I understand where he's coming from as an investor, but it's kind of a shame that there doesn't seem to be a mechanism by which these sorts of features can be turned into reality.
Companies can't really accept unsolicited ideas from the general public due to the IP can-of-worms that that opens, and it's a huge amount of mostly wasted effort to build a company that's essentially a clone of another firm but with this one extra feature.
Unfortunately you can't un-know an idea once someone tells you it and you decide it's not worth what they wanted for it. And nine times out of ten it's probably not worth what they wanted for it in the first place. So maybe the problem is inherent.
One probably-unworkable solution would be a sort of IP escrow company, where, say, Facebook tells them every idea they've ever had, and when an unsolicited idea comes in, they compare it against the list and advise Facebook whether it's worth paying for. But conflicts of interest in that sort of setup are probably unavoidable.
This is certainly true, especially when you're only trying to create a feature of an existing product. There are however, recent examples of features from other products turned into successful services on their own (FriendFeed is one example that comes to mind).
The flip side is that there is very little new under the sun, even in technology, so successful products are more often than not the result of innovation as opposed to invention. I know Paul Graham has written about this, and it has certainly worked for companies like Google.
One important feature of new tech products is the whole product concept from Crossing the Chasm. In our field, where things are changing on almost a daily basis, it's hard to determine what the whole product may be but companies with this mindset and approach are much more likely to succeed than others.
Having written an improvement to finding text areas in bitmap images for OCR software, and not selling a single copy, or even licensing it to whatever OCR monopoly there is, this rings somewhat true.
Actually, the same goes for recorded music. You can copy the Aphex Twin's sound, but he's already done it. Even if he depends on a brand of synth to sound like that, that means you can't use the same synth without sounding the same. It's a no-win situation.
Google didn't win by having a better particular feature, it won by being better at the core competency of its competitors: delivering relevant search results with a minimum of spam.
True. Though it seems like there's a fine line between "being better at your competitor's core competency", which is a great business plan, and "being just like X except with an extra Y", which isn't.
Companies can't really accept unsolicited ideas from the general public due to the IP can-of-worms that that opens, and it's a huge amount of mostly wasted effort to build a company that's essentially a clone of another firm but with this one extra feature.
Unfortunately you can't un-know an idea once someone tells you it and you decide it's not worth what they wanted for it. And nine times out of ten it's probably not worth what they wanted for it in the first place. So maybe the problem is inherent.
One probably-unworkable solution would be a sort of IP escrow company, where, say, Facebook tells them every idea they've ever had, and when an unsolicited idea comes in, they compare it against the list and advise Facebook whether it's worth paying for. But conflicts of interest in that sort of setup are probably unavoidable.