Self driving buses, trains, trucks etc

markodaly · 17-11-2018 10:11am

Meanwhile, in Ireland 50,000 people over a two year period have been charged with using a mobile phone while driving.
https://www.thesun.ie/news/3392743/mobile-phones-driving-irish-roads/

This included drivers who were taking selfies, live streaming and even watching Netflix while driving.
Humans are indeed a stupid careless animal and the sooner self-driving cars take over the majority of driving activity the better for road safety.

TheQuietFella · 17-11-2018 10:50am

orangerhyme wrote: »

It'll be an economic catastrophe.
Self-driving electric cars will lead to huge job losses in taxi drivers, van, truck, pizza delivery, also petrol stations, mechanics, car dealerships, car parts manufactures etc etc.
Some traditional big name manufacturers will go bust.
Car ownership and sales will fall off a cliff sometime in the 2020s

I wouldn't be worrying too much about it because it's never going to happen!

cnocbui · 17-11-2018 12:07pm

TheQuietFella wrote: »

I wouldn't be worrying too much about it because it's never going to happen!

Never say never, but I think the optimism among many is far too high and it isn't coming soon beyond major arterial roads like motorways.

Self-Driving Cars Can Handle Neither Rain nor Sleet nor Snow

As things stand today, the driverless car of the future can’t handle more than a dusting of snow.

It’s a known problem in the field, and vaguely embarrassing when the end result is supposed to be robots sophisticated enough to navigate the uncertainties of traffic and improve on lackluster human perception. In Boston, where NuTonomy has been road testing autonomous vehicles in cooperation with city planning officials, snow and seagulls have emerged as two of the biggest obstacles.

I'm sorry Dave, I can no longer take you to your destination. Dave you must take back control now.

Twenty Grand · 17-11-2018 12:23pm

cnocbui wrote: »

Never say never, but I think the optimism among many is far too high and it isn't coming soon beyond major arterial roads like motorways.

I'm sorry Dave, I can no longer take you to your destination. Dave you must take back control now.

And from the same article:
"Xavier Mosquet, senior partner at the Boston Consulting Group, thinks it will be at least two years before even an elite player such as Waymo is able to code a car brain to consistently handle the challenges of a Boston-like climate."

2 whole years?! Cancel everything!

markodaly · 17-11-2018 3:03pm

Twenty Grand wrote: »

And from the same article:
"Xavier Mosquet, senior partner at the Boston Consulting Group, thinks it will be at least two years before even an elite player such as Waymo is able to code a car brain to consistently handle the challenges of a Boston-like climate."

2 whole years?! Cancel everything!

If a Waymo car can consistently handle a Boston type climate and traffic scenario, that means it can handle the vast vast majority of car traffic. Game changer indeed.

dense · 17-11-2018 5:10pm

markodaly wrote: »

If a Waymo car can consistently handle a Boston type climate and traffic scenario, that means it can handle the vast vast majority of car traffic. Game changer indeed.

IF it can, great.

But it will take "at least" two years to find out, so that could be 5, 9, or 15 years in reality.

cnocbui · 17-11-2018 5:23pm

Twenty Grand wrote: »

And from the same article:
"Xavier Mosquet, senior partner at the Boston Consulting Group, thinks it will be at least two years before even an elite player such as Waymo is able to code a car brain to consistently handle the challenges of a Boston-like climate."

2 whole years?! Cancel everything!

Have they managed to finish the software they thought could be written to get the F35 fighter to do all the things they said it could do?

Didn't think so.

Twenty Grand · 17-11-2018 6:50pm

cnocbui wrote: »

Have they managed to finish the software they thought could be written to get the F35 fighter to do all the things they said it could do?

Didn't think so.

Didnt know Google was working on the F35?
Do they have literally the best software people in the world working on the F35 fighter?

markodaly · 18-11-2018 2:46am

dense wrote: »

IF it can, great.

But it will take "at least" two years to find out, so that could be 5, 9, or 15 years in reality.

2 years is 24 months. That is nothing. As I said, that is the end goal.

We pretty much already know that it can handle dry climates like Phoneix and traffic conditions like that now easily enough.

In my opinion, as I said. 2025-2030 is when we will reach the point of inflection in regards to self-driving cars and EV's.

The fact that some think this tech is like 50 years away are utterly deluded.

RayCun · 18-11-2018 7:56am

markodaly wrote: »

2 years is 24 months. That is nothing. As I said, that is the end goal.

We pretty much already know that it can handle dry climates like Phoneix and traffic conditions like that now easily enough.

In my opinion, as I said. 2025-2030 is when we will reach the point of inflection in regards to self-driving cars and EV's.

The fact that some think this tech is like 50 years away are utterly deluded.

"At least two years" means it is definitely not happening in the next 24 months. Beyond that, they haven't a clue.

18-11-2018 8:22am

RayCun wrote: »

"At least two years" means it is definitely not happening in the next 24 months. Beyond that, they haven't a clue.

Semantics

12,24,36 months

Whichever number it is, it's still a lot sooner than originally anticipated by a lot of people

As a pedestrian, cyclist and motorist, I am looking forward to seeing these become the norm

RayCun · 18-11-2018 8:29am

[Deleted User] wrote: »

Semantics

12,24,36 months

Whichever number it is, it's still a lot sooner than originally anticipated by a lot of people

As a pedestrian, cyclist and motorist, I am looking forward to seeing these become the norm

How can you say it is a lot sooner than anticipated when you don't know how soon it will be?

All he has said is that it _won't_ be in the next two years.

18-11-2018 8:40am

RayCun wrote: »

How can you say it is a lot sooner than anticipated when you don't know how soon it will be?

All he has said is that it _won't_ be in the next two years.

Like I said, Semantics

magicbastarder · 18-11-2018 9:57am

cnocbui wrote: »

Have they managed to finish the software they thought could be written to get the F35 fighter to do all the things they said it could do?

Didn't think so.

there's a lot more going on with the F35 than just avionics software though:
https://www.popularmechanics.com/military/a21957/wtf-35/

to suggest that because they're having trouble writing the software for it means we can't have self driving cars (?) would also suggest that because they've had trouble designing the wing for it, means wings are not possible to be built dependably.

cnocbui · 18-11-2018 12:48pm

magicbastarder wrote: »

there's a lot more going on with the F35 than just avionics software though:
https://www.popularmechanics.com/military/a21957/wtf-35/

to suggest that because they're having trouble writing the software for it means we can't have self driving cars (?) would also suggest that because they've had trouble designing the wing for it, means wings are not possible to be built dependably.

I didn't say 'can't'. I gave that as an example of how the unbridled optimism of certain people that assumes if you can think it you can do it - applied to software - is misguided.

That's what people think about self driving cars. The worst thing is that I think a huge number of people who comment on this think driving is a simple task and so should be trivial to write software to replicate what a lot of people can do with great competence.

I think the worst thing is that a lot of people think people are generally bad drivers and that autonomous driving systems will be a lot better.
Personally I suspect that general intelligence is required for full driving competence and that it will be found to be beyond the neural network approach that most of the companies are using.

If the target was actually self-aware general intelligence - real AI - who would be thinking that's only a decade away from full reality?

cnocbui · 18-11-2018 12:55pm

magicbastarder wrote: »

there's a lot more going on with the F35 than just avionics software though:
https://www.popularmechanics.com/military/a21957/wtf-35/

to suggest that because they're having trouble writing the software for it means we can't have self driving cars (?) would also suggest that because they've had trouble designing the wing for it, means wings are not possible to be built dependably.

I didn't say 'can't'. I gave that as an example of how the unbridled optimism of certain people that assumes if you can think it you can do it - applied to software - is misguided.

That's what people think about self driving cars. The worst thing is that I think a huge number of people who comment on this think driving is a simple task and so should be trivial to write software to replicate what a lot of people can do with great competence.

Personally I suspect that general intelligence is required for full driving competence and that it will be found to be beyond the neural network approach that most of the companies are using.

If the target was actually self-aware general intelligence - real AI - who would be thinking that's only a decade away from full reality?

The worst thing is that a lot of people seem to think people in general are bad drivers and that AI autonomous driving systems will be far better to the point of infallibility. Just wait for the first malicious hack to fix that notion.

markodaly · 18-11-2018 12:56pm

cnocbui wrote: »

I think the worst thing is that a lot of people think people are generally bad drivers and that autonomous driving systems will be a lot better.
Personally I suspect that general intelligence is required for full driving competence and that it will be found to be beyond the neural network approach that most of the companies are using.

If the target was actually self-aware general intelligence - real AI - who would be thinking that's only a decade away from full reality?

That is a more philosophical discussion really but I would image the best brains in Google and Waymo have thought of this and decided on what they think is the best course of action.

If you believe they are wrong, well send in your CV to them, get hired and tell them to do it right, as you obviously know better.

AndrewJRenko · 18-11-2018 1:00pm

cnocbui wrote: »

Personally I suspect that general intelligence is required for full driving competence and that it will be found to be beyond the neural network approach that most of the companies are using.

If the target was actually self-aware general intelligence - real AI - who would be thinking that's only a decade away from full reality?

The worst thing is that a lot of people seem to think people in general are bad drivers and that AI autonomous driving systems will be far better to the point of infallibility. Just wait for the first malicious hack to fix that notion.

I guess the hope comes from the idea that the neural network won't be phoning mammy at the same time, or eating breakfast, or doing makeup, or doing all the other things drivers are supposed to be doing when concentrating on their driving.

cnocbui · 18-11-2018 1:03pm

markodaly wrote: »

That is a more philosophical discussion really but I would image the best brains in Google and Waymo have thought of this and decided on what they think is the best course of action.

If you believe they are wrong, well send in your CV to them, get hired and tell them to do it right, as you obviously know better.

My software writing skills aren't up to it. My son on the other hand would be well up for it, but he's working on something else at the moment. He is probably more skeptical than I am, and he writes code probably 12 hours a day, 365.

cnocbui · 18-11-2018 1:15pm

AndrewJRenko wrote: »

I guess the hope comes from the idea that the neural network won't be phoning mammy at the same time, or eating breakfast, or doing makeup, or doing all the other things drivers are supposed to be doing when concentrating on their driving.

Well people's hopes are beyond simplistic. Neural networks have their own problems, such as the black box problem wherein if anything goes wrong, you have absolutely no idea how or why.

People should read about Toyota's 'simple' brake software problem and how it proved impossible to find the fault, know why it happened or fix it with a simple tweak.

Boeing just wrote some software for the new 737 max that killed 189 people.

RayCun · 18-11-2018 2:05pm

I think self-driving cars will get here eventually, but people underestimate the difficulties involved.

On one side, turning visual information into actionable data is very, very difficult. On the other, using network data (from traffic lights, other cars, traffic updates) is a security nightmare.

There's a big difference between a car that beeps when you are moving out of your lane, and a car that you get into, set your destination, and have a nap.

prinzeugen · 18-11-2018 3:48pm

cnocbui wrote: »

Have they managed to finish the software they thought could be written to get the F35 fighter to do all the things they said it could do?

Didn't think so.

The F35 needs hundreds of millions lines of code. It will take many people, many years to write that and test it.

The F35 in its current basic form is good to go and has already been combat proven.

Twenty Grand · 18-11-2018 5:22pm

cnocbui wrote: »

My software writing skills aren't up to it. My son on the other hand would be well up for it, but he's working on something else at the moment. He is probably more skeptical than I am, and he writes code probably 12 hours a day, 365.

I'd trust the folks over at Google more than you and your son.

As I said before, you can be as enthusiastic or skeptical as you like.
We'll know very soon what the future for autonomous cars looks like.

TheQuietFella · 19-11-2018 1:20am

Deleted User wrote: »

Don't know about the rest of you, but I am looking forward to seeing driverless buses, trains, taxis and luas

It's really not as far away as people think

Why do idiots continue with this drivel!

It will NEVER EVER happen 'cause they couldn't or just wouldn't be
able to factor in the human element or all the eventualities that
govern or surround transport!

Del.Monte · 19-11-2018 1:41am

TheQuietFella wrote: »

Why do idiots continue with this drivel!

It will NEVER EVER happen 'cause they couldn't or just wouldn't be
able to factor in the human element or all the eventualities that
govern or surround transport!

Probably because they watched too many SciFi movies.

Twenty Grand · 19-11-2018 3:35am

TheQuietFella wrote: »

Why do idiots continue with this drivel!

It will NEVER EVER happen 'cause they couldn't or just wouldn't be
able to factor in the human element or all the eventualities that
govern or surround transport!

We'll see how well the Waymo rollout next month works.

prinzeugen · 19-11-2018 3:39am

Twenty Grand wrote: »

We'll see how well the Waymo rollout next month works.

https://driving.ca/chrysler/auto-news/news/waymo-ceo-says-self-driving-cars-wont-be-ubiquitious-for-decades

Twenty Grand · 19-11-2018 3:44am

prinzeugen wrote: »

https://driving.ca/chrysler/auto-news/news/waymo-ceo-says-self-driving-cars-wont-be-ubiquitious-for-decades

Sure that's fair enough.

Most cars have a 20 odd year life, so decades for ubiquity is reasonable.
I also don't mind the limited human interaction. Even if you have to take control 1% of the time, that's 99% autonomous.

prinzeugen · 19-11-2018 4:47am

Twenty Grand wrote: »

Sure that's fair enough.

Most cars have a 20 odd year life, so decades for ubiquity is reasonable.
I also don't mind the limited human interaction. Even if you have to take control 1% of the time, that's 99% autonomous.

But not 100% driverless. That will never happen in our lifetime.

RayCun · 19-11-2018 7:47am

Twenty Grand wrote: »

Sure that's fair enough.

Most cars have a 20 odd year life, so decades for ubiquity is reasonable.
I also don't mind the limited human interaction. Even if you have to take control 1% of the time, that's 99% autonomous.

If the 1% of time you have to take control is not predictable, then you have to spend 100% of the journey with your hands on the wheel, paying attention to your surroundings. Not snoozing, reading, watching movies...

magicbastarder · 19-11-2018 9:33am

TheQuietFella wrote: »

It will NEVER EVER happen 'cause they couldn't or just wouldn't be able to factor in the human element or all the eventualities that govern or surround transport!

humans are unable to factor in all the eventualities too.

markodaly · 19-11-2018 11:51am

cnocbui wrote: »

My software writing skills aren't up to it. My son on the other hand would be well up for it, but he's working on something else at the moment. He is probably more skeptical than I am, and he writes code probably 12 hours a day, 365.

Good for him, there is a fortune to be made in IT for those with the right skills. If he is as good as you claim him to be, a job in Silicon Valley paying north of $250,000 a year awaits him.

markodaly · 19-11-2018 11:53am

TheQuietFella wrote: »

Why do idiots continue with this drivel!

It will NEVER EVER happen 'cause they couldn't or just wouldn't be
able to factor in the human element or all the eventualities that
govern or surround transport!

Never Ever? And you call other people idiots? ROFL

fly_agaric · 19-11-2018 10:59pm

cnocbui wrote: »

My software writing skills aren't up to it. My son on the other hand would be well up for it

...

If the target was actually self-aware general intelligence - real AI - who would be thinking that's only a decade away from full reality?

Kudos - an excellent humble-brag.:D
Am wondering about one thing. Why do you think a successful (safer than human) self driving car would need to be "self aware" to do the job? (that's kind of what that comment suggests I think?)
That seems like a very high bar.
I suppose it may involve getting closer to a sort of (limited) "general intelligence" than machines have gotten before (seeing [via sensors] + interpreting what is seen and acting on it at a human level of ability for specific task of driving a vehicle on a public road - edit: it's one job but there is quite a lot involved in it), but why self aware?

cnocbui · 19-11-2018 11:35pm

fly_agaric wrote: »

...

Kudos - an excellent humble-brag.:D
Am wondering about one thing. Why do you think a successful (safer than human) self driving car would need to be "self aware" to do the job? (that's kind of what that comment suggests I think?)
That seems like a very high bar.
I suppose it may involve getting closer to a sort of (limited) "general intelligence" than machines have gotten before (seeing [via sensors] + interpreting what is seen and acting on it at a human level of ability for specific task of driving a vehicle on a public road - edit: it's one job but there is quite a lot involved in it), but why self aware?

Well I suppose it doesn't have to be self aware, I just suspect general intelligence might be inseparable from self awareness. Why do I think GI - because driving isn't wrote: ever observed another car, vehicle cyclist or pedestrian and read their intentions from their 'behaviour'; That 6-8 year old girl sprinting down the footpath in the direction of pedestrian crossing. No neural network is going to predict the next sequence of events. I look cars ahead often to be prepared. Anyone who thinks real driving in all weather conditions with multiple types of vehicles, objects, children pedestrians can be distilled to basically a complex but predictable scalextric set is simple minded.

A large amount of driving is entirely predictable, but the edge cases can be doozies.

20-11-2018 6:34am

cnocbui wrote: »

Well I suppose it doesn't have to be self aware, I just suspect general intelligence might be inseparable from self awareness. Why do I think GI - because driving isn't wrote: ever observed another car, vehicle cyclist or pedestrian and read their intentions from their 'behaviour'; That 6-8 year old girl sprinting down the footpath in the direction of pedestrian crossing. No neural network is going to predict the next sequence of events. I look cars ahead often to be prepared. Anyone who thinks real driving in all weather conditions with multiple types of vehicles, objects, children pedestrians can be distilled to basically a complex but predictable scalextric set is simple minded.

A large amount of driving is entirely predictable, but the edge cases can be doozies.

You are talking about reaction times to random issues

A machine, with a multitude of sensors, will always out perform humans. This is old news

Folks can say this will not be coming for x many years but it will be coming sooner than most realise.

The big push to spread mass adoption will be from the insurance industry i.e. Drive yourself if you want, but expect to pay a farcical premium for the privilege

AlmightyCushion · 20-11-2018 12:13pm

cnocbui wrote: »

Well I suppose it doesn't have to be self aware, I just suspect general intelligence might be inseparable from self awareness. Why do I think GI - because driving isn't wrote: ever observed another car, vehicle cyclist or pedestrian and read their intentions from their 'behaviour'; That 6-8 year old girl sprinting down the footpath in the direction of pedestrian crossing. No neural network is going to predict the next sequence of events. I look cars ahead often to be prepared. Anyone who thinks real driving in all weather conditions with multiple types of vehicles, objects, children pedestrians can be distilled to basically a complex but predictable scalextric set is simple minded.

A large amount of driving is entirely predictable, but the edge cases can be doozies.

That's exactly what a neural net does. It learns from experience just like humans do. The neural net for self driving cars won't be based off the information and experiences of just one car, it will be based on the millions of cars they have on the road collecting data.

Alpha Go is a great example of this. It was just given the rules of the game and then played game after game. It eventually discovered effective techniques and strategies that humans had developed. Over time it even discovered ones of its own that humans now use.

cnocbui · 20-11-2018 2:12pm

AlmightyCushion wrote: »

That's exactly what a neural net does. It learns from experience just like humans do. The neural net for self driving cars won't be based off the information and experiences of just one car, it will be based on the millions of cars they have on the road collecting data.

Alpha Go is a great example of this. It was just given the rules of the game and then played game after game. It eventually discovered effective techniques and strategies that humans had developed. Over time it even discovered ones of its own that humans now use.

A neural net can learn how to do things, only if it has encountered the circumstances before. No neural network can anticipate or predict, particularly by observing human behaviour - their sensor systems couldn't even 'see' the little girl running down the footpath, let alone understand how children behave nor anticipate or predict what could happen next.

I think general intelligence is required to match human driving capabilities. Neural nets don't come close.

Remember Musk saying he would have an autonomous car cross the US? His latest tweet:

"V9 moving to wide release now. Holding back Autopilot drive on navigation for a few more weeks of validation. Extremely difficult to achieve a general solution for self-driving that works well everywhere."

Another shining example of how just thinking something can be done with technology is faith-based rather than reality based is Musk's belief that Tesla could have factories stuffed full of robots and clever software that could make cars at blindingly fast speeds that humans couldn't begin to match - 24/7 365, even though Fiat, GM, Toyota and Nissan have all tried that. End result was Musk employing lots more people, sidelining a lot of robots and tweeting 'humans are underrated'. No kidding.

cnocbui · 20-11-2018 2:13pm

Deleted User wrote: »

You are talking about reaction times to random issues

A machine, with a multitude of sensors, will always out perform humans. This is old news

Folks can say this will not be coming for x many years but it will be coming sooner than most realise.

The big push to spread mass adoption will be from the insurance industry i.e. Drive yourself if you want, but expect to pay a farcical premium for the privilege

Load of tosh from start to finish.

TheQuietFella · 20-11-2018 4:30pm

markodaly wrote: »

Never Ever? And you call other people idiots? ROFL

It will never, ever, ever, ever, happen!

plodder · 20-11-2018 4:42pm

We're only a few years from seeing self driving lorry convoys on motorways. In truth though, automation has been creeping in for a few years already, like the emergency braking technology on trucks. The lowest hanging fruit will be picked first. Eg, self driving cars on motorways. No pedestrians, predictable, consistent road design, far fewer "targets" to track. There is huge value in being able to drive yourself to the m-way entrance and let the machine take over until you leave the motorway where you take over again. As regards self driving outside of highways/motorways, it will happen sooner in the US than here; that's for certain. Road design standards are much more consistent (and even rational) in the US. Though again, you might find that it will happen here on a subset of roads that have been electronically surveyed first and the automation knows what to expect as regards road/junction layout. So, the default would always be manually operated, but some segments of a journey might be automatable.

RayCun · 20-11-2018 4:48pm

plodder wrote: »

We're only a few years from seeing self driving lorry convoys on motorways.

How many is 'a few'? Three?

How much would you bet that by the end of 2021 there will have been
- at least three convoys
- containing at least three trucks per convoy
- commercially operated (not a controlled test, but trucks carrying normal freight for an existing distributor)
- completely self-driving, no human driver required once they enter the motorway (even if the driver has hardly anything to do, or is only there 'just in case', if a driver is required in the cab ready to take over it is not a self-driving vehicle)

plodder · 20-11-2018 4:57pm

RayCun wrote: »

How many is 'a few'? Three?

How much would you bet that by the end of 2021 there will have been
- at least three convoys
- containing at least three trucks per convoy
- commercially operated (not a controlled test, but trucks carrying normal freight for an existing distributor)
- completely self-driving, no human driver required once they enter the motorway (even if the driver has hardly anything to do, or is only there 'just in case', if a driver is required in the cab ready to take over it is not a self-driving vehicle)

I'd say if the driver is not driving, it is a self-driving vehicle (at least some of the time).

The point of my post was that the technology has already begun creeping in and will continue to do so.

The article below says we would have self-driving convoys by the end of this year. So, three years seems a cautious enough prediction.

https://www.theguardian.com/politics/2017/aug/25/semi-automated-truck-convoy-trials-get-uk-go-ahead-platooning

cnocbui · 20-11-2018 4:58pm

They would love convoy automation in Australia. As it is now, big semi's and road trains slipstream crossing the country, at least on the Nullabor bit, where one truck will drive with a bare 2m gap between it and the back of the one in front. The following truck saves a lot of fuel as it has no wind resistance. After a few hours of that, tail end charlie will overtake and take over making a hole in the air while the former lead coasts behind.

Potentially very dangerous, you would think, but they have been at it forever.

RayCun · 20-11-2018 5:04pm

plodder wrote: »

I'd say if the driver is not driving, it is a self-driving vehicle (at least some of the time).

I disagree, not if a driver is needed in the cab and might have to take over at any time.
Putting your car in cruise control and taking your hands off the wheel doesn't make your car self-driving

RayCun · 20-11-2018 5:10pm

plodder wrote: »

The article below says we would have self-driving convoys by the end of this year. So, three years seems a cautious enough prediction.

https://www.theguardian.com/politics/2017/aug/25/semi-automated-truck-convoy-trials-get-uk-go-ahead-platooning

Note that the article was published in 2017 and said there would be trials in 2018, but the trials don't seem to have taken place.

dense · 20-11-2018 5:38pm

RayCun wrote: »

Note that the article was published in 2017 and said there would be trials in 2018, but the trials don't seem to have taken place.

Paper never refuses ink.

plodder · 20-11-2018 5:42pm

RayCun wrote: »

Note that the article was published in 2017 and said there would be trials in 2018, but the trials don't seem to have taken place.

I predict they will within three years though :pac:

On the definition of "self driving" there seems to be a consensus on how it is classified with five different levels.

In a way the truck convoy is a special case, in that the convoy is led by a human driver and the following vehicles aren't truly autonomous. But, it's an interesting case, because I'm convinced we will see them before any other kind of self-driving and the idea could spread to car convoys on motorways as well, with many of the same benefits as "true" autonomous cars.

RayCun · 20-11-2018 5:43pm

plodder wrote: »

There is huge value in being able to drive yourself to the m-way entrance and let the machine take over until you leave the motorway where you take over again.

This is what I would call self-driving. You get to the motorway and you don't have to lift a finger until you are leaving the motorway. Take a nap, read a book, watch a movie, eat a picnic - no engagement or attention required until you leave the motorway.

fly_agaric · 20-11-2018 10:10pm

cnocbui wrote: »

Well I suppose it doesn't have to be self aware, I just suspect general intelligence might be inseparable from self awareness. Why do I think GI - because driving isn't wrote: ever observed another car, vehicle cyclist or pedestrian and read their intentions from their 'behaviour'; That 6-8 year old girl sprinting down the footpath in the direction of pedestrian crossing. No neural network is going to predict the next sequence of events. I look cars ahead often to be prepared. Anyone who thinks real driving in all weather conditions with multiple types of vehicles, objects, children pedestrians can be distilled to basically a complex but predictable scalextric set is simple minded.

A large amount of driving is entirely predictable, but the edge cases can be doozies.

We'll see I suppose.
As said many times already the technology companies think the approaches they have now or refinements/extensions of them (which presumably won't ever deliver us a self aware machine, just one very good at doing the task it has learned/been trained to do) are enough for a problem like driving + are shovelling resources at it to be "first".
Assume a lot of heads will roll + large fortunes will go down the drain if they have gotten it as badly wrong as you think...

I'm not an expert at all but as far as I understand one of the difficulties with these neural network model approaches is almost the opposite of "predictability", that they are not easily understandable for humans. It is hard to pin down exactly why they have made a particular decision (which becomes very important when they make an incorrect decision that needs to be "debugged"/fixed) and simpler models are used to try and "explain" in a rough way what the neural network is doing/why it makes the choices it does. As mentioned above (new strategies found by ones trained to play games), they have thrown up surprises for their programmers/creators when applied to problems because of this unpredictable aspect.

Self driving buses, trains, trucks etc

Comments