THE_NEW_RULES

Free Labor for AI Training: The Great Open Source Betrayal?

Chapter 11: Open Source Economics

"Every open source contribution is now training data for models that companies sell back to us. We're not building commons; we're providing free R&D for trillion-dollar corporations."

The book exposes how our open source contributions become training data for commercial AI. GitHub Copilot, trained on our code, now competes with us. Is open source dead, or just transformed into unpaid AI training?

Questions for Debate:

The Exploitation Reality

  • Did we consent to our code training commercial AI?
  • Should contributors be compensated when their code trains models?
  • Is this the biggest theft in software history?

The Contribution Dilemma

  • Why contribute to open source if it trains your replacement?
  • Are we funding our own obsolescence?
  • Should we stop contributing until licensing is fixed?

The Corporate Capture

  • Have corporations killed the open source spirit?
  • Is "open source" now just free labor for big tech?
  • Can open source survive corporate exploitation?

Share Your Experience:

The Contributors:

  • How do you feel seeing your code in AI suggestions?
  • Has AI training changed your contribution behavior?
  • Would you contribute differently knowing it trains AI?

The Maintainers:

  • Should projects forbid AI training on their code?
  • How do you balance openness with protection?
  • What licenses actually prevent AI training?

The Economic Analysis:

The Value Transfer:

  • Billions in volunteer hours → Trillion dollar AI companies
  • Who captured the value of open source?
  • Should there be retroactive compensation?

The New Models:

  • Should open source require AI training licenses?
  • Can we create "AI-resistant" licenses?
  • Would paid open source be better than free?

The Incentive Breakdown:

  • What incentive remains to contribute?
  • Are we selecting for developers who don't understand exploitation?
  • Will quality contributors abandon open source?

The Legal Questions:

The License Loopholes:

  • Do current licenses permit AI training?
  • Is training "use" or "distribution"?
  • Can licenses be retroactively changed?

The Copyright Crisis:

  • Who owns AI output trained on open source?
  • Is AI generation derivative work?
  • Are we heading for massive litigation?

The Philosophical Divide:

The Purist Position:

  • Information wants to be free, including for AI
  • Restricting training betrays open source principles
  • This is evolution, not exploitation

The Pragmatist Position:

  • Open source needs protection from extraction
  • Commercial AI should pay for training data
  • Sustainability requires new models

The Radical Position:

  • Burn it all down and start over
  • No more free contributions to corporate infrastructure
  • Create alternative, protected ecosystems

The Future Scenarios:

Scenario 1: Mass Exodus

  • Contributors leave for paid or protected platforms
  • Open source becomes corporate-only
  • Innovation moves behind paywalls

Scenario 2: New Licenses

  • AI-specific licenses emerge
  • Training requires payment or permission
  • Balance between open and protected

Scenario 3: Status Quo

  • Nothing changes, exploitation continues
  • Developers accept training as inevitable
  • Open source becomes AI feeding ground

Your Choice:

Will you keep contributing to open source knowing it trains AI that might replace you?

Is this evolution or exploitation?

Loading comments...