Alternate title: Self-Driving Cars vs. The World. That's a long and
pretty thorough article in the latest (January/February 2024) hardcopy
issue of Popular Mechanics magazine.

The author is pretty skeptical about the safety promises of the various
levels of electronic driver assistance. I think that skepticism is
justified.

But for this group, it's interesting that the article focused repeatedly
on the various systems' confusion regarding bicyclists on the road, as
well as pedestrians. Roughly half the article's photos showed cyclists
on the road. And there were comments by an official of the League of
American Bicyclists.

--
- Frank Krygowski

Frank Krygowski <frkrygow@sbcglobal.net> wrote:
> Alternate title: Self-Driving Cars vs. The World. That's a long and
> pretty thorough article in the latest (January/February 2024) hardcopy
> issue of Popular Mechanics magazine.
>
> The author is pretty skeptical about the safety promises of the various
> levels of electronic driver assistance. I think that skepticism is
> justified.
>
> But for this group, it's interesting that the article focused repeatedly
> on the various systems' confusion regarding bicyclists on the road, as
> well as pedestrians. Roughly half the article's photos showed cyclists
> on the road. And there were comments by an official of the League of
> American Bicyclists.
>
>

Seems to be that way, in that it’s been “soon” for what a decade or so by
now.

While parts of the technology are being used ie lane assist and so on on
roads such as Motorways which are fairly homogeneous ie vehicle all
travelling broadly similar speeds and so on.

Which is a very different thing to roads with the variety of folks using it
and crossing it.

Roger Merriman

On Sunday, December 17, 2023 at 4:02:37 AM UTC-8, Roger Merriman wrote:
> Frank Krygowski <frkr...@sbcglobal.net> wrote:
> > Alternate title: Self-Driving Cars vs. The World. That's a long and
> > pretty thorough article in the latest (January/February 2024) hardcopy
> > issue of Popular Mechanics magazine.
> >
> > The author is pretty skeptical about the safety promises of the various
> > levels of electronic driver assistance. I think that skepticism is
> > justified.
> >
> > But for this group, it's interesting that the article focused repeatedly
> > on the various systems' confusion regarding bicyclists on the road, as
> > well as pedestrians. Roughly half the article's photos showed cyclists
> > on the road. And there were comments by an official of the League of
> > American Bicyclists.
> >
> >
> Seems to be that way, in that it’s been “soon” for what a decade or so by
> now.
>
> While parts of the technology are being used ie lane assist and so on on
> roads such as Motorways which are fairly homogeneous ie vehicle all
> travelling broadly similar speeds and so on.
>
> Which is a very different thing to roads with the variety of folks using it
> and crossing it.
>
> Roger Merriman
Self driving cars using AI can ONLY perform correctly if ALL cars and trucks use it. Anyone driving on the freeway these days can see that people are too stupid and careless for ANY AI to predict.

Tom Kunich <cyclintom@gmail.com> wrote:
> On Sunday, December 17, 2023 at 4:02:37 AM UTC-8, Roger Merriman wrote:
>> Frank Krygowski <frkr...@sbcglobal.net> wrote:
>>> Alternate title: Self-Driving Cars vs. The World. That's a long and
>>> pretty thorough article in the latest (January/February 2024) hardcopy
>>> issue of Popular Mechanics magazine.
>>>
>>> The author is pretty skeptical about the safety promises of the various
>>> levels of electronic driver assistance. I think that skepticism is
>>> justified.
>>>
>>> But for this group, it's interesting that the article focused repeatedly
>>> on the various systems' confusion regarding bicyclists on the road, as
>>> well as pedestrians. Roughly half the article's photos showed cyclists
>>> on the road. And there were comments by an official of the League of
>>> American Bicyclists.
>>>
>>>
>> Seems to be that way, in that it’s been “soon” for what a decade or so by
>> now.
>>
>> While parts of the technology are being used ie lane assist and so on on
>> roads such as Motorways which are fairly homogeneous ie vehicle all
>> travelling broadly similar speeds and so on.
>>
>> Which is a very different thing to roads with the variety of folks using it
>> and crossing it.
>>
>> Roger Merriman
> Self driving cars using AI can ONLY perform correctly if ALL cars and
> trucks use it. Anyone driving on the freeway these days can see that
> people are too stupid and careless for ANY AI to predict.
>
It’s advantages there ie auto brake and so on is reaction speed which is
comfortably faster than human, the problem is it making the right choice
this said various smart cruse control has been on various cars for good few
years and seems not to be issue.

It’s about only place that it’s being used commonly if at a less advanced
point.

Roger Merriman

On Sunday, December 17, 2023 at 10:43:31 AM UTC-8, Roger Merriman wrote:
> Tom Kunich <cycl...@gmail.com> wrote:
> > On Sunday, December 17, 2023 at 4:02:37 AM UTC-8, Roger Merriman wrote:
> >> Frank Krygowski <frkr...@sbcglobal.net> wrote:
> >>> Alternate title: Self-Driving Cars vs. The World. That's a long and
> >>> pretty thorough article in the latest (January/February 2024) hardcopy
> >>> issue of Popular Mechanics magazine.
> >>>
> >>> The author is pretty skeptical about the safety promises of the various
> >>> levels of electronic driver assistance. I think that skepticism is
> >>> justified.
> >>>
> >>> But for this group, it's interesting that the article focused repeatedly
> >>> on the various systems' confusion regarding bicyclists on the road, as
> >>> well as pedestrians. Roughly half the article's photos showed cyclists
> >>> on the road. And there were comments by an official of the League of
> >>> American Bicyclists.
> >>>
> >>>
> >> Seems to be that way, in that it’s been “soon” for what a decade or so by
> >> now.
> >>
> >> While parts of the technology are being used ie lane assist and so on on
> >> roads such as Motorways which are fairly homogeneous ie vehicle all
> >> travelling broadly similar speeds and so on.
> >>
> >> Which is a very different thing to roads with the variety of folks using it
> >> and crossing it.
> >>
> >> Roger Merriman
> > Self driving cars using AI can ONLY perform correctly if ALL cars and
> > trucks use it. Anyone driving on the freeway these days can see that
> > people are too stupid and careless for ANY AI to predict.
> >
> It’s advantages there ie auto brake and so on is reaction speed which is
> comfortably faster than human, the problem is it making the right choice
> this said various smart cruse control has been on various cars for good few
> years and seems not to be issue.
>
> It’s about only place that it’s being used commonly if at a less advanced
> point.
>
> Roger Merriman

The Tesla self driving feature actually is a problem. No other companies advertise a fully self driving technology though they are being treated so by their owners. Like I say - this is ALL about software with is my strong suit and you cannot make an all encompassing system that can see far enough around you and take into account what really stupid things people will do. https://www.caranddriver.com/news/a44185487/report-tesla-autopilot-crashes-since-2019/

Yesterday on a 30 mile driving in twilight leading to dark I watched everything from motorcycles passing dangerously to while I was driving behind a guy who was driving 10 car lengths behind a car that was riding on the bumper of a AAA wrecker, slow up for no discernable reason. Since we were going 40, I was driving 4 or a bit more behind him so it wasn't as if he was getting irritated that I was tailgating him. After he did this for the 30th time, I moved to the open lane on the right and sped up to 40 mph at which time this guy accelerated and acted like I was going to pull in front of him even though my lane was wide open. He drove straight up on the bumper of the car in front of him who was tailgating the wrecker. I had to pace him so that he didn't pull in front of me again. AFTER I passed the wrecker and he could pull over if he wanted he continued from that point a car length or less behind the car tailgating the wrecker. Since there was NO reasoning behind any of this but some sort of animal reaction possibly to challenge which wasn't there, there was no reasons for his actions and a self driving feature would be reduced to thinking it had superior reaction time to a human driver which is impossible because it is reacting and not acting.

In my entire life I've had three auto accidents and all three were when I was having seizures due to the concussion. I am a very good and careful driver. I watched actions yesterday that used to bring cops from miles around. Now they are too commonplace for the Highway Patrol to worry about. Today, cars in the high speed lane doing 15 or 20 mph over the posted speed limit don't even slow down when they see the CHP at the side of the road.

So do not believe that you can get a "self driving" feature to work as long as there is one single human driven car or truck. Though I will say that long distance truckers are generally very careful drivers.

On Sunday, December 17, 2023 at 11:32:27 AM UTC-8, Tom Kunich wrote:
> On Sunday, December 17, 2023 at 10:43:31 AM UTC-8, Roger Merriman wrote:
> > Tom Kunich <cycl...@gmail.com> wrote:
> > > On Sunday, December 17, 2023 at 4:02:37 AM UTC-8, Roger Merriman wrote:
> > >> Frank Krygowski <frkr...@sbcglobal.net> wrote:
> > >>> Alternate title: Self-Driving Cars vs. The World. That's a long and
> > >>> pretty thorough article in the latest (January/February 2024) hardcopy
> > >>> issue of Popular Mechanics magazine.
> > >>>
> > >>> The author is pretty skeptical about the safety promises of the various
> > >>> levels of electronic driver assistance. I think that skepticism is
> > >>> justified.
> > >>>
> > >>> But for this group, it's interesting that the article focused repeatedly
> > >>> on the various systems' confusion regarding bicyclists on the road, as
> > >>> well as pedestrians. Roughly half the article's photos showed cyclists
> > >>> on the road. And there were comments by an official of the League of
> > >>> American Bicyclists.
> > >>>
> > >>>
> > >> Seems to be that way, in that it’s been “soon” for what a decade or so by
> > >> now.
> > >>
> > >> While parts of the technology are being used ie lane assist and so on on
> > >> roads such as Motorways which are fairly homogeneous ie vehicle all
> > >> travelling broadly similar speeds and so on.
> > >>
> > >> Which is a very different thing to roads with the variety of folks using it
> > >> and crossing it.
> > >>
> > >> Roger Merriman
> > > Self driving cars using AI can ONLY perform correctly if ALL cars and
> > > trucks use it. Anyone driving on the freeway these days can see that
> > > people are too stupid and careless for ANY AI to predict.
> > >
> > It’s advantages there ie auto brake and so on is reaction speed which is
> > comfortably faster than human, the problem is it making the right choice
> > this said various smart cruse control has been on various cars for good few
> > years and seems not to be issue.
> >
> > It’s about only place that it’s being used commonly if at a less advanced
> > point.
> >
> > Roger Merriman
> The Tesla self driving feature actually is a problem. No other companies advertise a fully self driving technology though they are being treated so by their owners. Like I say - this is ALL about software with is my strong suit and you cannot make an all encompassing system that can see far enough around you and take into account what really stupid things people will do. https://www.caranddriver.com/news/a44185487/report-tesla-autopilot-crashes-since-2019/
>
> Yesterday on a 30 mile driving in twilight leading to dark I watched everything from motorcycles passing dangerously to while I was driving behind a guy who was driving 10 car lengths behind a car that was riding on the bumper of a AAA wrecker, slow up for no discernable reason. Since we were going 40, I was driving 4 or a bit more behind him so it wasn't as if he was getting irritated that I was tailgating him. After he did this for the 30th time, I moved to the open lane on the right and sped up to 40 mph at which time this guy accelerated and acted like I was going to pull in front of him even though my lane was wide open. He drove straight up on the bumper of the car in front of him who was tailgating the wrecker. I had to pace him so that he didn't pull in front of me again. AFTER I passed the wrecker and he could pull over if he wanted he continued from that point a car length or less behind the car tailgating the wrecker. Since there was NO reasoning behind any of this but some sort of animal reaction possibly to challenge which wasn't there, there was no reasons for his actions and a self driving feature would be reduced to thinking it had superior reaction time to a human driver which is impossible because it is reacting and not acting.
>
> In my entire life I've had three auto accidents and all three were when I was having seizures due to the concussion. I am a very good and careful driver. I watched actions yesterday that used to bring cops from miles around.. Now they are too commonplace for the Highway Patrol to worry about. Today, cars in the high speed lane doing 15 or 20 mph over the posted speed limit don't even slow down when they see the CHP at the side of the road.
>
> So do not believe that you can get a "self driving" feature to work as long as there is one single human driven car or truck. Though I will say that long distance truckers are generally very careful drivers.

I should add that accidents are RARE because it can take a life and so most people are not only careful for themselves but for others as well. But yet there are some 4% of the vehicles on the road EACH year involved in traffic accidents and the statistics point to the same people getting into accidents each year,