Why generative AI can make creative destruction more creative but less destructive
研究了机器学习应用于运营数据如何提高创业壁垒，使创造性破坏过程破坏性减弱（业务被抢走更少），但促使创业者承担更多风险、更具创造性，并讨论了数据开放与支持创业者获取机器学习技术的政策权衡。
Abstract The application of machine learning (ML) to operational data is becoming increasingly important with the rapid development of artificial intelligence (AI). We propose a model where incumbents have an initial advantage in ML technology and access to (historical) operational data. We show that the increased application of ML for operational data raises entrepreneurial barriers that make the creative destruction process less destructive (less business stealing) if entrepreneurs have only limited access to the incumbent’s data. However, this situation induces entrepreneurs to take on more risk and to be more creative. Policies making data generally available may therefore be suboptimal. A complementary policy is one that supports entrepreneurs’ access to ML, such as open source initiatives, since doing so would stimulate creative entrepreneurship.