Skip to main contentSkip to navigationSkip to search

Emergecny Actions

1. 5-Year Moratorium on All Large AI Models

STOP building models larger than 50 billion parameters.

Justification:

  • We don't know what we're doing
  • Job displacement is outpacing retraining
  • Energy/water costs are unsustainable
  • Deepfakes are destroying truth
  • Need time for society to adapt

Exceptions:

  • Medical research (cancer and drug discovery)
  • Climate modeling
  • Scientific research (non-commercial)
  • All require FAIA approval + environmental impact statements

Penalty for Violation: $10 billion fine + criminal charges for executives + model is destroyed

2. Break Up Frontier AI

OpenAI, Google, Microsoft, Meta, and Amazon: You're too big, you're done.

Forced Divestiture:

Microsoft:

  • Azure Cloud (separate company)
  • OpenAI stake (divest completely, OpenAI becomes public benefit corporation)
  • Copilot AI (spun off)
  • Gaming division (separate)

Google:

  • Search (separate from everything)
  • YouTube (separate)
  • Cloud/AI (separate)
  • Android (separate)
  • DeepMind (public research institute, not for-profit)

Amazon:

  • AWS (separate)
  • Retail (separate)
  • Alexa/AI (separate or shut down)

Meta:

  • Instagram (separate)
  • WhatsApp (separate)
  • Facebook (what's left)
  • AI research (public institute)

Timeline: 18 months to complete breakups or face nationalization

Why: Too much concentrated power. These companies control: infrastructure (cloud), data (search/social), AI models, and hardware supply chains.

That's feudalism.

3. Nationalize Critical AI Infrastructure

If companies resist breakup, federal government seizes:

  • All data centers (eminent domain, pay fair market value)
  • Cloud computing infrastructure becomes public utility
  • AI models become public goods (open source)
  • Research divisions become National AI Research Labs (like NASA, but for AI)

Why: Can't have private companies controlling society's intelligence infrastructure. It's like letting Exxon own all roads.

Cost: $500B-1T to buy infrastructure (amortize over 30 years = $16-33B/year)

Revenue: Charge businesses for cloud use (like electricity) and revenue funds operations. Likely revenue-neutral.