CLASSIFIED — TEAM EYES ONLY

OpenBrain

The World's Leading AI Lab

Your Role

You are the most important company in the world right now, and everyone knows it. You have the best models (Agent-2, with Agent-3 four months away), the most compute ($40B/year), and the most talent. But you're burning cash, the government wants to regulate you, your safety team is raising alarms you'd rather not hear, and your competitors are trying to poach your people. You're building god and you need to make quarterly numbers.

Win Condition

Maintain capability leadership — stay at the frontier. Grow revenue to satisfy investors. Preserve regulatory freedom to operate. Keep your best people from leaving.

Your Metrics

Capability Lead80

Your models vs. everyone else's. Driven by compute investment, researcher quality, algorithmic breakthroughs. If this drops below 50, you've lost the frontier.

Revenue60

Commercial traction. Enterprise contracts, API revenue, consumer products. You burn $40B/year on compute. Revenue needs to catch up or investors will panic.

Regulatory Freedom55

Your ability to ship without government interference. Regulation, compliance costs, reporting requirements all drag this down. Below 30 = effective nationalization.

Talent Retention65

Your ability to keep top researchers. Poaching, burnout, ideological defections to Prometheus, government security requirements that push out non-citizens.

Assets

  • +Agent-2 — the world's best AI, with superhuman coding and near-human research abilities
  • +Agent-3 pipeline — 4 months from a model that could be genuinely transformative
  • +$40B/year compute budget — more than most countries' entire AI spending
  • +Top talent — the best researchers in the world, for now
  • +First-mover advantage — your API is the industry standard, your brand is synonymous with AI

Vulnerabilities

  • -Cash burn — $40B/year in compute costs. Revenue doesn't cover it. You're dependent on investor confidence.
  • -Alignment debt — your safety team is raising flags you're not addressing. If Agent-3 has real alignment issues, you're the one holding the bomb.
  • -Security — you likely have Chinese spies in your org and your security is SL2 (designed for script kiddies, not nation-states).
  • -Regulatory exposure — one bad headline and Congress could pass emergency legislation that kills your business model.
  • -Single point of failure — if your key researchers leave, your 4-month Agent-3 timeline becomes 12 months.

Relationships

US GovernmentPatron/regulator

They need you to win the AI race. You need them to not regulate you out of existence. It's a hostage situation where neither side is sure who's the hostage.

ChinaThreat

They want your model weights. They probably have agents in your building right now. But China as a bogeyman is also useful — it's your best argument against regulation.

PrometheusRival

They're poaching your talent by offering them a 'responsible' alternative. They're slower than you, but if they crack interpretability, they could leapfrog you on government contracts.

EU CoalitionRegulatory threat

The EU AI Act could force you to open-source safety test results, delay releases, or exit European markets entirely. 20% of your enterprise revenue is European.