摘要
人工智能在建筑和城市设计领域有着较长历史。在20世纪70和80年代,人们曾对其抱有较高的期望。但事实证明,那是过于雄心勃勃的,并在一定程度上是具有挑衅意味的。人工智能应用科学的发展得到主流建筑师和城市设计师支持的案例少之又少。人工智能的初期主张引发了对于建筑师和城市设计师的核心工作和竞争力将被取代的担忧。而当期望未能兑现时,又导致建筑和城市设计领域对人工智能的兴趣和资金支持丧失。然而,自21世纪的第二个十年开始,对人工智能的兴趣又有所回升。技术的提升、对计算机科学整体态度的转变、大数据和数据分析的出现、公民学的广泛应用、公民设计的重新发现、公民设计科学的出现以及大型开放式在线设计方法,正在一同推进建筑和城市设计领域对于人工智能更为积极的展望,并可能导向令人信服的成果。文章基于案例,展示从早期的人工智能生成式方法向机器学习方法的转变。
Artificial intelligence（AI） has a long history in architecture and urban design. Expectations were high in the 1970 s and 1980 s. As it turned out, they were overambitious and to a degree confrontational.In only few cases it was possible to gain the support of leading architects and urban designers in the scientific development of AI applications. The claims of early AI raised the concern of being replaced by AI in their core activities and competencies.When the promises did not materialize, it led to the loss of interest and funding of AI in architecture and urban design. Yet interest has risen again since the beginning of the second decade of the 21 st century. Improvements in technology,changes in attitude towards computers in general, the advent of big data and data analytics, the wider reach of citizen science, the rediscovery of citizen design,the emergence of citizen design science and massive open online design approaches are promoting a more positive outlook on artificial intelligence in architecture and urban design and could lead to convincing results.The article will introduce examples showing the shift from early AI generative methods to machine learning approaches.
出处
《时代建筑》
2018年第1期32-37,共6页
Time + Architecture
