> Good ideas are rare, and nobody knows how to come up with them reliably.
The operative word there is "good ideas are rare". The claim is that you don't know whether a idea is good until it's been executed on (which seems plausible, and at any rate you haven't contradicted that), and that good ideas are rare, so most ideas are crap (which you yourself just asserted). Thus, the expected value of a given idea (absent execution) is very low, aka cheap.
I mean, I think it's obvious that the OP meant "good ideas" when he said "ideas are cheap." His entire point was "novelty" and "many really clever ideas" are not worth anything without the execution to go along with them.
> The claim is that you don't know whether a idea is good until it's been executed on
I don't think his claim was just epistemological the way you're interpreting it, and if it was, then I will withdraw since it's not the idea I meant to engage with.
> I mean, I think it's obvious that the OP meant "good ideas"
Eh, I consider "Ideas that are known for a fact to be good (ie useful) are useful (ie valuable, provided they aren't already devalued by ubiquity[0])." to be sufficiently tautological that I would not expect someone (intellegent and good-faith enough to be worth engaging with) to be arguing against it. Although note the distinction between merely good ideas, versus ideas that you actually know are good, since the work of turning the former into the latter was what I assumed they were attributing to execution.
> I don't think his claim was just epistemological the way you're interpreting it, and if it was, then I will withdraw since it's not the idea I meant to engage with.
Fair enough; I'm not sure we even have a disagreement over facts so much as over how we're categorizing things and the terminology therefor.
0: Eg, "writing" or "general-purpose computers" are almost incalculably valuable ideas, but they're not worth anything in a supply-and-demand sense because everyone in the world already knows about them.
> Good ideas are rare, and nobody knows how to come up with them reliably.
The operative word there is "good ideas are rare". The claim is that you don't know whether a idea is good until it's been executed on (which seems plausible, and at any rate you haven't contradicted that), and that good ideas are rare, so most ideas are crap (which you yourself just asserted). Thus, the expected value of a given idea (absent execution) is very low, aka cheap.