Will cars still have steering wheels in 2026?

[Brian Howell]

We had a good dinner and a rather heated discussion last night. 

David Fetter argued that within ten years, insurance companies would incentivize new car buyers to forgo steering wheels. His thesis was that technology, especially AI and sensors, is advancing rapidly enough that cars will be capable of driving themselves by 2026—and be better at it than humans. And that as soon as they are, insurance companies will start penalizing cars having steering wheels.

Alan Pearson responded that there are simply too many obstacles, both physical and figurative for that to happen. 

I agreed with David, mostly. There is a growing cache of facts, figures, and projections to support his estimate. Prototype self-driving cars are rapidly becoming safer.

Rob Allen seemed to side with Alan. So we were well matched: two on two.

I'd like to hear from other members of I-D: how long do you think it will be before most new cars are driverless?

[alanjoh...@gmail.com]

Yes a fair synopsis although I don't think that most new cars will be without steering wheels in 2026

Alan Pearson - Sent from my iPhone 510-334-8648

--
You received this message because you are subscribed to the Google Groups "Ipse Dixit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to Ipse-dixit+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/Ipse-dixit/c901bc26-e987-418c-a621-d1a50632a1b1%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Craig Good]

> On Jul 29, 2016, at 09:38 AM, Brian Howell <bdho...@gmail.com> wrote:
>
> I'd like to hear from other members of I-D: how long do you think it will be before most new cars are driverless?

Predictions are hard, especially when they’re about the future. Trying to pin a year to this prediction would be, for me at least, a WAG. Ten years is possible, but twenty wouldn’t surprise me. It does seem clear that the day is coming when letting an ape drive a car will be viewed as a very risky activity. Insurance companies will likely provide the first disincentives.




--
--Craig WWJGD?
clg...@me.com http://www.craig-good.com

"To disarm the people - that was the best and most effective way
to enslave them ...."
- George Mason ( Framer of the Declaration of Rights, Virginia,
1776, which became the basis for the U.S. Bill of Rights )

[Brian Howell]

Here is one optimistic view: http://www.techinsider.io/chris-dixon-future-of-self-driving-cars-interview-2016-6

--

Brian Howell
www.linkedin.com/in/bdhowell

[David Fetter]

On Fri, Jul 29, 2016 at 09:38:11AM -0700, Brian Howell wrote:
> We had a good dinner and a rather heated discussion last night.
>
> David Fetter argued that within ten years, insurance companies would
> incentivize new car buyers to forgo steering wheels.

Actually, what I said was slightly stronger. I don't think that new
cars will be available with steering wheels.

> His thesis was that technology, especially AI and sensors, is
> advancing rapidly enough that cars will be capable of driving
> themselves by 2026—and be better at it than humans.

That part should happen within the next five years. Self-driving cars
on actual roads are already pretty close to par with human drivers.
The fact that their crashing and killing an occupant is newsworthy
demonstrates this.

> And that as soon as they are, insurance companies will start
> penalizing cars having steering wheels.

Right. Somewhat by accident, while I was getting my degree, I
happened to take the entire math part of the curriculum required to
become an actuary. These are the people whose entire professional
lives embodied the phrase "data driven," long before it was coined.

Over the next 3-5 years, probably the lower end, actuaries at
insurance companies will be instrumental to the following:

1. Each feature that will eventually add up to cars driven without
driver intervention that costs them less in payouts will have its
financial impact measured. This measurement will quickly be
transformed into economic signals a.k.a. "discounts" on insurance,
then to penalties for not having the features.

2. Meanwhile, a longer term study will kick off, namely comparing how
automation performs in emergencies to how people do, whether it's
while driving "on their own," whatever that comes to mean, or, more
importantly, when they "take over in an emergency," which is a
euphemism for "substitute their poor judgement and bad reflexes for
the automatic ones." This will relatively quickly yield the
information that machines drive better, even in emergencies, when not
interfered with by people.

3. Somewhat off to the side, MADD and similar organizations will
demand that cars refuse to be driven by drunk people. At the moment,
this is done with expensive custom hardware that prevents the car from
starting, but that's crude and unprofitable for the auto company.
Soon, a "Clippy(R)" type software will say something along the lines
of, "I see you are unable to drive at the moment. Please relax while
I get you home," and refuses to yield control of the moving vehicle to
the driver.

> Alan Pearson responded that there are simply too many obstacles,
> both physical and figurative for that to happen.
>
> I agreed with David, mostly. There is a growing cache of facts,
> figures, and projections to support his estimate. Prototype
> self-driving cars are rapidly becoming safer.
>
> Rob Allen seemed to side with Alan. So we were well matched: two on
> two.

The facts aren't a democracy. I bet Alan $100 that new cars with
steering wheels in them will no longer be available in ten years.

> I'd like to hear from other members of I-D: how long do you think it
> will be before most new cars are driverless?

I'd be curious, too. "Ten years" is a pretty facile timeline, but it
was the best I could come up with based on what I know about the
situation.

Best,
David.
--
David Fetter <david(at)fetter(dot)org> http://fetter.org/
Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter
Skype: davidfetter XMPP: david(dot)fetter(at)gmail(dot)com

Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate

[Craig Good]

> On Jul 29, 2016, at 11:51 AM, David Fetter <da...@fetter.org> wrote:
>
> That part should happen within the next five years. Self-driving cars
> on actual roads are already pretty close to par with human drivers.
> The fact that their crashing and killing an occupant is newsworthy
> demonstrates this.

Except that there are no commercially available self-driving (autonomous) cars on the market yet. Tesla’s Autopilot is just at Level 2. We’re talking about Level 4 cars, which are still in development.

http://www.nhtsa.gov/About+NHTSA/Press+Releases/U.S.+Department+of+Transportation+Releases+Policy+on+Automated+Vehicle+Development

Also, there is no way that the Florida fatality was Autopilot’s fault.





--
--Craig WWSJD?
clg...@me.com http://www.craig-good.com
Nietzsche is pietzsche, but Sartre is smartre.

[alanjoh...@gmail.com]

I believe that Google could develop an emergency steering system using wood and ropes attached to the front wheels

Alan Pearson - Sent from my iPhone 510-334-8648

> --
> You received this message because you are subscribed to the Google Groups "Ipse Dixit" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to Ipse-dixit+...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/Ipse-dixit/52ADAE67-7695-4D67-855A-E4CAAAAACF57%40me.com.