So, according to this article by Mac rumours:
https://www.macrumors.com/2022/12/06/apple-car-project-scaled-back-delayed/
the apple car will no longer have full self driving capability. I know it was something that I was interested in, despite the price tag, and wanted to know other's thoughts.
I'm unsure if blind people can even operate full self driving cars, especially if human control is required during an emergency. One thing that was interesting in this rather sparse article was that apple might have a centre for car control and emergencies.
I wonder if, in the future, we could have a car that is human controlled at a distance. IE, people driving from home whilst we travel in the vehicle. All pie in the sky stuff but as the UK faces yet another train flurry of train strikes during the winter months, I do wonder about the future of alternatives. Long distance independence does seem appealing though uber works well in the short distance around my city of Bristol.
Comments
No a chance
Hi,
There is no way insurance companies would ever allow a totally blind person to use these cars. As you say there will always be potential where humans need to take over. Love the idea but I think not.
Reading between the lines,…
Reading between the lines, apple was looking to have a vehicle that had no controls at all. We can assume that this was therefore a fully automated vehicle... Which raises interesting issues when considering insurance. If a crash occurs, who''s fault is it? I'd like to think, Tim Cooks.
The other issue...
I'm a lot more worried about these so called, self-driving, electric vehicles running me over. It's hard to hear the electric cars as they are now if there's other traffic noise, and you throw in shoddy algorithms that creep in like bugs in an iPhone... It's just scary.
I'd go full throttle if I had the money
I'm always a person who doesn't let anyone else influence anything that happens to them, effect my decision. If someone used an iPhone for a week, says it's awful, no worries, it works for me. I expected there not to be fully self driving capabilities so this didn't surprise me. as for insurence, it wouldn't matter as my state has no fault insurence. As for electric cars and hearing them, I think by law they have to emit some sound so we can hear them.
Quanton computers
Today's AI and computers are not capable of doing so. You need one of those quantum computers that google and MIT are working on it. Self driving cars will not happen in this century unless Quantum computers come out.
Very little sound
I've had a few electric-hybrid taxis pick me up, and they're about as loud as a bicycle coasting by, almost no noise. That was on a very quiet street. There's no way those cars could be heard on a moderately busy street, like at a four-way stop where you need to hear the traffic flow. Now if there were only electric vehicles, and they were all being controled byAI, it might be different. More of us need to be run over to get this issue addressed.
Self-Driving Cars
No self-driving cars for me thank you. When I think of self-driving cars, I often laugh when thinking about the movie Total Recall, not that piece of garbage remake but the original, with that automated cab called Johnny Cab. Go watch/listen to it if you haven't. Anyway, the real reason no self-driving cars for me is because I do not believe for a second that artificial intelligence could ever keep up with human stupidity on the roads. Then again, perhaps we, as humans, are the real artificial intelligence? LOL
Computing power and quantum
It's my understanding that though quantum computers are excellent for certain tasks, they fall short when it comes to others when compared with solid state, contemporary computing. I think the computing is there, billions of calculations a second over multiple processors can handle pretty much anything, the key here is, however, what they are asked to do. It is the software that is always in question, the thought experiment, what should a computer do, protect the rider or the pedestrian.
the AI cars will, in theory, be more perceptive than human drivers so the electric car issue is moot. There are electric cars and there will be more of them, which is a good thing on a wider scale but admittedly more worrying on a mobility basis.
People will most certainlly be run over by AI controlled cars as people will most certainly be run over by bad drivers. the key question here is not if they will run people over, it is, will it be less than humans. If so, then it is a win, however we are a little touchy about our species so pointing out logic may fall on deaf ears.
Also, the thinly veiled 'Nuts to everyone else's opinion' whilst clearly needing to express their own by one of these replies, amused me.
If you meant me,
If that last part was aimed at me, yep, i amuse, annoy to whatever because i'm like that. Lol. seriously, like anything, there's going to be so many issues with the first car coming out so I think we're all barking up trees.
Yes!!!
Ever since I was little, I dreamed of having a car that a blind person could drive. So yes!!! If it becomes possible, I will find a way to get a self-driving car.
Re: Self Driving Cars
In fact, there are already driverless vehicles on the road. The company "Cruise" (which is a subsidiary of GM) has been operating driverless robo vehicles in San Francisco for some time now. They have been operating as an on-demand taxi / mini-bus service from 10:30 pm to 5:30 am overnight as a test. The program has recently expanded to daytime service. The taxi holds several people but doesn't have a backup driver.
Yes, self driving vehicles will have their problems just as human driven vehicles do. But considering that about 40,000 people in the U.S. die each year from auto accidents, I'm thinking that we can do a whole lot better once this technology improves a bit. As someone said, 50 years from now we'll think that these were the dark ages when we allowed so many people to die because we needed cars!
Interestingly, when elevators first began being used in high rise buildings, they all had operators. People were afraid to get into an elevator without an operator when self-operated elevators came onto the scene.
Considering how many of my sighted friends drive these days, I'd have a lot more faith in an autonomous vehicle!
--Pete
Lost all interest
The only way I'll have a car of my own is if they have no driving controls, so this is a bust. I'm not surprised though. Oh well, guess we'll have to keep yelling at American lawmakers to invest in public transit
Apple car documentary!
Interesting article and thoughts. Fifty years from now, I'm sure there will be a documentary about the evolution of the Apple car and how it enabled blind people to travel more independently. Maybe a good title would be:
From double tap and hold to double tap and hold on!
Jim
Computing power and quantum
When Quantom computers comes and mature, it will be the bases of real AI and more. It suppose to be similar to the brain. Capable of doing so much compare to current AI and CPU. It will be so far advance that it will be able to control a car and also the trafic similar to how Air control manage airplanes.
Quantum computing is a long…
Quantum computing is a long way off and, according to recent reports, rather over blown with its promises. Multi core linked nodes with standard computing will do it just fine and a vast amount cheaper. Ocado, known for its supermarket offerings here in the UK, are actually a robotics company. They have huge warehouses full of robots on intersecting rails that travel up to ten metres per second. A fully integrated automated transport system would be no more challenging, especially as you'd over lay areas. Quantum computing may help with the learning part of the AI, but that certainly waits to be seen.
QC and autonomous vehicles
The view on quantum computing is moving to be that there are a number of areas where it can be dominant but that best case scenario will see quantum coprocessors used by conventional computers which still do the bulk of the work. The environmental challenges alone (some of the most common QC techniques require temperatures only a few milikelvin above absolute zero) suggest we'll not have QCs at home any time soon, possibly never. The impact QC has on machine learning tasks is still quite theoretical but again, applicability likely to be very focused on particular problems.
But if that sounds negative I'd not be at all surprised if we see level 4 or 5 autonomous vehicles this decade. The current progress by the leaders in the space has been astonishing. And as someone mentioned earlier the system doesn't need be perfect, it only needs be better than the average human driver to be a net win. Not a high bar.
I suspect regulatory hurdles will be harder. The reason Cruise can do what it does is that the vehicles work only in very well defined areas that have been fully mapped by LiDAR, they are effectively on rails. Getting government agencies to allow cars with no controls loose in any road environment is not something I'd want to be responsible for.
To bring this back to Apple though, I always thought the idea of them having an autonomous system in a few years was ludicrous. They are massively behind the market leaders here (Tesla being furthest ahead) in having actual data from cars on the road. This data asset is in my opinion what will really make the difference for companies trying to build autonomous systems and Apple is pretty much at the back of the pack. And with other manufacturers building out their fleets its not hard to see that being an unassailable gap over time.
We'll see - I want independent travel without relying on others as much as anyone here!
Re: QC and autonomous vehicles
Agreed.
Another thing that will make it easier for self driving cars in the future is when humans aren't driving their own cars unpredictably and the infrastructure for roads is better set up to accomodate self driving cars and provide better feedback.
--Pete
Really, as you say, to make…
Really, as you say, to make it work we need ubiquity. There needs to be a conformity to how vehicles are tracked, how they act, and every vehicle to be curated by a master mind, someone mentioned an air traffic controller earlier, which is a good analogue. For the highest possible level of safety and efficiency apple can't afford to have a system that is competing with other systems. The road ways will literally become virtual airspace/train tracks, optimised for distance between vehicles and speed.
I think there will be an era of competing systems, google, apple, tesla etc, but I also foresee a conflict of AI imperative, a kind of playing chicken. An AI can probably predict the reactions of a human operator, has the advantage of reaction time, but what happens if it is reacting to another AI's road decisions. We end up with a very rapid loop of autocorrections on either side.
I'm not sure if that is how it would work, and I'm certain at least one of you could clarify this for me, but it does seem that there needs to be some communication between platforms and yes, prioritisation too. It puts me in mind of that guy who managed to get siri, Alexa and okay google to talk to one another in a loop by setting a couple of simple requests.
I'd imagine, in the end, infrastructure such as cellular will be the key factor here. I do love the idea of a contained vehicle that can work things out itself, but that processing is incredibly inefficient. We want cars to learn from each other and talk to one another... Mainly so we don't have to.
When you put it that way...
It could only go horribly wrong. The AI vehicles will start fighting each other in the streets when they have traffic accidents, and this will probably become entertainment for us if we can get out of the way before the tires start flying. After all, if people are programming the AI, at least at its beginning, our behaviors will somehow sneak in and run a muck. Only then will we come up with an AI traffic control system, like with aircraft.
Beset with problems
I'm no technological expert so I will steer clear of those problems but there are plenty of others, such that although I am 41 at time of writing I do not expect to be able to drive a car during my lifetime - and no, that is not because I have had my life expectancy shortened by personal injuries caused by an automated vehicle, for example. Indeed, notwithstanding that I live the sort of life that might have been enjoyed by the man whose sobriquet I have adopted as my username, there's no reason to suppose that I'm anything less than halfway through, like the average jonny.
So, to these difficulties:
the problem with the air traffic control analogy is, I think, one of numbers. There are far fewer aircraft than there would be vehicles in your imaginary landscape. Air traffic control is supposed to keep aircraft roughly five miles apart, as I understand it, so in addition to the numbers there's a little more margin for error to play with, even allowing that aircraft travel faster. So I'm not sure that analogy entirely stacks up.
The second problem is one of tortious liability. I don't agree with Oliver that the question is whether fewer people will be run over. That might be the purely humanitarian question, but economics invariably trumps that. The crucial question will be: who pays for it if Siri runs someone over? Will it be the blindy behind the virtual wheel? that can't be right as how could it be said that the blindy was negligent? Would there have to be a new statutory regime of strict liability that would nevertheless hold the blindy liable? If so, then it's not worth getting one of these anyway as you might be saddled with having to pay an award of damages when you, put it bluntly, had done nothing wrong; and what if the person run over were killed? Could you be criminally liable for causing death by dangerous driving? Hardly fair, you might think, when it was Siri driving dangerously; but it's no more fair to allow a victim to go uncompensated. There would, at the very least, have to be a radical overhaul of driving offences. You couldn't abolish them altogether but nor could you keep them as they are. That would take a huge amount of time, at the very least.
The third problem is conservatism, with a small C. Even were the technology absolutely right, even if we'd overcome the various legal issues, there would still be that doubt about allowing blindies to be behind the virtual wheel. The 'what if something goes wrong' argument would be king for a very, very long time. Your aircraft analogy wouldn't help you out here - even Ryan Air flights have pilots!
As for train strikes, Oliver, they'll still go on strike even in the era of fully automated vehicles. They would do better to have a close look at the report into trade union overspending published yesterday. Read A Very British Coup, which is an excellent novel and available on Audible. This has a few very wise things to say about public sector strikes.
Final point, and this is just in answer to the noise issue: I asked Phil Hammond, otherwise known as Spreadsheet Phil, this very question when he was Secretary of state for Transport: in the UK, there are now regulations as to noise levels required as a minimum from electric vehicles. Perhaps that minumum is too low but of all the problems we've highlighted I think the noise one is actually one of the easier ones to solve.
Too Good to Be True
I thought the promise of an autonomous driving car sounded too good to be true. Now we know it is. I moved to a big city soon after being blind with a transit system that operates 24 hours a day, seven days a week. My neighborhood is highly walkable so I can just walk to many places without needing a car or even using public transit. For this privilege, I pay some of the highest taxes in America and endure high levels of crime. Maybe in another lifetime, I can live in the country and drive into the city.
Possibly, if you're a youngster
Kelly, if you're a youngster then it is a possibility that you will have this dream fulfilled. AS i have said above, I don't see that this is going to happen for someone of my vintage i.e. in their 40s. Just to draw upon some valuable historical context - and this relates directly to my conservatism argument above: the motorcar was introduced to Britain in 1896 but was subject to very strict coonditions. if you owned a motorcar you had to assign one of your servants the duty of walking ahead of you on the road bearing a flag to alert coachmen on the highway of your approach. This naturally meant your chauffeur's having to drive rather slowly - he could not overtake the flag man in front. Such conditions were not altered until sometime after the passing of Her late majesty Queen Victoria - they were gone by 1907, that much I do know..but anyway, the precise date is unimportant for the point I seek to make. We would, I suggest, need to undergo the equivalent of the flag-bearing period before anything like this were allowed. Now, throw another factor into the mix: in Cambridge, England, there has recently been a trial of robots delivering groceries to the residents of that noble university town - although oxford is better. How have said robots been getting on? Well, not too bad except they've not managed certain aspects very well. They haven't known how to deal with children fiddling with them or cats investigating them. As for pedestrian crossings, there have been mixed results. Some robots have appeared rather shy and have waited until they can cross the road in company with several of their colleagues. Others have not been able to reach the button and have had to ask for help from their human inferiors - "Will you please press the traffic button? thank you." the upshot is they've been viewed as even more eccentric than some of the dons of Cambridge colleges who are normaly never surpassed in eccentricity save, perhaps, by some of the dons of Oxford colleges. In short, if our delivery robots still need to ask for help from a human being and are put off by kids and cats, we're a long way short of blindies being let loose on the roads and, I'm afraid, even a long way short of the man with the flag stage. This is good news for Bingo, of course, who need not worry about taking his valet away from his usual duties.
100
By 100 it will be ready. All depend on whether we have not made a big booboo of the planet. Need the structure such as control of trafic by AI, and a powerful AI that can do much better than the one today. We are not going to be around for it. Sad but true.
100?
Holger, the year 100? You must be using a calendar with which I am not familiar. The Gregorian Calendar has this year some 1,922 years ago, likewise the Julian. For the Masonic or Roman calendars it's more than 2,500 years ago. if the calendar of revolutionery France were still in use I believe we would now be in the year 230 or thereabouts. It's a long time since Bingo attended his lectures on le droit constitutionel de la France at l'Universite Pantheon-Assas, paris, and he can't recall whether the revolutionery calendar began on the death of Louis XVI or whether it started four years earlier with the commencement of the revolution itself. Either way, we're in 230 or 234 or thereabouts...demonstrably the year 100 is in the revolutioneries' past.
Which calendar, I ask, still has the year 100 sometime in the future? Or perhaps you have started one of your own? If you have, bravo - bingo admires the spirit of adventure, so long as his valet doesn't follow your example.
I guess, in the time of ride…
I guess, in the time of ride hailing apps such as uber, we do need to consider what we would actually gain from a self driving vehicle. If the goal is to get from point A to point B in safety and with relative alacrity, then, for the most part, this is satisfied. Having a car we can summon to us, in which we can be transported to a destination and dropped off, without any human assistance, EG taking us to the door of the building, then, even if everything works correctly with self driving technology, we gain nothing. WE, in fact, lose the key aspect of human interaction, the ability to ask for help.
Bingo Little
100 years from now. Even Bella the cat got it.
Creativity
I'm sorry, Holger. I adhere to the orthodoxy, that's my problem - the orthodoxy that dictates that there is a difference between the year 100 and 100 years hence. I have just ended an email to one of my line reports with: see you in 2023. Clearly, I need to ensure she does not understand this as meaning I will see her 2,023 years from now i.e. in the year 4045. Still, on the plus side it's good to see Bella the Cat again after her traumatic experiences in IOS 13 - and I don't mean IOS 13 versions into the future, which would be IOS29, I believe.
oliver, would you please press the traffic button?