Results for "provable safety"
SaMD
Intermediate
Software regulated as a medical device.
Artificial General Intelligence
Frontier
AI capable of performing most intellectual tasks humans can.
x-Risk
Advanced
Existential risk from AI systems.
Slow Takeoff
Advanced
Incremental capability growth.
AI Boxing
Advanced
Isolating AI systems.
Tripwire
Advanced
Signals indicating dangerous behavior.
Shutdown Problem
Advanced
Ensuring AI allows shutdown.
Power-Seeking Behavior
Advanced
Tendency to gain control/resources.
AI Treaty
Intermediate
International agreements on AI.
Model Release Control
Intermediate
Restricting distribution of powerful models.
Alignment Research
Intermediate
Research ensuring AI remains safe.