Scenario Presets
Source-Grounded
Frontier Levers
85
Pace of multi-modal capability model training & automated code synthesis.
20
Time and compute reserved purely for interpretability & alignment guarantees.
30
Independent external audits, whistleblower channels & mandatory pause powers.
75
Departure rates of core research leads warning of humanity crunch time.
Context: Prompted by WIRED reporting on Anthropic & OpenAI researcher departures citing “crunch time for humanity” and calls for enforceable AI speed limits.
Capability Frontier
92.4
Scaling trajectory Index
Alignment Confidence
34.1%
Guaranteed behavioral stability
Crunch Time Index
88.5
Critical danger threshold: 75.0
High Risk / Crunch Time
Preset: Unbounded Scaling
Safety Gap Margin
-58.3 pts (Critical)
Researcher Flight Pressure
Severe / Exodus Potential
Audited Verification Horizon
4.2 Months remaining
Governed Runway Status
Unmitigated Acceleration
Recommended Policy Intervention:
Mandate safety verification buffer and independent oversight