AI safety can learn from past sociotechnical disasters, says new paper

A new paper published on arXiv argues that the development and deployment of AI systems can learn critical lessons from past sociotechnical disasters. The authors highlight that catastrophic events like Chernobyl, Three Mile Island, and the Challenger disaster were not solely due to unforeseen technical interactions but also stemmed from unaddressed social, political, and economic factors. The paper proposes that AI development should incorporate improved risk perception, traceability of responsibilities, and a holistic approach to safety that includes social and organizational dynamics as engineering concerns. AI

IMPACT Suggests a more holistic approach to AI safety, integrating social and organizational factors alongside technical design.

RANK_REASON The cluster contains an academic paper discussing AI safety and lessons from past technological failures. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• Bhopal
• Chernobyl disaster
• Fukushima Daiichi Nuclear Power Plant
• Space Shuttle Challenger disaster
• Three Mile Island

• paper
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →