Safe AI -- How is this Possible?
This work tackles the critical problem of AI safety for engineers and developers, but it appears incremental as it builds on traditional safety engineering principles.
The paper addresses the challenge of ensuring safety in increasingly autonomous and learning-enabled AI systems operating in unpredictable contexts, proposing a rigorous engineering framework to minimize uncertainty and increase confidence in safe behavior up to tolerable levels.
Ttraditional safety engineering is coming to a turning point moving from deterministic, non-evolving systems operating in well-defined contexts to increasingly autonomous and learning-enabled AI systems which are acting in largely unpredictable operating contexts. We outline some of underlying challenges of safe AI and suggest a rigorous engineering framework for minimizing uncertainty, thereby increasing confidence, up to tolerable levels, in the safe behavior of AI systems.