AI Safety for Everyone

AI safety encompasses far more than existential risk: The paper's systematic literature review identifies "a vast array of concrete safety work that addresses immediate and practical concerns with current AI systems," including adversarial robustness and interpretability—issues affecting AI systems in use today, not just theoretical future scenarios.

Overemphasizing existential risk may backfire: The authors warn that framing AI safety primarily around existential threats "could lead the public to mistakenly view AI safety as focused solely on existential scenarios" and "risks creating resistance to safety measures among those who disagree with predictions of existential AI risks," potentially undermining support for necessary safeguards.

AI safety builds on established safety practices: The research shows that AI safety "naturally extends existing technological and systems safety concerns and practices," meaning many safety principles already proven effective in other industries (aviation, nuclear power, etc.) can be adapted for AI systems.

Diverse perspectives strengthen the field: The paper advocates for "an epistemically inclusive and pluralistic conception of AI safety that can accommodate the full range of safety considerations, motivations, and perspectives," suggesting that researchers approaching safety from different angles—not just existential risk—are all making valuable contributions.

Current AI systems need safety attention now: Beyond future concerns, the paper emphasizes that practical safety challenges with existing AI systems deserve immediate focus and resources, making AI safety relevant to anyone using or deploying AI today.