The received view is that phones are getting GPS chips and location applications follow from that. Nokia put a receiver in the N95 and bought a mapping company. Everyone else is adding the hardware as fast as the bill of materials allows.
That is the visible half. The half that actually changed things has no chip in it at all.
Open Google Maps on a phone with no GPS whatsoever and press the button. It knows roughly where you are, immediately, indoors. It does this by looking at which cell towers the handset can hear and comparing them against a database of where those towers are. On a device with wifi it does the same trick with the names of nearby networks, which is more precise because a wifi access point covers a much smaller area than a mast.
The result is accurate to a few hundred metres rather than a few metres, and it is the better product for almost everything anyone actually does.
A GPS receiver, in a phone, in a city, is a frustrating object. It needs a clear view of enough satellites, which a street with tall buildings on both sides does not provide. From cold it can take thirty or forty seconds to work out where it is, and that is the good case. It does not work inside, which is where people are most of the time. And it costs meaningful battery to keep running.
Set against that, a technique that answers in under a second, works in a basement, costs almost no power, and is wrong by two streets. For finding a restaurant, checking which way is home, or filtering a search to your city, being wrong by two streets does not matter at all.
Precision was never the constraint. Latency and indoor coverage were, and nobody was measuring those because the spec sheet has a line for accuracy in metres and no line for whether it works where you are standing.
The consequence is that location stops being a feature you invoke and becomes a thing the phone simply knows. It fills your city in on a search form without asking. It sorts results by distance by default. You stop noticing the feature, which is what happens to every technology that ends up mattering.
Two things follow, and I have seen nobody discuss either.
The database is the asset. Knowing where every mast and every wifi access point in a country sits is expensive to build and gets more valuable as it grows. Whoever owns that owns something a competitor cannot simply decide to have.
And the privacy conversation has not started. When location required a GPS fix, using it was a deliberate act. When it is ambient, continuous, and free of battery cost, the question of who is keeping the log becomes a real one. Right now the terms attached to this are, as far as I can find, largely unwritten.
That will hold until the first time somebody produces a map of where a person has been.