部分AI能力的调用价格已经大幅下降，使更多任务具备采用条件，也使制约应用的因素逐渐显现：可按期接入的电力、用途明确的数据、专业判断以及组织实施能力。价值分配取决于这些条件在具体任务中是否不可替代、由谁控制，以及收益与成本如何约定。产业需求、资源权利、资本投入、经营现金流和资产估值之间存在多重转换，需求增长只能支持其中一部分。本文考察这些转换如何发生，并讨论资源持有者与财富拥有者如何识别可持续的贡献与回报。

The price of using some AI capabilities has fallen sharply. More tasks now meet the conditions for adoption, and the factors that constrain application are becoming visible: electricity that can be connected on schedule, data with a clearly defined use, professional judgement and the organisational capacity to implement. How value is distributed depends on whether these conditions are irreplaceable in a given task, who controls them, and how returns and costs are agreed. Between industry demand, resource rights, capital investment, operating cash flow and asset valuation lie several conversions, and growth in demand can support only some of them. This paper examines how those conversions occur, and how resource holders and wealth owners can identify contributions and returns that last.