> This is simply incorrect. Every economic output requires material inputs...There is simply no such thing as "immaterial" economic production.
Parent post was talking about marginal cost of production. E.g. where once we'd buy servers to run services on, we now share servers rented "by the cycle" from providers who amortize the cost over many users. Idle servers can be running at-the-moment economically valuable code for others. It's even possible that some economically valuable computation I might cause to be performed could consume an immeasurably small amount of power compared to the server sitting idle.
i'm partial to the labor theory of value, not for pricing (which is better modeled with marginalism, as you speak to), but for an intrinsic value by which we can evaluate how well a particular market might be working. a well-functioning market should tend toward the intrinsic (labor) value of a good (if, for example, relevant information is widely available and known, competitors can easily enter and exit, distortions are minimized, etc.).
that's the whole point though. price can be highly decoupled from cost, which i'm positing is a signal of the magnitude of the distortions in a given market.
but to your question, you'd just "measure" price as the prevailing price that you see in the marketplace, and labor is measured as the aggregate cost of all labor inputs, including parts and raw materials and amortized capital costs. in a well-functioning market, you'd expect prices to trend with the labor value. the industry profit percentage would then be an reasonable expression of the riskiness of the market/industry (in drug discovery, it might be 50% but in grocery, it might be 5%).
so for example, if apple sells iphones for $1000 but it costs them $400 in labor to make and sell, their labor-price markup would be 250%. in a functioning competitive market, you'd expect a competitor like samsung to pursue those profits and, as a result, eventually drive prices down.
Parent post was talking about marginal cost of production. E.g. where once we'd buy servers to run services on, we now share servers rented "by the cycle" from providers who amortize the cost over many users. Idle servers can be running at-the-moment economically valuable code for others. It's even possible that some economically valuable computation I might cause to be performed could consume an immeasurably small amount of power compared to the server sitting idle.
Your formulation reminds me of the "labour theory of value" (https://en.wikipedia.org/wiki/Labor_theory_of_value).