What happens when AI makes checking cheap, not just producing?
热度趋势
趋势数据积累中
百分比基于当前可用热度信号，而非评论数或独立用户人数。
人工智能大幅降低核查成本的能力，可能会通过降低“核查边界”来重塑商业格局。这种转变将减少因核查成本高昂而得以维持的“不透明租金”。除了审计之外，人工智能还可以使合规性和最优性检查变得经济实惠，从而引发重大的制度变革。审计师、认证机构和中介机构等围绕昂贵核查而发展的实体，将需要在“后不透明”世界中适应。…
Imagine a company overcharges you by €43, but proving it would cost €200. You ignore it. The money survives not because it is hidden, but because checking is too expensive.
Now imagine AI makes that check cost €2.
I call the boundary between what is worth checking and what is not the verification frontier. AI may push that frontier down by reducing both the cost of each check and the cost of building custom verification systems. The surplus that survives mainly because checking is uneconomic is what I call an opacity rent.
This goes beyond auditing. There are two questions: conformity: did the company follow the contract correctly? And optimality: even if it did, is this actually the right contract for you?
What interests me most is the institutional change this could cause. A lot of commerce developed around expensive checking: auditors sample, certifiers spread verification costs, intermediaries get paid for expertise, brands provide assurance when buyers cannot inspect quality themselves, and contracts are often not designed to be machine-verifiable. If checking becomes cheap, these institutions do not necessarily disappear, but their role and value should change.
That is what I mean by post-opacity: not perfect transparency, but a world where “nobody will bother checking” becomes a much weaker economic protection, forcing businesses and institutions to adapt.
Where do you think this applies first, and where am I overestimating the impact? Which institutions would actually resist cheap verification rather than adapt to it?
I’ve written the broader argument here: https://post-opacity.com