Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.forth > #33901 > unrolled thread

forth to drive automated fast food machines

Back to article view | Back to comp.lang.forth

Contents

  forth to drive automated fast food machines gavino_himself <visploveslisp@gmail.com> - 2014-10-14 16:47 -0700
    Re: forth to drive automated fast food machines Paul E Bennett <Paul_E.Bennett@topmail.co.uk> - 2014-10-15 22:12 +0100
      Re: forth to drive automated fast food machines foxaudioresearch@gmail.com - 2014-10-15 18:36 -0700
        Re: forth to drive automated fast food machines Jason Damisch <jasondamisch@yahoo.com> - 2014-10-16 07:52 -0700
          Re: Google to drive car (was forth to drive automated fast food machines) "Alex McDonald" <blog@rivadpm.com> - 2014-10-16 16:45 +0100
          Re: forth to drive automated fast food machines Bernd Paysan <bernd.paysan@gmx.de> - 2014-10-16 21:38 +0200
            Re: forth to drive automated fast food machines rickman <gnuarm@gmail.com> - 2014-10-16 15:49 -0400
              Re: forth to drive automated fast food machines Bernd Paysan <bernd.paysan@gmx.de> - 2014-10-17 01:26 +0200
                Re: forth to drive automated fast food machines Just Ian <superian@gmail.com> - 2014-10-16 17:09 -0700
                  Re: forth to drive automated fast food machines Bernd Paysan <bernd.paysan@gmx.de> - 2014-10-17 02:29 +0200
                Re: forth to drive automated fast food machines the_gavino_himself <visphatesjava@gmail.com> - 2014-11-08 23:32 -0800
            Re: forth to drive automated fast food machines hughaguilar96@yahoo.com - 2014-10-16 19:55 -0700
              Re: forth to drive automated fast food machines "Rod Pemberton" <buzzoff@nonamewhichexists.cmm> - 2014-10-17 22:40 -0400
            Re: forth to drive automated fast food machines the_gavino_himself <visphatesjava@gmail.com> - 2014-11-08 23:33 -0800
          Re: forth to drive automated fast food machines "Rod Pemberton" <buzzoff@nonamewhichexists.cmm> - 2014-10-17 22:35 -0400
      Re: forth to drive automated fast food machines the_gavino_himself <visphatesjava@gmail.com> - 2014-11-08 23:25 -0800
    Re: forth to drive automated fast food machines "Rod Pemberton" <buzzoff@nonamewhichexists.cmm> - 2014-10-17 22:34 -0400
      Re: forth to drive automated fast food machines the_gavino_himself <visphatesjava@gmail.com> - 2014-11-08 23:30 -0800
        Re: forth to drive automated fast food machines "Rod Pemberton" <buzzoff@nonamewhichexists.cmm> - 2014-11-09 06:08 -0500

#33901 — forth to drive automated fast food machines

Forth has a huge opportunity here I think


https://www.youtube.com/watch?v=FmXLqImT1wE

[toc] | [next] | [standalone]

#33928

gavino_himself wrote:

> Forth has a huge opportunity here I think
> 
> 
> https://www.youtube.com/watch?v=FmXLqImT1wE

So what are all the grad students supposed to do to earn an income when they 
finish college?

-- 
********************************************************************
Paul E. Bennett IEng MIET.....<email://Paul_E.Bennett@topmail.co.uk>
Forth based HIDECS Consultancy.............<http://www.hidecs.co.uk>
Mob: +44 (0)7811-639972
Tel: +44 (0)1235-510979
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
********************************************************************

[toc] | [prev] | [next] | [standalone]

#33931

On Wednesday, October 15, 2014 5:12:41 PM UTC-4, Paul E Bennett wrote:
> gavino_himself wrote:
> 
> 
> 
> > Forth has a huge opportunity here I think
> 
> > 
> 
> > 
> 
> > https://www.youtube.com/watch?v=FmXLqImT1wE
> 
> 
> 
> So what are all the grad students supposed to do to earn an income when they 
> 
> finish college?
> 
> 
> 
> -- 
> 
> ********************************************************************
> 
> Paul E. Bennett IEng MIET.....<email://Paul_E.Bennett@topmail.co.uk>
> 
> Forth based HIDECS Consultancy.............<http://www.hidecs.co.uk>
> 
> Mob: +44 (0)7811-639972
> 
> Tel: +44 (0)1235-510979
> 
> Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
> 
> ********************************************************************

Sidebar...

The exponential nature of the change we are in could mean that more than grad students are looking for jobs.  I heard an MIT prof in an interview say that "experts" in 2004 were saying, that driverless cars were not in the foreseeable future. They were being tested in 2010.  We struggle with exponential change, as a species.

[toc] | [prev] | [next] | [standalone]

#33945

On Wednesday, October 15, 2014 6:36:56 PM UTC-7, foxaudio...@gmail.com wrote:

> that driverless cars were not in the foreseeable future. 
> They were being tested in 2010. 

I'm sorry but I have a real problem believing that computer programs will even be good enough to drive a car in traffic.  There are just too many variables.  They will try. 

[toc] | [prev] | [next] | [standalone]

#33946 — Re: Google to drive car (was forth to drive automated fast food machines)

on 16/10/2014 15:52:42, Jason Damisch wrote:
> On Wednesday, October 15, 2014 6:36:56 PM UTC-7, foxaudio...@gmail.com
> wrote:
> 
>> that driverless cars were not in the foreseeable future.
>> They were being tested in 2010.
> 
> I'm sorry but I have a real problem believing that computer programs
> will even be good enough to drive a car in traffic. There are just too
> many variables. They will try.
> 


I drove past the Google car this summer heading north up to San Fransisco
past Sunnyvale on Route 101 (I work not far away from Google's HQ).
Search on "google self driving car on route 101" for a youtube video.

It's real, it's autonomous, and it appears to be as good a driver as
other road users on that stretch of road. How did I know it was a self
driving car? It says in big letters all over the bumpers "Self Driving
Car". I don't think they're kidding. The human inside wasn't in the
drivers seat.

The problem we have with predicting the future is that some things become
possible long before we expect them, and others that we expect to arrive
"real soon now" fail to arrive at all. It was ever thus.

[toc] | [prev] | [next] | [standalone]

#33952

Jason Damisch wrote:

> On Wednesday, October 15, 2014 6:36:56 PM UTC-7, foxaudio...@gmail.com
> wrote:
> 
>> that driverless cars were not in the foreseeable future.
>> They were being tested in 2010.
> 
> I'm sorry but I have a real problem believing that computer programs will
> even be good enough to drive a car in traffic.  There are just too many
> variables.  They will try.

I've a real problem believing that humans can be good enough to drive a real 
car in traffic. Come on, we are talking about some 10k people killed every 
year in the US, and accidents in the order of a million or so.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]

#33953

On 10/16/2014 3:38 PM, Bernd Paysan wrote:
> Jason Damisch wrote:
>
>> On Wednesday, October 15, 2014 6:36:56 PM UTC-7, foxaudio...@gmail.com
>> wrote:
>>
>>> that driverless cars were not in the foreseeable future.
>>> They were being tested in 2010.
>>
>> I'm sorry but I have a real problem believing that computer programs will
>> even be good enough to drive a car in traffic.  There are just too many
>> variables.  They will try.
>
> I've a real problem believing that humans can be good enough to drive a real
> car in traffic. Come on, we are talking about some 10k people killed every
> year in the US, and accidents in the order of a million or so.

*10k*!!???  More like 33k!!!!  It used to be a lot higher with a peak of 
54k in '72.  All the safety pushes including both car and highway 
designs have a lot to do with the decline.

I expect the problems with the autonomous cars will be legal rather than 
technical.

-- 

Rick

[toc] | [prev] | [next] | [standalone]

#33966

rickman wrote:
> *10k*!!???  More like 33k!!!!

That was more about the order of magnitue, I didn't look up the exact 
numbers.  33k indicates a lot of room for improvement.  Germany has a bit 
more than 3k per year now, so with one fourth of the population one tenth of 
the dead toll... and yes, we have that German Autobahn where you can drive 
as fast as you like (20% left or so, the rest is regulated).

> It used to be a lot higher with a peak of
> 54k in '72.

We had the peak about at the same time, 20k in 1970.

> All the safety pushes including both car and highway
> designs have a lot to do with the decline.

The compulsary wearing of the belt did show up in the statistics.  We have 
an almost linear decline since then, and if the decline continues as it is, 
by about 2020, there will be zero deaths in traffic...

> I expect the problems with the autonomous cars will be legal rather than
> technical.

You mean political.  The legal problems can easily be solved by requiring 
that the company who develops the autopilot is responsible for the 
accidents, and therefore has to pay the insurance.  The political problems 
like e.g. the taxi drivers preventing autonomous taxis to take their place 
or the car manufacturers who like to sell a car to everybody, which is then 
parking most of the time... while the number of autonomous cars just have to 
satisfy the transpor needs...  These problems are political, and 
challenging.

Rooting your autopilot of course voids that warranty... but making this 
illegal can have legal problems, too (e.g. what about the GPL'd software 
supporting the autopilot?).

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]

#33967

Bernd Paysan said:

> You mean political.  The legal problems can easily be solved by requiring 
> that the company who develops the autopilot is responsible for the 
> accidents, and therefore has to pay the insurance.  

Would you like to insure the developers? I know I wouldn't.

  Ian

[toc] | [prev] | [next] | [standalone]

#33968

Just Ian wrote:

> Bernd Paysan said:
> 
>> You mean political.  The legal problems can easily be solved by requiring
>> that the company who develops the autopilot is responsible for the
>> accidents, and therefore has to pay the insurance.
> 
> Would you like to insure the developers? I know I wouldn't.

We already have quite some high-reliable software out there where someone 
needs an insurance to deploy it.  This might not be the actual developer, 
but the company who builds the car - someone has to take over that risk.  If 
your car crashes due to a failure of the anti-blocking system, your 
insurance will go after the car manufacturer, and his insurance will pay.

I think that part of the legal stuff is already in place, it needs some 
minor modification so that self-driving cars are covered, as well.

Of course, an insurance company needs some data to adjust their police; that 
data is obtained during testing and initial roll-out (it will take more than 
a decade to convert to 99% fully autonomous driving vehicles on the 
streets).

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]

#34373

On Thursday, October 16, 2014 4:26:25 PM UTC-7, Bernd Paysan wrote:
> rickman wrote:
> > *10k*!!???  More like 33k!!!!
> 
> That was more about the order of magnitue, I didn't look up the exact 
> numbers.  33k indicates a lot of room for improvement.  Germany has a bit 
> more than 3k per year now, so with one fourth of the population one tenth of 
> the dead toll... and yes, we have that German Autobahn where you can drive 
> as fast as you like (20% left or so, the rest is regulated).
> 
> > It used to be a lot higher with a peak of
> > 54k in '72.
> 
> We had the peak about at the same time, 20k in 1970.
> 
> > All the safety pushes including both car and highway
> > designs have a lot to do with the decline.
> 
> The compulsary wearing of the belt did show up in the statistics.  We have 
> an almost linear decline since then, and if the decline continues as it is, 
> by about 2020, there will be zero deaths in traffic...
> 
> > I expect the problems with the autonomous cars will be legal rather than
> > technical.
> 
> You mean political.  The legal problems can easily be solved by requiring 
> that the company who develops the autopilot is responsible for the 
> accidents, and therefore has to pay the insurance.  The political problems 
> like e.g. the taxi drivers preventing autonomous taxis to take their place 
> or the car manufacturers who like to sell a car to everybody, which is then 
> parking most of the time... while the number of autonomous cars just have to 
> satisfy the transpor needs...  These problems are political, and 
> challenging.
> 
> Rooting your autopilot of course voids that warranty... but making this 
> illegal can have legal problems, too (e.g. what about the GPL'd software 
> supporting the autopilot?).
> 
> -- 
> Bernd Paysan
> "If you want it done right, you have to do it yourself"
> http://bernd-paysan.de/

but germany doesn't have american africans and illegal latinos

[toc] | [prev] | [next] | [standalone]

#33970

On Thursday, October 16, 2014 12:38:30 PM UTC-7, Bernd Paysan wrote:
> Jason Damisch wrote:
> 
> > On Wednesday, October 15, 2014 6:36:56 PM UTC-7, foxaudio...@gmail.com
> > wrote:
> 
> >> that driverless cars were not in the foreseeable future.
> >> They were being tested in 2010.
> 
> > I'm sorry but I have a real problem believing that computer programs will
> > even be good enough to drive a car in traffic.  There are just too many
> > variables.  They will try.
> 
> I've a real problem believing that humans can be good enough to drive a real 
> car in traffic. Come on, we are talking about some 10k people killed every 
> year in the US, and accidents in the order of a million or so.

The problem I foresee, is that all of the cars will be using the same algorithm, and there will be harmonic-resonance (a feedback-loop) between the cars as they respond to each other.

Of course, we have the same problem with people. One person slows down to gawk at a traffic accident on the other side of the road, and the people behind have to slow down to avoid rear-ending him. But, while they are driving slow they become distracted by gawking at the traffic accident on the other side of the road, and so they don't speed up as soon as the should --- pretty soon you have miles of traffic jam and the vast majority don't know why they are driving so slow, and there is no physical problem with the road to cause anybody to drive slow.

Maybe computers will actually be smarter than people because they are less likely to be distracted. When the AI enthusiasts predict that computers will someday be as smart as people, I always ask: "Does this mean that the computers will smoke crack, watch television, and collect welfare?" We can add "cause traffic jams" to that list as well.

To avoid the feedback-loop problem, the cars' algorithm would need a random element, so it is not doing the same thing as every other car, and a harmonic-resonance will be disrupted. So, if you're riding in an automated car, and it suddenly switches lanes for no apparent reason or does some other pointless maneuver, that would be the Random-Number-God taking control.

This reminds me of the story of how some guys were herding sheep along a narrow path. There was a tree branch in the path, so each sheep had to jump over it, which was slowing down the whole process. One of the guys grabbed the branch and tossed it into the ditch, hoping to speed the process up. Weirdly enough, the sheep continued to perform a jump at the location where the branch had been. Apparently, one sheep had seen the sheep in front jump over the branch, and had assumed that there must be a good reason for jumping at that spot, so it did a jump at that spot too --- this was despite the fact that the guy had pulled the branch away in the meantime. Then the next sheep saw the sheep in front jump at that spot, and assumed that there must be a good reason for jumping at that spot...

You know, there is a reason why the word "sheeple" got invented!

[toc] | [prev] | [next] | [standalone]

#34002

On Thu, 16 Oct 2014 22:55:26 -0400, <hughaguilar96@yahoo.com> wrote:
> On Thursday, October 16, 2014 12:38:30 PM UTC-7, Bernd Paysan wrote:
>> Jason Damisch wrote:
>> > On Wednesday, October 15, 2014 6:36:56 PM UTC-7, foxaudio...@gmail.com
>> > wrote:

>> >> that driverless cars were not in the foreseeable future.
>> >> They were being tested in 2010.
>
> [...]
>
> The problem I foresee, is that all of the cars will be using the same  
> algorithm,

Unlikely, the Germans would never use an algorithm the Americans
developed.  It's a matter of pride or nationalism or racism or some
other insane irrationality to them.  The Japanese and Chinese would
develop their own simply because they can, superiority complexes, but
claiming that cost or control was needed.

> Of course, we have the same problem with people. One person slows down  
> to gawk at a traffic accident on the other side of the road, and the  
> people behind have to slow down to avoid rear-ending him. But, while  
> they are driving slow they become distracted by gawking at the traffic  
> accident on the other side of the road, and so they don't speed up as  
> soon as the should --- pretty soon you have miles of traffic jam and the  
> vast majority don't know why they are driving so slow, and there is no  
> physical problem with the road to cause anybody to drive slow.

If the cop never turned on his lights or the news announcer never
announced snow or rain, most traffic jams wouldn't occur either.
So, this is a cascade failure, part of human nature in groups,
fearing something that isn't likely to happen to them.

> Maybe computers will actually be smarter than people because they are  
> less likely to be distracted.

There is no doubt in my mind that computers will excel with math
and logic.  It's the irrational and emotional and religious aspects
of human nature which we classify as part of our consciousness that
will not be mastered by machines.  This will also allow us to declare
otherwise sentient machines as being non-sentient allowing us to murder
them without cause and without it being a crime, much like we do with
hunting of animals or killing of livestock.

> When the AI enthusiasts predict that
> computers will someday be as smart as people, I always ask: "Does this
> mean that the computers will smoke crack, watch television, and collect
> welfare?" We can add "cause traffic jams" to that list as well.

Smart is not the same as rational.  Don't conflate the two.
There are many smart, but irrational liberals.  Some are lame
comedians on TV.

> To avoid the feedback-loop problem, the cars' algorithm would need a
> random element, so it is not doing the same thing as every other car,
> and a harmonic-resonance will be disrupted.

It's a design feature.  When cars are moving in a pack and controlled
by computers, no one can travel faster than the slowest moving vehicle
for undetermined safety reasons ...

> So, if you're riding in an automated car, and it suddenly switches
> lanes for no apparent reason or does some other pointless maneuver,
> that would be the Random-Number-God taking control.

More likely it would be a hacker or rogue government agency with
too much authority to hide it's crimes.  We can't keep Russian
and Chinese hackers out of our computer systems now.  Obama can't
keep American Mexicans from jumping the fence at the White House.
There is no way he could stop Mexican Mexicans from jumping the
fence at the border.


Rod Pemberton

[toc] | [prev] | [next] | [standalone]

#34374

On Thursday, October 16, 2014 12:38:30 PM UTC-7, Bernd Paysan wrote:
> Jason Damisch wrote:
> 
> > On Wednesday, October 15, 2014 6:36:56 PM UTC-7, foxaudio...@gmail.com
> > wrote:
> > 
> >> that driverless cars were not in the foreseeable future.
> >> They were being tested in 2010.
> > 
> > I'm sorry but I have a real problem believing that computer programs will
> > even be good enough to drive a car in traffic.  There are just too many
> > variables.  They will try.
> 
> I've a real problem believing that humans can be good enough to drive a real 
> car in traffic. Come on, we are talking about some 10k people killed every 
> year in the US, and accidents in the order of a million or so.
> 
> -- 
> Bernd Paysan
> "If you want it done right, you have to do it yourself"
> http://bernd-paysan.de/

alcohol n drugs

[toc] | [prev] | [next] | [standalone]

#34001

On Thu, 16 Oct 2014 10:52:46 -0400, Jason Damisch <jasondamisch@yahoo.com>  
wrote:

> On Wednesday, October 15, 2014 6:36:56 PM UTC-7, foxaudio...@gmail.com  
> wrote:
>
>> that driverless cars were not in the foreseeable future.
>> They were being tested in 2010.
>
> I'm sorry but I have a real problem believing that computer programs  
> will even be good enough to drive a car in traffic.  There are just too  
> many variables.  They will try.

I'm not sure why you say that.  It's primarily a matter of edge detection
which video games have mastered.  Don't hit the walls in any maze or Quake
like game.  Of course, the walls are the roadlines and other vehicles,
computer vision processing is needed to identify the roadlines and vehicles
in real-time.  Modern video cards have the ability to do physics  
computations.


Rod Pemberton

[toc] | [prev] | [next] | [standalone]

#34371

On Wednesday, October 15, 2014 2:12:41 PM UTC-7, Paul E Bennett wrote:
> gavino_himself wrote:
> 
> > Forth has a huge opportunity here I think
> > 
> > 
> > https://www.youtube.com/watch?v=FmXLqImT1wE
> 
> So what are all the grad students supposed to do to earn an income when they 
> finish college?
> 
> -- 
> ********************************************************************
> Paul E. Bennett IEng MIET.....<email://Paul_E.Bennett@topmail.co.uk>
> Forth based HIDECS Consultancy.............<http://www.hidecs.co.uk>
> Mob: +44 (0)7811-639972
> Tel: +44 (0)1235-510979
> Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
> ********************************************************************

shake that thing 

[toc] | [prev] | [next] | [standalone]

#34000

On Tue, 14 Oct 2014 19:47:35 -0400, gavino_himself  
<visploveslisp@gmail.com> wrote:

> Forth has a huge opportunity here I think
>

There are factory pancake machines that do much the same.

http://www.youtube.com/watch?v=vc-Ra-wlZzQ

No Forth is required for such systems.


Rod Pemberton

[toc] | [prev] | [next] | [standalone]

#34372

On Friday, October 17, 2014 7:33:08 PM UTC-7, Rod Pemberton wrote:
> On Tue, 14 Oct 2014 19:47:35 -0400, gavino_himself  
> <visploveslisp@gmail.com> wrote:
> 
> > Forth has a huge opportunity here I think
> >
> 
> There are factory pancake machines that do much the same.
> 
> http://www.youtube.com/watch?v=vc-Ra-wlZzQ
> 
> No Forth is required for such systems.
> 
> 
> Rod Pemberton

yummy

what are they programmed in?

[toc] | [prev] | [next] | [standalone]

#34375

On Sun, 09 Nov 2014 02:30:07 -0500, the_gavino_himself  
<visphatesjava@gmail.com> wrote:
> On Friday, October 17, 2014 7:33:08 PM UTC-7, Rod Pemberton wrote:
>> On Tue, 14 Oct 2014 19:47:35 -0400, gavino_himself
>> <visploveslisp@gmail.com> wrote:

>> > Forth has a huge opportunity here I think
>> >
>>
>> There are factory pancake machines that do much the same.
>>
>> http://www.youtube.com/watch?v=vc-Ra-wlZzQ
>>
>> No Forth is required for such systems.
>>
>>
>
> yummy
>
> what are they programmed in?


It appears to be pancake batter ...


RP

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.forth

csiph-web