Well for one thing, the term "idea" isn't very well defined. If you are talking about something like Elon Musks Hyperloop, I can see that it might be worth money. But someone else might consider "Uber for dog walkers" an idea. And it is to some extent, but how would you know before you buy? No worries, you say, you'll just make a clause that says that you only pay for an idea if you execute it? Right, but then you are looking to buy an idea without knowing anything about it, and if you hear it but don't like it, you have now exposed yourself to the anger of the idea-guy if you ever go on to do something related.
I have once upset an acquaintance because he felt that I stole his idea. After that incident, I have actually started to ask people not to tell me their ideas unless they are cool with me possibly using them some day. Otherwise, I just risk making enemies with very little upside.
I think it's possible to sufficiently reduce this problem. For example, "needs to be an idea I can execute on with skill set X (or hire for)", "needs to be something I can execute on with $Y or includes a plan to acquire $Y that I can execute on", "needs to have a potential ROI of Z%".
Also, enough information can be given in advance that I'd be able to filter out ideas that I'm not interested in pursuing. General industry it's in, whether or not it's legal, etc.
I have plenty of information/ideas that would be valuable to software engineers, so I know that there exist scenarios in which information/ideas are valuable. I would've benefitted from the same information/ideas a few years back, and I suspect there's more ideas I don't have yet that would provide more gains. However, as I mentioned in another post, not many people willing to pay for information/ideas so I've had difficulty finding buyers.
I have once upset an acquaintance because he felt that I stole his idea. After that incident, I have actually started to ask people not to tell me their ideas unless they are cool with me possibly using them some day. Otherwise, I just risk making enemies with very little upside.