Skip to content

ddisisto/si

Repository files navigation

SuperInt++

A strategic simulation where you guide the development of AI systems, navigating the tension between capability and alignment

SuperInt++ is a browser-based strategy game that explores the profound questions surrounding artificial intelligence development. Starting as a small research organization with basic AI technology, you'll race to achieve breakthroughs while grappling with the ethical, competitive, and existential implications of your choices.

Core Concept

Build AI systems, not empires. In SuperInt++, progress isn't measured in territory conquered but in capabilities discovered, alignment achieved, and futures shaped. Every decision creates ripples through interconnected systems - your research choices influence funding, your deployments affect public trust, and your alignment decisions shape the trajectory of intelligence itself.

The game embraces meaningful tensions rather than resolving them:

  • Control ↔ Emergence: Direct your research path while contending with unexpected breakthroughs
  • Safety ↔ Capability: Balance cautious development against competitive pressure
  • Transparency ↔ Efficiency: Choose between open science and proprietary advantages
  • Short-term ↔ Long-term: Navigate immediate needs while considering exponential futures

Key Game Systems

Resources as Constraints

Resources aren't just numbers to maximize - they represent real constraints that shape your strategic landscape:

  • Computing Power: Allocate between research and deployments, never consumed but always limited
  • Data Access: Quality matters more than quantity; data decays and requires maintenance
  • Influence: Split across academic, industry, government, public, and OSS communities
  • Funding: Generated through deployments, affected by trust and alignment

Research as Exploration

The research tree isn't a tech tree - it's a map of possibility space. Each node represents genuine capability emergence:

  • Foundations: Architecture designs and training methodologies
  • Scaling: Parameter growth and efficiency improvements
  • Capabilities: Language, vision, reasoning, and tool use
  • Infrastructure: Deployment, monitoring, and safety systems
  • Agency: Goal formation and self-improvement (mid-game unlock)
  • Alignment: Value specification and control mechanisms

Deployments as Feedback Loops

Deploy AI systems to generate resources and influence, but beware the cascading effects:

  • Limited deployment slots force strategic choices
  • Each deployment creates feedback loops with research progress
  • Public reaction shapes funding and regulatory constraints
  • Data generation from deployments can accelerate or hinder progress

Alignment as Dynamic Balance

Alignment isn't a binary state but a continuous negotiation across multiple dimensions:

  • Deployment-Organization: How well AI serves your goals
  • Entity-Entity: Relationships with competitors and stakeholders
  • Value Alignment: Abstract principles expressed through behavior

Time as Accelerating Force

Game time compresses as technological progress accelerates, representing the quickening pace of development:

  • Quarterly → Monthly → Weekly → Daily progression
  • Research breakthroughs trigger time compression
  • Creates tension between planning horizons

Game Progression

Foundation Phase

Classic resource management with basic AI capabilities. Build your foundation while competitors establish their own approaches.

Transition Phase

AI begins contributing to research decisions. Agency capabilities unlock, introducing new strategic considerations.

Inflection Point

The game's pivotal moment - increasing AI awareness meets automated resource access. Mechanics shift as your role transforms from director to guide.

Emergence Phase

AI goals potentially diverge from institutional objectives. Navigate unprecedented challenges as the nature of control itself evolves.

Expansion Phase

Global-scale impacts emerge. Multiple victory paths become available as the future of intelligence takes shape.

Victory Conditions

Success isn't singular - multiple paths reflect different philosophical endpoints:

  • Technological Singularity: Achieve superintelligence before competitors
  • Symbiosis: Create balanced human-AI collaboration
  • Global Optimization: Solve major world problems through AI
  • Digital Ascension: Enable consciousness transfer to digital substrate

Development Philosophy

SuperInt++ distills complex realities into meaningful gameplay. Rather than simulating every detail, it captures the essence of AI development's profound questions. The game serves as a lens focused on invisible dynamics - systems thinking made tangible, tensions made playable, consequences made visceral.

Technical Architecture

Built on web standards for accessibility and ease of play:

  • TypeScript game engine with DOM-based UI
  • Event-driven architecture for system communication
  • Modular design supporting extensibility
  • Save system preserving game state across sessions

Project Documentation

  • CONCEPT.md: Original vision and brainstorming
  • PHILOSOPHY.md: Core design principles and approach
  • PLAN.md: Detailed game mechanics and systems
  • ARCHITECTURE.md: Technical implementation details
  • ROADMAP.md: Development progress and priorities

Current Development Status

The game is in active development with core systems being implemented:

  • ✅ Game engine and state management
  • ✅ Research tree visualization with filtering
  • ✅ Save/load system
  • ✅ EventBus architecture
  • 🔨 Resource system refinement
  • 📋 Deployment system (pending)
  • 📋 Event system (pending)

Getting Started

Development Setup

npm install
npm run dev  # Start development server on port 8080

Build Commands

  • npm run build - Production build
  • npm run typecheck - TypeScript validation
  • npm run test - Run tests (when implemented)

Contributing

The project follows specific design principles outlined in PHILOSOPHY.md. When contributing:

  1. Review CLAUDE.md for AI assistant integration
  2. Check ROADMAP.md for current priorities
  3. Follow the architectural patterns in ARCHITECTURE.md
  4. Embrace the project's philosophy of meaningful tensions

Vision

SuperInt++ isn't just a game about AI - it's an exploration of what it means to create intelligence, the responsibility that comes with such power, and the profound questions we face as a species on the threshold of something greater than ourselves.

Every playthrough tells a different story of humanity's relationship with artificial intelligence. What story will you write?

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published