Managing the risks posed by Artificial General Intelligence

管理通用人工智能带来的风险

• 批准号: DP200100399
• 负责人: Prof Paul Salmon
• 金额: $ 32.36万
• 依托单位: University of the Sunshine Coast
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Projects
• 财政年份: 2020
• 资助国家: 澳大利亚
• 起止时间: 2020-12-16 至 2023-12-15
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DP200100399
• 关键词: Managing; risks; posed; Artificial; General

It is widely acknowledged that a failure to implement appropriate controls for the next generation of Artificial Intelligence, Artificial General Intelligence (AGI), could have catastrophic consequences, including in the worst case - the extinction of the human race. This research aims to forecast the risks associated with AGI systems and identify the controls required to ensure that risks and existential threats are minimised. The expected outputs will provide designers, organisations, regulators and governments with a framework to support the design, implementation, and management of safe and efficient AGI systems. This will ensure that the potential far-reaching benefits of AGI are realised without undue threat to society.

人们普遍认为，未能对下一代人工智能，即人工通用智能（AGI）实施适当的控制，可能会产生灾难性的后果，包括在最坏的情况下-人类的灭绝。本研究旨在预测与AGI系统相关的风险，并确定确保风险和存在威胁最小化所需的控制措施。预期成果将为设计者、组织、监管机构和政府提供一个框架，以支持安全高效的AGI系统的设计、实施和管理。这将确保AGI的潜在深远利益得以实现，而不会对社会造成不当威胁。