ITR---Reinventing Artificial Intelligence

ITR---重塑人工智能

• 批准号: 0218861
• 负责人: Patrick Winston
• 金额: --
• 依托单位: Massachusetts Institute of Technology
• 依托单位国家: 美国
• 项目类别: Continuing grant
• 财政年份: 2002
• 资助国家: 美国
• 起止时间: 2002-09-01 至 2005-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0218861&HistoricalAwards=false
• 关键词: ITR; Reinventing; Artificial; Intelligence

The Genesis Group of the MIT Artificial Intelligence Laboratory is workingon a computational theory of human intelligence. Their work is grounded intwo key assumptions: first, that humans think with their language, vision,and motor systems, and their interaction; and second that humans can thinkabstractly because they can build on a foundation of thinking about concreteevents in the physical world.Their plan includes the development of a testbed that features paths,agents, causes, both language and visual inputs and outputs, complex statetransitions, and support for abstract reasoning in abstract worlds.They propose to exploit both symbolic and nonsymbolic representations. Thesymbolic representations will include a representation for describing themovement of animals and artifacts along trajectories and a representationfor describing state transitions in terms of a vocabulary of qualitativechanges. The nonsymbolic representations will include memory traces lyingclose to experienced sensory inputs.The Genesis Group expects success to lead not only to a betterunderstanding of natural intelligence buy also, in the long term, toimportant practical results, such as computer applications with genuinecommonsense and educational applications that exploit an understanding ofhow best to engage human linguistic and visual problem solving faculties.

麻省理工学院人工智能实验室的创世纪小组正在研究人类智能的计算理论。他们的工作基于两个关键假设：首先，人类通过语言、视觉和运动系统以及它们之间的相互作用进行思考；第二，人类可以抽象地思考，因为他们可以建立在思考物理世界中具体事件的基础上。他们的计划包括开发一个测试平台，该平台具有路径、代理、原因、语言和视觉输入和输出、复杂状态转换以及对抽象世界中的抽象推理的支持。他们建议同时利用符号和非符号表征。符号表示将包括用于描述动物和人工制品沿着轨迹运动的表示，以及用于描述质变词汇的状态转换的表示。非符号表征将包括靠近经验感官输入的记忆痕迹。创世纪小组希望成功不仅能带来对自然智能的更好理解，而且从长远来看，还能带来重要的实际结果，比如具有真正常识的计算机应用，以及利用对如何最好地利用人类语言和视觉解决问题能力的理解的教育应用。