Fair enough. regarding hacking....I'm much more comfortable with Google doing this since they are keenly aware of security aspects. Car companies? Not at all.
Heh, yea... remember when they first started putting Bluetooth in cars? Every single one had the PIN defaulted to either 0000 or 1111, and most of them can't be changed. That's the sort of brilliance to be expected from the auto makers.
And frankly I wouldn't want them to be security experts - I just want their systems to be certified by security experts. Likely there needs to be some standard/regulation/testing levels established so that the car companies have requirements they meet to sell such cars.
If done right, yea. But I think we can both figure out how it would actually go down - a bunch of industry insiders, like former GM CEO's or whatever, end up being appointed to the regulating committee, and nothing positive happens.
The emergency situation will be handled by the system. In fact, that will be the greatest reason for increased driving safety. The.5% is for places like parking garages, dirt trails, and other edge cases.
To which the human operator responds, "WTF? I don't know how to drive, this things supposed to do everything for me!"
The solution is simple once self driving cars are common, the law will change to make them not liable, unless they fail to maintain the vehicle properly.
Yea, kinda like how copyright law changed with the advent of digital media.
Oh, wait, that didn't work out for the consumer at all...
Google's cars are already interacting with real world drivers quite well.
Where? I'm curious as to what sort of conditions they're doing the trials in.
I.e., are the tests limited to the city roads around Google's office? How many interstate highways have they tested on, during what hours, and for how many miles?
I forget which manufacturer it is, but they've begun to equip cars with IR sensors that can identify people, deer, etc at a much greater distance than the car's headlights penetrate.
Cadillac has been advertising such a feature for some time now, though I have no idea how well it works.
I tried to think of some more scenarios, but to be honest most of them would be just as treacherous and difficult to handle for human drivers (chunk of asphalt suddenly flying towards the windshield, for example).
So, I guess my real problem with "predictive driving," which is another way of saying what an auto-car does, is this: drivers and the environment are wholly unpredictable, and there's absolutely no way to prepare for everything. Then again, considering how most people consider driving some kind of wasteful chore unworthy of their complete attention, I do support the notion of putting the control of their cars into the computer's digits.
Hacking is another issue; the notion that some punk-ass in the next car over could, somehow, take control of my vehicle, is a pretty damn terrifying concept.
OK, Tip #1? Don't use Youtube as a reference, unless you like being wrong a lot.
I haven't seen the video you posted, A) because the firewall blocks it, and B) a simple Google search of, say, "bird through windshield" offers no shortage of evidence that you and whoever you got your info from is incorrect.
Only if you ignore how hard it is to wind up with a lower total death toll if your method starts by killing a sizable portion of the population. Notice I made no distinction between the two sides, merely noting that while a more effective weapon will be more devastating, it may well reduce the overall effect of a war. Ignoring very possible human mistakes, of course.
Mistakes nothing, you've ignored a very important, ever-present aspect of war - the people who wage them. Or, more importantly, their ideologies.
So long as the people in charge feel that murder is good policy, making weapons more destructive will lead to greater loss of life. Just look at drone usage in Pakistan if you need a real-world example - entire villages, wiped off the face of the earth, because there was supposedly 1 guy in the entire group worth killing.
PS: If you didn't get the memo, genocide is a word frowned upon. Nowadays, we prefer the much more eloquent term of collateral damage.
Two terms that mean different things; Political correctness can go suck turds out of my ass.
I would think the point would be that machines, once properly programmed, can be the worlds safest drivers...statistically. You, as a human, will still be responsible for taking over when the machine doesn't know what to do. But, for the other 99.5% of the time, the self-driving car will make the best decisions and always be completely alert.
The problem with that is, how much notice do you think a computer is going to give the operator when it comes across one of those situations it doesn't know how to handle? 5 minutes, probably not; 5 seconds*, maybe. That's not a lot of time to switch gears from "casually reading a book" to "OMFGABIRDISCOMINGTHROUGHTHEWINDSHIELD!"
Maybe in America. There are other areas on the world where you can get soldiers as young as 6. They may not be as efficient, but they are cheaper than robots.
Oh, how I'd love for you to present said evidence, that "proves" people like my brother are mindless killing machines that do everything the government "programs" them to do...
The irony being, of course, that you just described a robot, rather than a human.
That is a serious strawman and if you think I said anything like that at all, you either lack reading comprehension or are just looking for a fight.
Your statement, verbatim:
All evidence that I've seen shows that a large number - possibly even the majority - of soldiers have been brainwashed in to following orders unconditionally and will commit the most horrendous crimes against humanity when ordered to do so.
If there's a strawman here, it's one you built.
I've elected to ignore the rest of your post as it's obvious you don't even know what you are saying, let alone what I am.
Seems like a waste of resources to me. I mean, if war is going to become robot v robot, why bother with physical assaults to begin with? Rather, have both sides agree to be bound by the result of a couple rounds of Call of Duty.
A pleasant idea, yes, but wholly unlikely considering the mentalities of the people who tend to rise to power.
This above scenario is quite doable with modest amounts of your own code - most has already been written. I would estimate a minimum cost for the above device - a device that holds say a SKS with a 75 round drum and can shoot an aimed round once a second at targets close to each other - at around $300.
You might be surprised; after all, using chemical weapons on unarmed civilians is a violation of the Geneva Conventions, yet the US government does it all. The. Time.
Yea, but they CANT destroy the US. It's not possible. It's like we live in a mansion and a rat ran in and shit on our floor. So now we have the entire staff chopping up the floorboards and taring the plaster off the walls looking for the fucking thing. We're doing far more damage than the stupid rat ever could. Some pests just don't go away, so you have to keep the cheese in the fridge, put out some traps and deal with it. Don't burn the house down around you just to win.
So... they actually can destroy the US, just not directly.
With cruise missiles, you can just fire away. Directly, it causes more deaths; but all together, it can be quite a bit less gruesome. It's always the application - not the weapon itself - that makes the distinction.
Ironically, the same statement can be made to justify genocide - if you completely wipe out the "other guys," there's no chance they'll rise up later and cause deaths to your guys, right?
because all evidence shows that the weak point always lies with the soldier that has to pull the trigger and decide to kill a fellow human being.
All evidence that I've seen shows that a large number - possibly even the majority - of soldiers have been brainwashed in to following orders unconditionally and will commit the most horrendous crimes against humanity when ordered to do so.
Oh, how I'd love for you to present said evidence, that "proves" people like my brother are mindless killing machines that do everything the government "programs" them to do...
The irony being, of course, that you just described a robot, rather than a human.
Also, philosophically speaking, I'd say a human's decision making is just a really complex set of algorithms that we don't understand particularly well at this point. What we do know is that humans make significant mistakes with regularity, so the test isn't whether or not these autonomous systems make mistakes in difficult circumstances, but rather the ratio of mistakes compared to human agent.
Right, humans fuck up.
And these killbots would be programmed and controlled by humans.
Therefore, it stands to reason that killbots would do everything their fuck-up human masters tell them, except without the compassion, remorse, shame, and other emotions that prevent many humans from doing fucked up things.
If a general tells a soldier, "go murder your entire family," said soldier will likely not follow that order. A robot, conversely, would always do what its master tells it, regardless of whether the master says, "go pick some daisies," or "go commit genocide."
Slashdot Mirror
Fair enough. regarding hacking....I'm much more comfortable with Google doing this since they are keenly aware of security aspects. Car companies? Not at all.
Heh, yea... remember when they first started putting Bluetooth in cars? Every single one had the PIN defaulted to either 0000 or 1111, and most of them can't be changed. That's the sort of brilliance to be expected from the auto makers.
And frankly I wouldn't want them to be security experts - I just want their systems to be certified by security experts. Likely there needs to be some standard/regulation/testing levels established so that the car companies have requirements they meet to sell such cars.
If done right, yea. But I think we can both figure out how it would actually go down - a bunch of industry insiders, like former GM CEO's or whatever, end up being appointed to the regulating committee, and nothing positive happens.
Just like the FCC.
The emergency situation will be handled by the system. In fact, that will be the greatest reason for increased driving safety. .5% is for places like parking garages, dirt trails, and other edge cases.
The
To which the human operator responds, "WTF? I don't know how to drive, this things supposed to do everything for me!"
The solution is simple once self driving cars are common, the law will change to make them not liable, unless they fail to maintain the vehicle properly.
Yea, kinda like how copyright law changed with the advent of digital media.
Oh, wait, that didn't work out for the consumer at all...
Google's cars are already interacting with real world drivers quite well.
Where? I'm curious as to what sort of conditions they're doing the trials in.
I.e., are the tests limited to the city roads around Google's office? How many interstate highways have they tested on, during what hours, and for how many miles?
I forget which manufacturer it is, but they've begun to equip cars with IR sensors that can identify people, deer, etc at a much greater distance than the car's headlights penetrate.
Cadillac has been advertising such a feature for some time now, though I have no idea how well it works.
I tried to think of some more scenarios, but to be honest most of them would be just as treacherous and difficult to handle for human drivers (chunk of asphalt suddenly flying towards the windshield, for example).
So, I guess my real problem with "predictive driving," which is another way of saying what an auto-car does, is this: drivers and the environment are wholly unpredictable, and there's absolutely no way to prepare for everything. Then again, considering how most people consider driving some kind of wasteful chore unworthy of their complete attention, I do support the notion of putting the control of their cars into the computer's digits.
Hacking is another issue; the notion that some punk-ass in the next car over could, somehow, take control of my vehicle, is a pretty damn terrifying concept.
A bird won't penetrate a windshield, no driver intervention would be needed:
https://www.youtube.com/watch?feature=player_detailpage&v=V-xoDBEO1EI#t=38
OK, Tip #1? Don't use Youtube as a reference, unless you like being wrong a lot.
I haven't seen the video you posted, A) because the firewall blocks it, and B) a simple Google search of, say, "bird through windshield" offers no shortage of evidence that you and whoever you got your info from is incorrect.
Only if you ignore how hard it is to wind up with a lower total death toll if your method starts by killing a sizable portion of the population. Notice I made no distinction between the two sides, merely noting that while a more effective weapon will be more devastating, it may well reduce the overall effect of a war. Ignoring very possible human mistakes, of course.
Mistakes nothing, you've ignored a very important, ever-present aspect of war - the people who wage them. Or, more importantly, their ideologies.
So long as the people in charge feel that murder is good policy, making weapons more destructive will lead to greater loss of life. Just look at drone usage in Pakistan if you need a real-world example - entire villages, wiped off the face of the earth, because there was supposedly 1 guy in the entire group worth killing.
PS: If you didn't get the memo, genocide is a word frowned upon. Nowadays, we prefer the much more eloquent term of collateral damage.
Two terms that mean different things; Political correctness can go suck turds out of my ass.
I would think the point would be that machines, once properly programmed, can be the worlds safest drivers...statistically. You, as a human, will still be responsible for taking over when the machine doesn't know what to do. But, for the other 99.5% of the time, the self-driving car will make the best decisions and always be completely alert.
The problem with that is, how much notice do you think a computer is going to give the operator when it comes across one of those situations it doesn't know how to handle? 5 minutes, probably not; 5 seconds*, maybe. That's not a lot of time to switch gears from "casually reading a book" to "OMFGABIRDISCOMINGTHROUGHTHEWINDSHIELD!"
* I'm probably being quite generous.
Maybe in America. There are other areas on the world where you can get soldiers as young as 6. They may not be as efficient, but they are cheaper than robots.
Still takes at least 6 years to grow 'em.
Oh, how I'd love for you to present said evidence, that "proves" people like my brother are mindless killing machines that do everything the government "programs" them to do...
The irony being, of course, that you just described a robot, rather than a human.
That is a serious strawman and if you think I said anything like that at all, you either lack reading comprehension or are just looking for a fight.
Your statement, verbatim:
All evidence that I've seen shows that a large number - possibly even the majority - of soldiers have been brainwashed in to following orders unconditionally and will commit the most horrendous crimes against humanity when ordered to do so.
If there's a strawman here, it's one you built.
I've elected to ignore the rest of your post as it's obvious you don't even know what you are saying, let alone what I am.
Robots cost lots of money and take time to build. People less so.
People take about 18 years and millions of dollars to 'build' into soldiers.
Seems like a waste of resources to me. I mean, if war is going to become robot v robot, why bother with physical assaults to begin with? Rather, have both sides agree to be bound by the result of a couple rounds of Call of Duty.
A pleasant idea, yes, but wholly unlikely considering the mentalities of the people who tend to rise to power.
This could easily be the case as they would not be so affected by emotions, fear, revenge...
compassion, remorse, empathy....
I wonder if the Geneva conventions were being drawn up now, the American federal government would boycott them?
FTFY. The vast majority of Americans have absolutely no say in the matter.
BTW, the answer is 'no, they'd go ahead and sign, then ignore the parts they don't like, the way they do now.'
On the bright side, algorithm-driven machines are unlikely to pull their guns just because they have an attitude problem like some cops do.
It also wouldn't have to worry about any of those pesky emotions like compassion or remorse slowing down its murder spree.
You overestimate the development required.
This above scenario is quite doable with modest amounts of your own code - most has already been written.
I would estimate a minimum cost for the above device - a device that holds say a SKS with a 75 round drum and can shoot an aimed round once a second at targets close to each other - at around $300.
Indeed
I suppose that the argument can be made that they fill the same role as land mines,
Well, we're fucked in that case
You might be surprised; after all, using chemical weapons on unarmed civilians is a violation of the Geneva Conventions, yet the US government does it all.
The.
Time.
Yea, but they CANT destroy the US. It's not possible. It's like we live in a mansion and a rat ran in and shit on our floor. So now we have the entire staff chopping up the floorboards and taring the plaster off the walls looking for the fucking thing. We're doing far more damage than the stupid rat ever could. Some pests just don't go away, so you have to keep the cheese in the fridge, put out some traps and deal with it. Don't burn the house down around you just to win.
So... they actually can destroy the US, just not directly.
How about: Murder of innocent citizens.
95% of them aren't americans (me included). Why would the distinction be important?
Americans don't seem to think that non-Americans are people, therefore not deserving of rights
You don't actually know any Americans, do you?
With cruise missiles, you can just fire away. Directly, it causes more deaths; but all together, it can be quite a bit less gruesome. It's always the application - not the weapon itself - that makes the distinction.
Ironically, the same statement can be made to justify genocide - if you completely wipe out the "other guys," there's no chance they'll rise up later and cause deaths to your guys, right?
Except looking at history, they will probable lead to fewer soldier deaths, fewer bystander deaths, more accurate targeting.
I don't know why people think they are bad.
Really? You can't see where an emotionless machine programmed to kill whatever its masters tell it to could be a bad thing?
I find that philosophy kinda messed up and poorly thought out, to be honest.
Failsafe system will be contracted out to the people who profited by writing and then fixing the Affordable Healthcare websites.
So the Canadians will be responsible for SkyNet?
That actually kinda makes sense.
The road to Hell is paved with good intentions, after all.
because all evidence shows that the weak point always lies with the soldier that has to pull the trigger and decide to kill a fellow human being.
All evidence that I've seen shows that a large number - possibly even the majority - of soldiers have been brainwashed in to following orders unconditionally and will commit the most horrendous crimes against humanity when ordered to do so.
Oh, how I'd love for you to present said evidence, that "proves" people like my brother are mindless killing machines that do everything the government "programs" them to do...
The irony being, of course, that you just described a robot, rather than a human.
Also, philosophically speaking, I'd say a human's decision making is just a really complex set of algorithms that we don't understand particularly well at this point. What we do know is that humans make significant mistakes with regularity, so the test isn't whether or not these autonomous systems make mistakes in difficult circumstances, but rather the ratio of mistakes compared to human agent.
Right, humans fuck up.
And these killbots would be programmed and controlled by humans.
Therefore, it stands to reason that killbots would do everything their fuck-up human masters tell them, except without the compassion, remorse, shame, and other emotions that prevent many humans from doing fucked up things.
If a general tells a soldier, "go murder your entire family," said soldier will likely not follow that order. A robot, conversely, would always do what its master tells it, regardless of whether the master says, "go pick some daisies," or "go commit genocide."