Uber is already hamstrung by their inexperienced drivers's reliance on Google Maps for navigation. It is in no way equivalent to actually knowing your way around.
The difficulty of making an urban self-driving car aside, Google would have to achieve a quantum leap forward in the quality of their navigation platform. Otherwise every auto-taxi in San Francisco will proceed single file down Van Ness, with turns onto Market or Mission only. All other streets will be empty and filled with pigeons.
It's not just taking traffic into account, which it can (sort of) already do.
They need to actually use the accident data. Use a diverse set of routes to get to the same place, so that there isn't one path for every car on the road. Understand stoplight patterns and where it's hard or easy to make a turn. Not focus obsessively on shortest path rather than most tolerable path.
Fundamentally what a good cab driver (which are not that common) offers you is knowledge of the best route to a destination from experience. Google Maps doesn't have that. And if it did, and it controlled a significant portion of the cars, it would no longer be the best route.
You know that you're describing things that machines are inherently better at than humans? And the problems you described stem mostly from the fact that maps also serve informative function (to learn the route in advance), and apparently Google didn't decide yet to compute routes in their navapp based on other active users of said navapp in the neighbourhood. But the data, infrastructure and algorithms are there.
> And if it did, and it controlled a significant portion of the cars, it would no longer be the best route.
Of course it would, because it would be able to run a global route optimization on all the cars simultaneously, thus outperforming even the most experienced human drivers by orders of magnitude. Getting above human level seems like a college-level exercise.
> inexperienced driver's reliance on Google Maps for navigation
When I lived in Sydney, I used to take the taxi often. And the rule was pretty much: If your address doesn't exist on Google Maps, they don't know how to get there. Even "At the corner of Hyde Park and Oxford St", which is in the CBD, returned a 404 from the driver's vocal API. They were all officially registered drivers, I just think cab's over reliance on Google Maps makes them unaware of the street names.
> Uber is already hamstrung by their inexperienced drivers's reliance on Google Maps for navigation. It is in no way equivalent to actually knowing your way around.
I don't know about you, but my experience is that Google Maps is far more reliable than a cabbie who purports to know their way around.
My experience is the exact opposite in most cities in the UK - Google Maps regularly gets me going insanely stupid routes, while the cabbies always seem to know every street and the best way to get to it.
Perhaps the US has a different culture for its cab drivers? I can't imagine why, though. We have all the taxi licensing schemes and whatnot that the US does, so it can't be a case of "there's more competition so they're better". Perhaps it literally does just come down to culture?
Not sure how it works elsewhere in the UK, but in London the cab drivers have to take an in-depth course called 'The Knowledge' which covers street routes, alternate routes during street closures, commercial destinations, etc. See some sample questions : http://www.taxiknowledge.co.uk/mock.html
This is majorly helpful beyond usual satnav use case. Eg, to get dropped off during high-traffic area, the cabbies will know well the surrounding streets and at your request can drop you off in a reasonable place 1-2 blocks away, for much less time and expense.
In the US, cabbies are often relatively recent immigrants. They may know parts of the city OK, but they can't be relied upon to have the city memorized.
There's licensing, but it's a far cry from the Knowledge of London.
I guess we'll find out after Uber has exhausted its VC money, if Google doesn't replace them with self-driving cars first.