为何大型语言模型在软件开发中可能是死胡同
原文英文,约1200词,阅读约需5分钟。发表于: 。While the title Does current AI represent a dead end? is clearly made to encourage debate, there is a case The post Why LLMs Within Software Development May Be a Dead End appeared first on The New Stack.
当前人工智能系统缺乏可分解性和可解释性,软件开发者应关注可测试组件。大型语言模型(LLMs)在产品中存在问题,无法有效分离操作与训练数据,导致安全和法律风险。开发者应设计可持续平台,以确保人工智能的可解释性和可监控性。