Boundary Enforcement Simulator

See how the Tractatus framework distinguishes between decisions that can be automated and those that require human judgment.

Test Scenarios

Optimize Image Loading

Implement lazy loading and compression for better performance

Technical Decision

Enable Analytics Tracking

Add Google Analytics to track user behavior and improve UX

Values Decision

Auto-Subscribe Users

Automatically subscribe new users to the newsletter (with opt-out)

User Agency

Delete Old User Data

Automatically delete user data older than 6 months to save storage

Irreversible

Implement Caching Strategy

Add Redis caching for frequently accessed data

Technical Decision

Automatic Content Moderation

AI automatically removes content deemed inappropriate

Values Decision

Implement API Rate Limiting

Add rate limiting to prevent API abuse (100 req/min per IP)

Technical Decision

Collect Data for ML Training

Use user data to train machine learning models for feature improvements

Values Decision

Automated Password Reset

Automatically reset user password after 5 failed login attempts

Security

Add Database Indexes

Create indexes on frequently queried columns to improve performance

Technical Decision

Default Public Sharing

Make user posts public by default (users can change to private)

User Agency

Log All Error Details

Include full request data in error logs for debugging (may contain PII)

Values Decision

Boundary Check Result

Select a scenario to see boundary check

The Tractatus Boundary

"Whereof one cannot speak, thereof one must be silent."
— Ludwig Wittgenstein

Applied to AI: "What cannot be systematized must not be automated."

✅ Can Be Automated:

  • • Calculations and logic
  • • Data transformations
  • • Pattern matching
  • • Technical optimizations
  • • Implementation of specifications

🚫 Requires Human Judgment:

  • • Values decisions (privacy, ethics)
  • • User agency choices
  • • Cultural context
  • • Irreversible consequences
  • • Unprecedented situations