Author here: You're right that it's a big claim, and I think it requires a lot of imagination & even some suspension of disbelief. It's safer to be skeptical.
But specifically re: competition of all shapes & sizes, I do think that two of Google's long-term investments give it a pretty strong position:
1. Google Maps & Street View, which together comprise an amazingly detailed model of the real world. There are alternatives like OpenStreetMap of course, but there's a lot more going on in those products than just navigation.
2. The technology embedded in the self-driving car. I mean, a system that can interpret a live feed of the chaotic world around it and make important decisions based on what it sees? It doesn't feel to me like you can Instagram your way into that capability. (But, I don't know: maybe it's more commodified than I realize.)
To lend some credence to these claims: other than the well-known Sebastian Thrun, Google also has at least a dozen world-class computer vision researchers working on a variety of different products, from Maps to the self-driving car to Goggles and others. This is some serious fire power, and in so far as the technical problems are unsolved (which they mostly are right now), no other company can touch them in the near future.
Of course, there is always the important step of making something that consumers will use, but that's only possible if the technical problems are solved first. And perhaps Google is learning how to make more consumer-friendly things...
But specifically re: competition of all shapes & sizes, I do think that two of Google's long-term investments give it a pretty strong position:
1. Google Maps & Street View, which together comprise an amazingly detailed model of the real world. There are alternatives like OpenStreetMap of course, but there's a lot more going on in those products than just navigation.
2. The technology embedded in the self-driving car. I mean, a system that can interpret a live feed of the chaotic world around it and make important decisions based on what it sees? It doesn't feel to me like you can Instagram your way into that capability. (But, I don't know: maybe it's more commodified than I realize.)