PDK vs Manual
#106
Three Wheelin'
^To add to it, one of the biggest advantages is if one car's AI learns about some traffic situation/pot hole and how to handle it, every other car can potentially learn the same almost instantaneously.
Udacity has a very interesting (and free) introductory course on the AI behind self driving cars by Peter Thrun (who was in charge of Google's self driving car program). Highly recommend it if anyone is curious.
https://www.udacity.com/course/artif...obotics--cs373
Also more generic courses on AI and Machine learning -
https://www.udacity.com/course/intro...ligence--cs271
https://www.udacity.com/course/intro...earning--ud120
Udacity has a very interesting (and free) introductory course on the AI behind self driving cars by Peter Thrun (who was in charge of Google's self driving car program). Highly recommend it if anyone is curious.
https://www.udacity.com/course/artif...obotics--cs373
Also more generic courses on AI and Machine learning -
https://www.udacity.com/course/intro...ligence--cs271
https://www.udacity.com/course/intro...earning--ud120
#107
Most probably many of us will also use these driverless cars for daily commute. That will be comfortable, maybe too comfortable.
But I know, I will crave for driving myself and analog, but also with some of the technologies wonder to compensate my driving skill deficit. Hence, for me it might be the case that I keep GT4 for a very very long time.
Tried to get a slot for gt3 RS next to my gt4. Unfortunately none available in Germany although we are a key account for PAG. I was too slow.
But I know, I will crave for driving myself and analog, but also with some of the technologies wonder to compensate my driving skill deficit. Hence, for me it might be the case that I keep GT4 for a very very long time.
Tried to get a slot for gt3 RS next to my gt4. Unfortunately none available in Germany although we are a key account for PAG. I was too slow.
#108
There is a huge difference between active cruise/collision avoidance/lane departure warning and a truly autonomous vehicle, that can operate anywhere, safely and efficiently. We are still a long way off from the latter being a reality, even longer before it becomes pervasive, and longer than you or I have on this planet before it replaces manually operated cars entirely. I'm quite certain we will all be happily driving our sports cars until the folks at Sunnyside Acres wrestle the keys out of our hands. There are so many hurdles still to cross and such enormous cost challenges on top of that.
Good thing there are a lot of fun older cars that will satisfy the cravings of enthusiasts who were imprinted by certain characteristic when they became enthusiasts. The prices will go up, but at least you can still get them, or build them.
#109
oh boy! i cant wait till we're all standing around the track, watching driver less cars, and doing what?
Placing bets! things that make you go hmmmm...
im glad ill be dead when this takes place... i already have a hard enough time playing a rigged game trying to beat algorythyms and hft's in the current environment on a daily basis...
as for me personally driving my own car at my free will in the future.... Molon Labe!
Placing bets! things that make you go hmmmm...
im glad ill be dead when this takes place... i already have a hard enough time playing a rigged game trying to beat algorythyms and hft's in the current environment on a daily basis...
as for me personally driving my own car at my free will in the future.... Molon Labe!
#110
^To add to it, one of the biggest advantages is if one car's AI learns about some traffic situation/pot hole and how to handle it, every other car can potentially learn the same almost instantaneously.
Udacity has a very interesting (and free) introductory course on the AI behind self driving cars by Peter Thrun (who was in charge of Google's self driving car program). Highly recommend it if anyone is curious.
https://www.udacity.com/course/artif...obotics--cs373
Also more generic courses on AI and Machine learning -
https://www.udacity.com/course/intro...ligence--cs271
https://www.udacity.com/course/intro...earning--ud120
Udacity has a very interesting (and free) introductory course on the AI behind self driving cars by Peter Thrun (who was in charge of Google's self driving car program). Highly recommend it if anyone is curious.
https://www.udacity.com/course/artif...obotics--cs373
Also more generic courses on AI and Machine learning -
https://www.udacity.com/course/intro...ligence--cs271
https://www.udacity.com/course/intro...earning--ud120
#111
Oh, I know it's coming. Just not as soon as most people think, at least in anything but small, mostly test numbers. And I also know that in no way will this threaten our ability to drive manually operated cars for the rest of our natural lives, so I all the hand wringing over this is foolish. The remaining technological, regulatory and, especially, financial hurdles, are enormous and will take a very long time to fully play out. Maybe 20 years from now we'll actually see a small number of fully autonomous vehicles, likely smaller cars, being sold and driven in meaningful numbers. But it'll be another 50+ years beyond that before they're a majority of the vehicles on the road. Even if the technology was absolutely perfected today, it would take decades and decades to fully turn over the vehicles on the road, purely for financial reasons. The vehicle cost is going to be staggering and it's likely to be linked to the alternative fuel movement, which has its own massive infrastructure costs and challenges.
This discussion reminds me of the people who argued 25 years ago that we'd all be driving electric vehicles by the year 2000.
Our kids kids may never own a manually operated car. But not us.
This discussion reminds me of the people who argued 25 years ago that we'd all be driving electric vehicles by the year 2000.
Our kids kids may never own a manually operated car. But not us.
#112
As an engineer and Chief Technology Officer of a tech company a stone's throw away from Tesla, count me among those who take the opposing view. Nissan's CEO has stated 2020 for affordable self driving cars, Mercedes said the same. Tesla is saying sooner, and current Teslas are built with the sensor suite needed- they just need software and redundancy. You can buy a long haul truck today that will drive itself on the freeway.
It's not a yes/ no proposition, instead the change will be incremental, and liability will shift in stages not instantly. However more miles will be driven in the U.S. by computers than humans in far less than 20 years. It's a common error to extrapolate past technical progress linearly into the future. It's not linear- it's exponential. Darpa's grand challenge was less than a decade ago. Today I'm dodging driverless cars daily.
Just my $.02.
It's not a yes/ no proposition, instead the change will be incremental, and liability will shift in stages not instantly. However more miles will be driven in the U.S. by computers than humans in far less than 20 years. It's a common error to extrapolate past technical progress linearly into the future. It's not linear- it's exponential. Darpa's grand challenge was less than a decade ago. Today I'm dodging driverless cars daily.
Just my $.02.
First, extrapolate forward to the point when this technology is truly fully deployable across all the major manufacturers product lines and all of the regulatory bodies, federal and state, have signed off on the safety of the vehicles. (Have the feds and the states even developed the testing methodology they're going to use to test the safety of these vehicles yet? What about light trucks that have to be safe in dirt/off road situations? What about inclement weather/snow testing?)
Now, say we're at that point, which I'd wager serious cash is at least 10+ years away. Given the fact that the average age of cars/light trucks on the road in the U.S. are 11.5 years old, with some much younger and some much older, it would be safe to assume that it would take over 10 years to even turn over the majority of the cars and light trucks, and that's assuming buyers were even willing to buy these cars in droves without being forced to do so through legislation, which I highly doubt they will. And on top of that, figure that all this technology costs money, so the cars will likely be a fair bit more expensive than current cars, which will make it even harder.
It is going to take a generation for this conversion to be completed.
BTW, explain to me how a light truck that needs to be used on a farm could ever really be fully autonomous, when it has to drive all over in random directions at the whim of the operator, into tight spots, pick up loads, back up a hay picker, cross a creek, etc. There are a thousand different situations that I could think of that are going to be a huge challenge for fully autonomous vehicles that have not been solved yet.
#115
First, extrapolate forward to the point when this technology is truly fully deployable across all the major manufacturers product lines and all of the regulatory bodies, federal and state, have signed off on the safety of the vehicles... Now, say we're at that point, which I'd wager serious cash is at least 10+ years away. Given the fact that the average age of cars/light trucks on the road in the U.S. are 11.5 years old, with some much younger and some much older, it would be safe to assume that it would take over 10 years to even turn over the majority of the cars and light trucks
You think people don't want to surf the web on their way to work? People are already buying this technology in droves, whether they know it or not.
Last edited by Petevb; 09-03-2015 at 01:48 PM.
#116
Rennlist Member
PDK vs Manual
People won't necessarily need to purchase one. A person will hail a driverless Cab from his or her smart phone. The car will buzz the phone as it approaches, and the owner will get inside the car and go where it's directed.
This becomes a more attractive option as the younger generation cares less and less about cars. Also, as all of these safety and autonomy features become more prevalent, the cost of owning and insuring a vehicle becomes less manageable, so there's more of an incentive to simply pay by the ride. We'll be buying these cars de facto.
This becomes a more attractive option as the younger generation cares less and less about cars. Also, as all of these safety and autonomy features become more prevalent, the cost of owning and insuring a vehicle becomes less manageable, so there's more of an incentive to simply pay by the ride. We'll be buying these cars de facto.
#117
PDK is faster, period. But less involving, and attractive for some. What else is there to say?
The "market value" of PDK vs Manual will be interesting, I don't think anyone can predict that. The only fact there is to consider is that a large amount of the demand for the GT4 is because it's a manual. Though certainly there is probably more demand for it if it had a PDK.
The "market value" of PDK vs Manual will be interesting, I don't think anyone can predict that. The only fact there is to consider is that a large amount of the demand for the GT4 is because it's a manual. Though certainly there is probably more demand for it if it had a PDK.
#118
[QUOTE=Petevb;12561931]You're assuming a linear timeline. In reality these things are happening in parallel. The sensors and hardware needed for self-driving cars are already being deployed in the form of lane departure, collision avoidance, blindspot detection, etc. Which is why Tesla only pushed a software update to its Beta testers to enable those cars to drive themselves. Within five years the majority of cars sold will come with similar technologies, some of them no doubt mandated by DOT for safety reasons. So when regulations change new routes and areas where the cars can legally drive themselves will simply be unlocked. The reg changes won't come at the start of the tech deployment cycle, but in the middle to end.
QUOTE]
No, I'm not assuming a linear timeline. There will certainly be things going on in parallel. It's just that all phases will take more time than you're assuming, there will be some dependencies, and there are likely to be numerous points of setback. And I'm fully aware that various technologies and elements of the system will be released over time. My point it that, truly fully autonomous vehicles across the spectrum of vehicles will not be on our roads in heavy numbers for a long time. And manually operated vehicles will not be gone or legislated out of existence for a lot longer still. We, the people in this forum who probably average about 50 years of age or older, have nothing to worry about. Nobody's going to take our right to drive these cars away. Nor likely do our children have to worry about it either.
On top of all the technical hurdles, at current vehicle costs, it would take about $9 trillion to replace every passenger vehicle in the U.S. That's gonna take some time.
QUOTE]
No, I'm not assuming a linear timeline. There will certainly be things going on in parallel. It's just that all phases will take more time than you're assuming, there will be some dependencies, and there are likely to be numerous points of setback. And I'm fully aware that various technologies and elements of the system will be released over time. My point it that, truly fully autonomous vehicles across the spectrum of vehicles will not be on our roads in heavy numbers for a long time. And manually operated vehicles will not be gone or legislated out of existence for a lot longer still. We, the people in this forum who probably average about 50 years of age or older, have nothing to worry about. Nobody's going to take our right to drive these cars away. Nor likely do our children have to worry about it either.
On top of all the technical hurdles, at current vehicle costs, it would take about $9 trillion to replace every passenger vehicle in the U.S. That's gonna take some time.
#119
People won't necessarily need to purchase one. A person will hail a driverless Cab from his or her smart phone. The car will buzz the phone as it approaches, and the owner will get inside the car and go where it's directed.
This becomes a more attractive option as the younger generation cares less and less about cars. Also, as all of these safety and autonomy features become more prevalent, the cost of owning and insuring a vehicle becomes less manageable, so there's more of an incentive to simply pay by the ride. We'll be buying these cars de facto.
This becomes a more attractive option as the younger generation cares less and less about cars. Also, as all of these safety and autonomy features become more prevalent, the cost of owning and insuring a vehicle becomes less manageable, so there's more of an incentive to simply pay by the ride. We'll be buying these cars de facto.
Let's revisit this thread in 10 years time. It's very similar to whole 'EVs will be everywhere within a few years' argument that I've been hearing most of my adult life and yet they're still today a drop in the bucket and probably still will be 10 years from now. Same arguments were being made; the technology's getting better every day, it's an exponential thing, people will want these cars, etc.
#120
Adoption cycles are exceedingly short when disruptive and enabling technology comes along. Remember flip phones? The first iPhone was released eight years ago. How many phones have you gone through since then? And that's just a phone, it doesn't give you 5% more useful time in your day, or measurably increase your life expectancy.
I'll say more miles will be driven autonomously in the US than by humans within 15 years, so before 2030. We can come back then and see who's right.