Re:"I can tell you wrote this yourself"
by FinnLobsien on Friday November 20, 2026 @05:59AM
(#50572985)
Yes, I totally agree. But yeah, with physical products, there's always a utilitarian function of "does it accomplish its physical purpose", e.g. keep feet warm and protected.
To what degree, for how long, under what circumstances, and with what externalities that is accomplished is a separate discussion.
But in knowledge work, the output's purpose is often merely to represent a cognitive process, in which case sidestepping the process negates the output's value.
Think of an architecture document. The entire point is to obviate downstream questions and discussions because the thinking process has already been done and agreed upon from multiple sides. If we AI-generate an architecture document, the humans evaluating downstream work output may still end up arguing because the cognitive process the document represents never happened, in which case the document was pointless from the start.
To what degree, for how long, under what circumstances, and with what externalities that is accomplished is a separate discussion.
But in knowledge work, the output's purpose is often merely to represent a cognitive process, in which case sidestepping the process negates the output's value.
Think of an architecture document. The entire point is to obviate downstream questions and discussions because the thinking process has already been done and agreed upon from multiple sides. If we AI-generate an architecture document, the humans evaluating downstream work output may still end up arguing because the cognitive process the document represents never happened, in which case the document was pointless from the start.