News1 min ago
No Deaths, This Time........
32 Answers
http:// www.bbc .com/ne ws/tech nology- 4192381 4
Hit by a truck, yes the fault of the trucker but that's cold comfort if you are injured or worse. Note to the programmers, it is entirely possible that other vehicles on the road will not obey the rules.
Hit by a truck, yes the fault of the trucker but that's cold comfort if you are injured or worse. Note to the programmers, it is entirely possible that other vehicles on the road will not obey the rules.
Answers
Best Answer
No best answer has yet been selected by ToraToraTora. Once a best answer has been selected, it will be shown here.
For more on marking an answer as the "Best Answer", please visit our FAQ.It's not the first crash of its kind.
https:/ /www.th elocal. ch/2016 0923/sw iss-dri verless -bus-tr ial-sus pended- after-c rash
https:/
you cannot prove that woofgang but often drivers would avoid a collision even if they are in the right.
bhg - yes and if we could wave a magic wand and remove all but autonomous vehicles from the road and all the pedestrians and all the domestic pets and all the squirrels and deer etc etc then no doubt all the computers would get on. Sadly that's an impossible dream.
bhg - yes and if we could wave a magic wand and remove all but autonomous vehicles from the road and all the pedestrians and all the domestic pets and all the squirrels and deer etc etc then no doubt all the computers would get on. Sadly that's an impossible dream.
The big difference between people and computers is that computers can learn from mistakes whereas people are often too stupid to do so. (Nobody uses a phone whilst driving now, do they, because they know it's dangerous). The computer controlling the damaged vehicle can have its program updated and the mods sent to all similar vehicles automatically; people continue to drink-drive, text etc even after they've been punished for doing it.
do you have any training in IT BHG? computers have no learning ability, contrary to popular belief AI does not yet exist. What you have in these vehicles is a very good set of rules programmed to cover most situations with zero creativity. That would be fine if all the vehicles where thus, like the DLR, a closed system totally automated. No doubt their programming will be continually improved but basically they are following a list of instructions at all times. Humans, whilst flawed, have the ability to improvise and adapt, computers do not. These vehicles will make it onto the road for a while until carnage ensues and they will be withdrawn. Yes statistically no doubt less people will die on the roads but that wont matter, when a person is killed by an autonomous vehicle the public outcry will demand action regardless of cold statistics.
I programmed my first computer in 1963 TTT - a Ferranti Pegasus, just like the one in the Science Museum. Since then I have worked on everything from PCs to machines in the Top 50 fastest in the world. I have been a user of other peoples' programs, a writer of programs for other people's use, a system administrator and the manager of a network system in operation before the internet existed. In all the cases, if a problem occurred the fault was traced, corrected and the revised program put out for use. In the case of vehicles an accident will be analysed, the controlling program fixed and put out to all the other vehicles using that code. Admittedly, the computer can't teach itself but then, neither can a lot of people.
so what was all this about then? "The big difference between people and computers is that computers can learn from mistakes whereas people are often too stupid to do so. " - you know computers can't learn! so was this some attempt at hyperbole?
"In the case of vehicles an accident will be analysed, the controlling program fixed " - what if it can't be "fixed"? there are no win situations out there all the time. How would you "fix" this situation? A human might back off if possible or maybe look in the mirror to see if we can pull out of the way, the computer just stopped and waited to be hit.
"In the case of vehicles an accident will be analysed, the controlling program fixed " - what if it can't be "fixed"? there are no win situations out there all the time. How would you "fix" this situation? A human might back off if possible or maybe look in the mirror to see if we can pull out of the way, the computer just stopped and waited to be hit.
If a person can look in the mirror or back-off, a computer can be taught to do that and remember that a computer can look backwards, forwards and sideways at the same time and make its decisions much more quickly. A computer drives with the cumulative experience of thousands of drivers; a person just his own experience and bits that he's remembered from being taught.
"if a person can look in the mirror or back-off, a computer can be taught to do that and remember that a computer can look backwards, forwards and sideways at the same time and make its decisions much more quickly." - agreed cleary this vehicle needs more programming.
"A computer drives with the cumulative experience of thousands of drivers;" err no, it drives with the defined rules of driving programmed by those that defined them to it.
" a person just his own experience and bits that he's remembered from being taught. " - and conscious creativity
"A computer drives with the cumulative experience of thousands of drivers;" err no, it drives with the defined rules of driving programmed by those that defined them to it.
" a person just his own experience and bits that he's remembered from being taught. " - and conscious creativity
naomi "^Oo-er! Is that really a rational question?"
Well I think so...."the shuttle did what it was supposed to do and stopped. Unfortunately the human element, the driver of the truck, didn’t stop.”
There is nothing there to suggest that a human shuttle driver would have done anything different....or do human drivers never ever hit stationary vehicles?
Well I think so...."the shuttle did what it was supposed to do and stopped. Unfortunately the human element, the driver of the truck, didn’t stop.”
There is nothing there to suggest that a human shuttle driver would have done anything different....or do human drivers never ever hit stationary vehicles?
https:/ /en.wik ipedia. org/wik i/Qanta s_Fligh t_32
can anyone tell me how a computer could have thought its way out of that? and land the plane with no injuries to its occupants?
can anyone tell me how a computer could have thought its way out of that? and land the plane with no injuries to its occupants?
bit of leap from cars to planes mush but the same principle applies. It could no doubt be programmed to return to an airfield or find one asap but would you get on an autonomous plane? Gawd we can't get cars right!
Oddly enough I watched "Sully" the other night and the simulator indicated a course of action would have killed the lot. Primarily because it did not have the human insight that the pilot did.
Oddly enough I watched "Sully" the other night and the simulator indicated a course of action would have killed the lot. Primarily because it did not have the human insight that the pilot did.
AI of the sort needed to drive cars may not be so far away. I think people have a tendency to overestimate their own abilities and underestimate computers'.
But we'll see. In the long run it seems pretty clear that self-driving cars don't need to be "perfect" to be a vast improvement on human drivers.
But we'll see. In the long run it seems pretty clear that self-driving cars don't need to be "perfect" to be a vast improvement on human drivers.