本文探讨人工智慧对市场运作的影响，而非仅聚焦于生产力提升。作者指出，AI本质上是一种资讯技术，能大幅降低获取、处理与产出资讯的成本。从资讯经济学的角度来看，AI或许能解决如阿克洛夫「柠檬市场」般的资讯不对称问题，让买方借助大型语言模型更准确地判断二手车品质，从而促成互利交易。

然而，AI也可能破坏市场中既有的「信号」机制。例如，大学学位长期作为求职者能力的信号，但若AI使每位学生的表现趋于一致，学历作为能力指标的效力将大打折扣。这将增加劳动市场中雇主与求职者配对的难度，反而可能降低整体经济效率，除非社会能找到新的有效信号来取代失效的旧信号。

此外，AI加剧了「搜寻成本」问题。当每个人都能利用AI撰写完美的求职信或交友档案时，个人的理性行为却造成集体的负面外部性——所有人都必须投入更多精力去筛选与辨别。作者总结道，AI虽然让资讯变得廉价，但「资讯过载」本身也带来效率损失，能更快完成任务并不代表市场整体互动更具生产力。

This article examines how AI may reshape market functioning beyond simple productivity gains. As fundamentally an information technology, AI dramatically lowers the cost of acquiring and processing information. While this could resolve classic information asymmetry problems—such as Akerlof's 'market for lemons,' where buyers armed with LLMs might accurately assess used car quality and enable mutually beneficial trades—the picture is not entirely positive.

AI threatens to undermine established signalling mechanisms that markets rely on. University degrees, for instance, serve as signals of worker quality, but if AI equalises student performance or access to selective institutions, these signals lose their differentiating power. The resulting difficulty in matching the right workers to the right jobs could reduce labour market efficiency, representing a net productivity loss unless new, effective signals emerge to replace the degraded ones.

The article also highlights how AI intensifies costly search problems. When everyone can produce flawless job applications or dating profiles using AI, each individual's rational choice creates a negative externality: everyone else must expend more effort to filter noise and stand out. The author concludes that while AI makes information cheaper, 'too much information' carries its own efficiency costs—reminiscent of propaganda strategies that flood the zone with plausible claims. Executing tasks more productively does not guarantee that overall market interactions become more productive.