Google & low-hanging fruit

Google has picked all the low-hanging fruit of search. Easily spidered pages on public websites? Done. Basically anything with machine-readable text is done.

This is both good and bad. It’s good because the low-hanging fruit is the best and most useful fruit (as far as I can tell). It’s bad because this kind of search is now something of a well-understood problem. Competitors are now popping up with their own spin on search.

The rest of the fruit is more difficult. Image search is more difficult. The relative lack of image search tools compared to text search tools speaks to that difficulty. Video search is more difficult still. Right now video search mostly involves indexing the pages the videos live on. Google’s purchase of YouTube makes sense in that context. But any other kind of video search (within video for instance), is an engineering and a computational challenge. Indexing of audio faces the same challenge. Indexing the dark web is difficult. Indexing information that isn’t yet online is nigh impossible.

I’m sure you can think of a few more difficult search arenas. (Leave them in the comments!)

If it doesn’t look like anyone’s enthusiastically pursuing these search technologies, it’s probably because of diminishing returns. Problems with a high computational component tend to be future technologies. You can argue Google is a product of low bandwidth cost, low storage cost, and low computation cost with a bit on engineering prowess mixed in.

We may not be at the point where these problems make sense to solve. We may never be. Computation cost is essentially energy cost, and energy cost is not trending down. There may never be a point at which (barring of course the kind of engineering breakthroughs that make these sorts of technologies suddenly and oddly viable) it makes sense to index video itself instead of the pages it lives on.

Google faces a choice, then. It can pursue indexing all information ever, with diminishing returns at ever higher costs, or it can branch out.

I think Google is branching out. They have been for a while. They keep throwing things at the wall. A few of them are sticking. And very few of them are directly related to search. Most are orthogonal to it. Most are instead directly related to ecosystem lock-in (Gmail, Docs, etc), and userbase protection (Chrome, Android, etc).

This is what you do when the low-hanging fruit is gone, after all.