with your API keys
DEEPSEEK_API_KEY=sk-REPLACE_ME
OPENAI_API_KEY=sk-REPLACE_ME
ANTHROPIC_API_KEY=sk-REPLACE_ME

# Learning Configuration
MAX_DAILY_PROPOSALS=50
AUTO_APPROVAL_THRESHOLD=0.8
LEARNING_SESSION_HOUR=9

# Notification Configuration
SMTP_HOST=smtp.gmail.com
SMTP_PORT=587
SMTP_USERNAME=your_email@gmail.com
SMTP_PASSWORD=REPLACE_ME_WITH_YOUR_APP_PASSWORD
TELEGRAM_BOT_TOKEN=REPLACE_ME_WITH_YOUR_BOT_TOKEN
TELEGRAM_CHAT_ID=your_chat_id

# Notification Settings
NOTIFY_LEARNING=true
NOTIFY_ERRORS=true

📊 API Endpoints

Core Learning APIs

• GET /api/learning/sessions - Get learning sessions
• POST /api/learning/sessions/run - Trigger learning session
• GET /api/learning/evolution/stage - Get current AI stage
• GET /api/learning/stats - Get learning statistics

Content Management APIs

• GET /api/learning/proposals - Get learning proposals
• POST /api/learning/proposals/{id}/approve - Approve proposal
• POST /api/learning/proposals/{id}/reject - Reject proposal
• GET /api/learning/rss/pipeline-stats - Get RSS pipeline stats
• POST /api/learning/rss/fetch-content - Fetch content manually

Ethical Safety APIs

• GET /api/learning/ethics/stats - Get ethical filter statistics
• POST /api/learning/ethics/check-content - Test content for ethical compliance
• GET /api/learning/ethics/violations - Get ethical violation history
• POST /api/learning/ethics/clear-violations - Clear violation log
• POST /api/learning/ethics/add-blacklist-keyword - Add keyword to blacklist
• GET /api/learning/ethics/blacklist-keywords - Get current blacklist

Advanced Features APIs

• GET /api/learning/knowledge/stats - Get knowledge consolidation stats
• POST /api/learning/knowledge/consolidate - Trigger knowledge consolidation
• GET /api/learning/memory/stats - Get advanced memory management stats
• POST /api/learning/memory/optimize - Optimize memory system

Analytics APIs

• GET /api/learning/analytics/historical - Get historical learning data
• GET /api/learning/analytics/comparison - Compare learning periods
• GET /api/learning/analytics/trends - Get learning trends analysis

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for details.

Areas for Contribution

• UI/UX Improvements: Dashboard enhancements, mobile responsiveness
• Learning Sources: Add new RSS feeds and API integrations
• Ethical Filtering: Improve safety algorithms and rules
• Documentation: API docs, tutorials, guides
• Testing: Unit tests, integration tests, performance tests

🤝 Join The Movement - Not Just A Project

🌍 Why You Can't Ignore StillMe

Whether you support or oppose it, StillMe forces you to pay attention:

• If you support: You're helping build the future of transparent AI
• If you oppose: You need to monitor what's happening
• Either way: You can't deny the impact of a 100% open project

💬 We Welcome All Voices

• Supporting? - Contribute code, ideas, resources
• Criticizing? - Point out flaws, suggest improvements
• Skeptical? - Check the code, test the system, find vulnerabilities
• Interested? - Follow, share, discuss

"In a world of AI 'black boxes', our transparency is our strongest weapon."

💖 Support The Mission

This project is maintained by passion and community contributions. If you believe in the mission:

• 📧 Donate: [Donation link will be announced] - help maintain servers and development
• 🔧 Contribute: Code, docs, testing, translations - every contribution matters
• 📢 Spread The Word: Share with developer and researcher communities
• 🎯 Provide Feedback: Criticize, suggest, propose - help us improve

🎯 The Bottom Line

StillMe isn't just a product - it's a STATEMENT:

1. AI must be transparent - no exceptions
2. Ethics must be foundational - not an add-on feature
3. Community must control - not corporations
4. Future must be discussed - not imposed

---

Join us. Watch us. Critique us. But you can't ignore us.

Because in the darkness of AI "black boxes", our transparency is the light.

📄 License

MIT License - see LICENSE for details.

🙏 Acknowledgments

StillMe is built with love and dedication to create a truly transparent, ethical AI system. Special thanks to:

• OpenAI for GPT models and API access
• DeepSeek for advanced AI capabilities
• Anthropic for Claude integration
• The Open Source Community for inspiration and support

---

StillMe - Self-Evolving AI System with Complete Ethical Transparency 🤖✨

"The future belongs to AI systems that can learn, adapt, and evolve. StillMe is that future, today."