On Mon, 26 Sep 2016 12:09:23 +0100, David Cantrell
wrote:

On Fri, Sep 23, 2016 at 11:01:31AM +0100, tim... wrote:

but they can afford one city as a trial on the basis of their current
funding

but scaling it up to 10,000 cities just isn't going to be cheap, and I defy
them to find the funding for such.

If they can show that it works in one place, and makes money, then I'm
sure that the funding will be available. Not enough to go straight to
10,000 cities, but to roll it out to another 10 and do a larger trial.
And then to expand that, and to expand that, and so on, until all 10,000
are covered.

You could have levelled the same criticism against bold plans 130 years
ago to do ridiculous things like connect every single house in the
country to the electricity supply.

And I don't know where Tim got the idea that Uber is aiming to
instantly roll out self-driving cars in 10,000 cities. It's currently
only active in about 500, and I suspect it only plans to generate
detailed maps of a small subset of these.

Perhaps it will one day find 10,000 cities to dominate, but that must
be many decades away, if ever, particularly now it's pulled out of
China.