Issues about Search are generally and wrongly presented as technological or computer challenges when in fact they are best viewed as *advertising* challenges. Ultimately the search winner will be the advertising winner (Now that winner is Google with Yahoo, MSN, and ASK working hard to catch up). I'm suspicious that innovation is now driven more by advertising than by "quality search" considerations. Certainly innovation is now mostly *funded* by advertising and bets placed on the quest for ad dollars.
I suggested in an email exchange recently:
…. a "perfect" search engine set up like Google would make much less *directly* from ads since it would always deliver a perfect organic (ad free) result. I suppose in some cases there would ALSO be a perfect ad match, but there is an interesting natural tension between profit, search quality, and market share.
Tom observed in response:
…let's assume a perfect search result is one where each search result has a bit of information that's of interest to the searcher; and since it's perfect, the search engine has gone over the results and found superficially similar results that don't contribute new information content, ranked the results according to useful information content, and generally done a perfect job. I think there's still room for product promotion in there, especially if I'm looking for a product, which I increasingly do on the 'Net.
I agree if we ssume as below that perfect search still does not really match us exactly to our query. But I'm more optimistic about search and think that when combined with personal histories and other inputs like query refinements, it'll come close to reading our query intention with extreme accuracy. People would still BUY stuff as a result of search but it would be hard to use the existing models for advertising which associate only those willing to pay with the relevant results lists.