Skip to content

Tracks Review Resources

Control

The high-stakes control roadmap (talk)
What is High Stakes?
Win/continue/lose scenarios and execute/replace/audit protocols
AI catastrophes and rogue deployments
Rogue internal deployments via external APIs
A basic systems architecture for AI agents that do autonomous research
Monitoring and Mitigation
Practical challenges of control monitoring in frontier AI deployments
Blocking live failures with synchronous monitorsOptional
Ctrl-Z: Controlling AI Agents via Resampling
Environment construction
Why it's hard to make settings for high-stakes control research
Untrusted advice for AI control (guided)
Basic legibility protocols improve trusted monitoring (guided)

AI ControlHigh-stakes controlThe high-stakes control roadmap (talk)

Module 5: High-stakes control

The high-stakes control roadmap (talk)

~30 min

The high stakes control roadmap: now to the singularity — Aryan Bhatt

Sign in to track progress

PreviousAn overview of areas of control work Next What is High Stakes?