Monopoly search is bad because it is easier to game. People have gotten so good at manipulating Google rankings, that there is room for competition. Merely being small has some inherent advantages. Being bigger and more popular can be inherently bad.
Yeah, I think of it as the Google uncertainty principle (after Heisenberg). As soon as Google measures a search signal, just the fact that it goes into the rankings means that it degrades in quality. People start reverse engineering the ranking and trying to game it. So it's kind of an arms race between Google and the text ad industry they've created.
He starts at 8 mins in, and although the whole video is 2 hours long, he makes a bunch of major points in the first 3 to 4 minutes of his talk, after which he answers questions for 3 to 5 minutes.
His central point is that there is a lot of potential value in getting the fixed costs of the search business down, and that he would prefer to invest there, rather than in playing the zero sum game of trying to take away make share in the $25 Bln search business.
Also worth noting: anybody who can lower the fixed costs of search by an order of magnitude is also going to be in a position to make money from businesses that have similar cost structures.
Yes and no. Pretty basic spelling correction is pretty good, and as Google makes its spelling operation more sophisticated, I actually find myself needing to reverse their corrections more often.
Indexing is more complicated. You could argue that 20% of Google's index contains at least 80% of the information people need. The problem, like the old advertising saying, is figuring out which 20%. So if you have a clever way to address content quality and uniqueness, suddenly your crawling and indexing costs plummet.
While it may be argued that `the obscure stuff' is harder to find or may be more valuable/important, I'd disagree with you.
First, the bulk of the search seems to be for the usual stuff. Thus the bulk of revenue from ads for the provider and _perhaps_ the bulk of added value for the users stems from searching for the usual stuff.
Second, as the obscure stuff oft uses uncommon terminology and names, it's technically easier to search for it and display relevant results. The trick is to give useful results for the most common things. Consider how hard it may be to automate (!) picking sensible results for words like `the' or `free'. Or for dates that are represented in one zillions of formats. Of course various approaches (like lists of & criteria for stopwords) have been implemented.
The idea that costs for a search engine go down with time is false.
Bandwidth, CPU, memory and disk sizes you have available to index and retrieve information do get cheaper. And this means they get cheaper for information-producers too. So the need to increase the amount of data you index eats away the decreasing cost of hardware.
Naturally this holds true for general search engines. Vertical ones are a very different beast. And I think Peter Thiel was not talking about specialized vertical searches like patent search.
It depends on whether you think you need to index the same fraction of web content over time. I think that's one of the unexamined principles of the search engine industry. Also, software costs are in some ways decreasing even faster than hardware costs because of open source infrastructure.
To go back to the Blekko example, I'd say it's pretty clear you couldn't build something like that with $10-15 MM (assuming they've gone through maybe half their funding) a few years ago.
It seems like you really need to keep up the index on the same %, but I guess dropping all the "designed for Ads only" sites would drop your number. I keep wondering if one of the shortening companies would come up with a search engine based on their links and some metric on repeats / reputation. Also, it might be work trying some touch setup with a person going good/bad for some % of your index.