Strategic Computing: DARPA and the Quest for Machine Intelligence, 1983-1993 (review)
Strategic Computing: DARPA and the Quest for Machine Intelligence, 1983-1993 (review)
复制标题
战略计算：DARPA 和机器智能的探索，1983-1993（评论）
DOI：
--
复制
发表时间：
2003
期刊：
影响因子：
--
通讯作者：
R. Seidel
中科院分区：
文献类型：
--
作者：
R. Seidel
沃伦·韦弗在D-2和第七师召集了来自战前所有地点的专家，并特别关注火力控制和“防空问题”。到战争结束时，韦弗和他的同事们已经将当地的知识和实践结合起来，设计了新一代的自动化雷达制导火炮指挥仪，并对人机系统中的计算、反馈、预测和控制等广义问题有了新的认识。当维纳声称一个新的世界观和学科在1948年出版的控制论，他创造了它的知识遗产，完全抹去了几十年前的努力，忽略了，事实上，他自己的战时贡献的姿态，明德尔建议，这是一个心理后果维纳的深刻的个人拒绝军事研究。明德尔进一步指出，乔治·菲尔布里克的“超级模拟器”可能更准确地反映了根植于早期工程实践中的赛博格和控制论思想。在国家发改委1946年的技术报告摘要中，菲尔布里克的设备概括了战时控制系统：“随着战争变得越来越机械化和自动化，菲尔布里克写道，超级模拟器将能够模拟整个战斗：因此，[制导]导弹之间的战争可以完全在实验室中进行和观察，可能有统计机构参与解释和评估。多年的经验，以及在控制和动态组件开发方面的试验和错误，可以被压缩成小时“（Mindell，第308页）。《人与机器之间》是一部典型的历史。一方面，明德尔以对细节和背景的精致关注勾勒出当地的技术文化;另一方面，他从未忽视更大的叙事或他的故事的史学相关性。他令人印象深刻地成功地充实了我们的电子人过去，恢复了标准账户中缺失的历史。
Warren Weaver assembled in D-2—and then Division 7—experts from all of the prewar sites and made fire control and the “anti-aircraft problem” his special concern. By the war’s end, Weaver and his colleagues had integrated local knowledge and practice, engineering both a new generation of automated, radar-guided gun directors and a new appreciation for the generalized problems of computation, feedback, prediction, and control in manmachine systems. When Wiener claimed a new worldview and discipline in 1948 with the publication of Cybernetics, he created for it an intellectual heritage that entirely erased the efforts of earlier decades and ignored, in fact, his own wartime contributions—a posture, Mindell suggests, that was a psychological consequence of Wiener’s profound personal rejection of military research. A more accurate reflection of cyborgian and cybernetic ideas rooted in earlier engineering practices might be, Mindell further suggests, George Philbrick’s “supersimulator.” Described in one of the 1946 volumes contributed to NDRC’s summary technical report, Philbrick’s device generalized wartime control systems: “As war became increasingly mechanical and automated, Philbrick wrote, the supersimulator would be capable of modeling entire battles: ‘Warfare among [guided] missiles could thus be staged and observed entirely in the laboratory, possibly with statistical machinery in attendance for interpretation and assessment. Years of experience, and of trial and error, on the development of controls and dynamic components could thus be collapsed into hours’” (Mindell, p. 308). Between Human and Machine is a model history. On the one hand, Mindell sketches local technical cultures with exquisite attention to detail and context; on the other, he never loses sight of the larger narrative or the historiographical relevance of his story. He has succeeded impressively in fleshing out our cyborgian past and restoring history missing from standard accounts.