"whisky-dave" wrote in message
...
On Sunday, 30 September 2018 17:04:57 UTC+1, Jimmy Wilkinson Knife wrote:
On Mon, 10 Sep 2018 11:17:16 +0100, whisky-dave
wrote:

On Friday, 7 September 2018 23:00:05 UTC+1, Jimmy Wilkinson Knife
wrote:
On Fri, 07 Sep 2018 10:44:19 +0100, whisky-dave
wrote:

On Friday, 7 September 2018 00:26:18 UTC+1, Jimmy Wilkinson Knife
wrote:
On Thu, 06 Sep 2018 10:04:00 +0100, whisky-dave
wrote:

On Wednesday, 5 September 2018 20:53:23 UTC+1, bert wrote:
In article
,
whisky-dave writes
On Tuesday, 4 September 2018 15:08:23 UTC+1, Jimmy Wilkinson
Knife wrote:
On Tue, 04 Sep 2018 10:45:01 +0100, whisky-dave
wrote:

On Monday, 3 September 2018 16:52:41 UTC+1, Jimmy Wilkinson
Knife wrote:
On Mon, 03 Sep 2018 16:06:21 +0100, whisky-dave
wrote:

On Monday, 3 September 2018 14:50:32 UTC+1, Jimmy
Wilkinson
Knife wrote:
On Mon, 03 Sep 2018 13:30:40 +0100, whisky-dave
wrote:

On Saturday, 1 September 2018 21:59:36 UTC+1, Jimmy
Wilkinson
Knife wrote:
On Sat, 01 Sep 2018 20:54:02 +0100, Rod Speed
wrote:

Jimmy Wilkinson Knife wrote
Rod Speed wrote





http://www.abc.net.au/news/2018-09-0...sting/10190804

The report doesn't say whose fault it was.

But it does list a hell of a lot of other similar
failures
and a well designed self driving car should drive
the
car so it doesn't cause human drivers to run into
the
back of the car its driving because of how it
drives.

And of course human drivers never **** up like
this....

Which is why people don't want to spend Ł1000s on a
self
driving car that ****s up or kills them, they can have
that


No, they get 20 times less ****ups.

The California DMV said it has received it has received
95
autonomous vehicle collision reports as of August 31.
Dozens of
companies have received permits to test self-driving
vehicles on
California roads, but those permits require the presence
of a human safety driver.

Just how amny of these atomomous cars are there a few
dozen
comparded to....

You say 95 collisions with autonomous vehicles, but you
don't say
how many manually driven cars have had collisions.

Because it's such a small sample, we don't narrow down the
number
of manual collisons of 100 or so cars and that is why there
are so
many more collisins with manual cars because we count the
cars in
millions NOT dozens.
What's needed is a rate of accidents a figure they will not
give out.

I've seen it, and it's 20 times larger for human driver than
automated cars.

Then yuo should be able to cite it then shouldn't you ?
Presenntly there are NO automated cars, anyway and the cars
that are
used as driverless aren't tested amonst real road users they
just go
round and around on test tracks, no wonder they don't have
accidents.

But if they are so good why do they require a human at the
wheel ?

Because they are not THAT good.

So humans are still better at driving cars than humans, that is
the point.

Are you drunk again? Rewrite that sentence.

Nothing wronfg humans are humans they are all difernt even if only
slightl;y so have one faulty human driver isnlt such a dig deal like
yourself who can't see a deer in front of them and can't stop in
time isnt such a big deal but when you have 100,000s of them and
none for them cane yuo end up with serious problems it;s why you
have to have recalls for faulty products.
With humans you remove their driving license if they can't drive
properly and it;s not big deal except for teh person who has their
license removed do that to 100,000 cars in a day and you have
problems.

Computerised cars can easily be updated with better programs, you
can't do that with humans.

Yes you can it's called training.

Not as easy as updating the software in every computer.

But that won;t work it's the semsors and hardware that will need updating.

Much more likely its what the software that uses those
sensors and hardware that needs the updating.

It's like phone Apple haven't just upgraded the software for their
iPhone3GS

Thats providing more capability. They have in fact radically changed
what the software does as well as improving the sensors and cameras.

so it's compatable with face recognition

It has in fact got a completely new way of looking at the face for that.

and everything else have they. So you wouldn't
be able to do that with cars either.

Even sillier than you usually manage, and thats saying something.

just software that might need changing, but then nyou;'d have to
understand
the technologies involved rather thsan just assume they work by magic.

of course if you can't tell a real car from a fake wooden one ?,
Could you tell the differnce between a faked wooden car and the real
thing ?

People make mistake...

The person saw the car, the co puter in the car didn't.
People are still better at driving cars than cars driving themselve,

Thats very arguable with the routine driving.

and that is why currently all driverc assisted cars come with
a steering wheel and pedals otherwise why put them in the
car just to waste money, to take up an extra seat, to make
the cars heavier and less efficient ?

Just saw an article in the BBC RSS feed which points out that
one of the real downsides with that approach of supposedly
having the human checking how the computer is driving is
that many humans will just doze off or read a book etc
when the computer is doing all the work and it will be
interesting to see how real the supervision actually is.