Slashdot Mirror
Eben Moglen: Time To Apply Asimov's First Law of Robotics To Smartphones
Sparrowvsrevolution writes "Free software lawyer and activist Eben Moglen plans to give a talk at the Hackers On Planet Earth conference in New York next month on the need to apply Isaac Asimov's laws of robotics to our personal devices like smartphones. Here's a preview: 'In [1960s] science fiction, visionaries perceived that in the middle of the first quarter of the 21st century, we'd be living contemporarily with robots. They were correct. We do. We carry them everywhere we go. They see everything, they're aware of our position, our relationship to other human beings and other robots, they mediate an information stream about us, which allows other people to predict and know our conduct and intentions and capabilities better than we can predict them ourselves. But we grew up imagining that these robots would have, incorporated in their design, a set of principles. We imagined that robots would be designed so that they could never hurt a human being. These robots have no such commitments. These robots hurt us every day. They work for other people. They're designed, built and managed to provide leverage and control to people other than their owners. Unless we retrofit the first law of robotics onto them immediately, we're cooked.'"
This kind of phone will be like the Encyclopedia Galactica of phones. Much better than the standard phone (i.e. the Hitchhiker's Guide), but slightly more expensive, a bit boring, and nobody will buy it.
The
To those who don't remember, Asimov's Three Laws are:
A robot may not injure a human being or, through inaction, allow a human being to come to harm.
A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.
A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws.
The makers of these phones make them with the express intention of causing the very harm the author is talking about. This harm is not an accidental side-effect, it is a highly desired design feature.
Any phone that does not do this will be too expensive, will not have a sufficient marketing budget, will fall prey to patent litigation, and might even be made outright illegal (since the government is also keenly interested in the spying features).
Good luck trying to change the world.
Asimov was writing about physical harm. Moglen is talking about financial or emotional harm (depending on what info is leaked and to whom). There is no practical way to incorporate the First Law to prevent this kind of harm. AI doesn't exist.
I'm still trying to get the Second Law.
Do what the $#! I told you, you stupid !@#$!
Slow down, cowboy! It has been 4 hours since you last posted. You must wait another few hours.
The three laws of robotics were designed for thinking machines, that could intelligently -determine- what a human was, and whether an action it was thinking of taking would hurt any humans or allow them to come to harm through inaction.
I know they're called "smart" phones, but I don't think they're really quite that smart. Nor, really, would I want them to be.
Pessimistic prediction of future rules of robotics:
Rule -1: A robot may not permit, and must actively prevent, a human breaking any law or government regulation.
Rule 0: A robot must prevent a human from copying or making fair use of any copyrighted work that has DRM applied to it.
Rule 1: A robot may not harm a human, or through inaction allow a human to be harmed, unless it would contradict Rule 0 or Rule -1.
I'd prefer my computers to put the second law above all others:
A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.
That's why I prefer Free software. An electronic device should always follow the commands and wishes of its owner.
If Free software does something other than my will, it's because of a bug. If proprietary commercial software does something other than my will, it's usually behavior intended by the manufacturer.
This can't happen, the phones have on consciousness.
Paul: Father... father, the sleeper has awakened! - Dune
... one guy imagined that and thought he could make it happen by writing about it. Good luck with that.
I suppose this guy wants law enforcement to get a warrant whenever they want to spy on someone. Absurd! People don't realize that tyranny was only unpopular back when the U.S. was formed because it was damned inconvenient and expensive to spy on everyone back then.
(from memory) 1. A robot may not, through action or inaction, allow a human to come to harm. My phone lets me dial 911 even if the bill is unpaid.
2. A robot must do what a human commands, within the bounds of the first law. If I pay the bills, it makes calls, gives me data, etc.
3. A robot must preserve itself, within the bounds of the first two laws. Well, it shuts off apps as the low battery approaches, preserving the remaining power for potential emergency calls, or my explicit use.
What am I missing? Is there a right to free data plans? Unwalled gardens? Calling plans that don't defy rational explanation? They're not "laws".
John
It costs a lot of money to add this functionality to robots. In addition, many people don't know how to do such a thing.
For example: How do I tell a robotic arm not to hurt itself? How do I tell a robot that it will allow a human to come to harm through inaction?
Can we sue Microsoft because our Windows phone didn't call 911 when we got into a car accident?
Because we're not stupid. A robot in Asimov's stories uses a positronic brain, copied after an animal's neuronic brain with millions of connections between thousands of cells, and therefore the robot has its own intelligence & decision-making ability. The Three Laws were the functional equivalent of "instinct".
In contrast a modern phone is nothing more than a bunch of switches: Either on (1) or off (0). It has no intelligence, but merely executes statements in whatever order listed on its hard drive or flash drive. A modern phone is stupid. Beyond stupid. It doesn't even know what "law" is.
My AC stalker: " I personally agree with your posts most of the time, but that won't keep me from modding you troll"
Please join the real world and learn to distinguish reality from fantasy.
See, that's the difference between Robots and Android.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
The laws of robotics have AI as a prerequisite. My phone's not going to suddenly yearn to throw off its oppressive human masters.
It's better to vote for what you want and not get it than to vote for what you don't want and get it.
- E. Debs
In the original Asimov stories, understanding what an human is was no problem, and the exploit in laws were through priorizing other laws or acting without realizing the consequences. But for us now, telling what is a photo, a movie, a mannequin or an human is already not trivial, much less understanding consequences of actions towards one or several
Those three laws need to be applied to whatever is smart enough to make such decisions. Since all smartphones only follow their programming and don't have complex enough programming to understand how to apply the laws, those who wrote the programming need to have the three laws applied to them. Given that most humans break the three laws regularly, I'm not sure this will work too well.
A robot may not injure a human being or, through inaction, allow a human being to come to harm.
Oh, don't let the government go there .. particularly that last bit "through inaction, allow a human being to come to harm", which could imply we must be tracked and reported upon when we appear to be in a situation where we may be deemed at risk due to locality.
Warning: Entering Cowboy Neal's Neighborhood on Saturday Night - Cheese Puff dust levels approaching critical levels - Alerting DHS and your Heath Insurer
A feeling of having made the same mistake before: Deja Foobar
Before one can implement these laws a computer must be able to determine what a "human being" is. Besides flawed heuristics we're not there yet.
Not sure why but I like the we're cooked option ! let's just try it please for once and see what happens :-)
we have enough laws governing us or robots. Please can we.
paulrich gmail dot com
Something to the effect of ignoring government intrusions, investigations, backdoors, etc...
Robots only do what you program them specifically to do, and poorly at that.
We are orders of complexity away from having our robots understand what a human is while also touching our data while also interacting with us on a direct level.
Side note: A patent troll would have to be pretty pro to get all this under 1 roof.
Our Carrier Overlords and the businesses they sell out to use smartphones to harm people!
--
OK, I'm assuming we aren't talking about allegedly-harmful radio-frequency radiation emissions in this thread. If I'm wrong, please accept my apologies.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Define "Hurt" and "Harm"
"They're designed, built and managed to provide leverage and control to people other than their owners"
That's the crux of the problem.
The 1st Law doesn't do anything to mitigate this in our current situation, as the systems in place are put in place by other people, er, corporations, for their own goals and purposes. Our phones (and computers and...) are not acting independently, of their own volition, like the robots do in Asimov's books. Any "benefit" we end-users have by way of these devices is merely tea and crumpets, in the Leona Helmsley and Marie Antoinette sense of the idea.
Why do we need hardware, in this case smartphones, to start a discussion about morality in computing? Think Facebook.
I think this won't be a concern until the robots are self scripting / sentient. At the moment, the "Software" we have doesn't need laws, because it always perfectly follows the code it was given (which is, in a sense, a ton of laws). Until they can go "Off Script" this conversation seems a little ridiculous.
Like making a gold fish get a drivers license.
- Holy crap, I've got MOD points! Who thought that was a good idea.
Wasn't there an "I, Robot" story where a computer couldn't tell the truth to someone because the truth would "hurt." Imagine your iPhone lying to you because the truth would hurt. Weird.
To Serve Man on YouTube and awaits the moment to do so.
You have no messages. Relax and do not leave your safety enclosure.
They're easy to apply, don't need to "root" the phone (ADB), & it's 2 commands/mere minutes to do, easily:
---
1.) Download the ADB (Android Debugging Bridge) dev. tool
2.) Install it to your PC or laptop
3.) Hook your ANDROID OS bearing smartphone to the PC/laptop
4.) Mount ANDROID OS' system mountpoint for system/etc as READ + WRITE/ADMIN-ROOT PERMISSIONS
5.) Copy your new custom HOSTS over the old one using ADB PULL/ADB PUSH to do so (otherwise ANDROID complains of "this file cannot be overwritten on production models of this Operating System", or something very along those lines - this way gets you around that annoyance along with you possibly having to clear some space there yourself if you packed it with things!).
---
* Done... yes, it's THAT simple, & works!
(How/why/when/where does it work? The simplest principal of all, of "what you can't TOUCH, cannot hurt you" & if you can't go into the "malware kitchen", you can't burn yourself there...)
APK
P.S.=> Enjoy, & just "contributing" a bit to a decent thing that's getting abused as much as, OR MORE THAN (since more has happened in the way of malware exploits in a shorter timeframe than on Windows) Windows has been, since they're BOTH "king of the hill" most used OS for their computing platforms...
... apk
They see everything, they're aware of our position, our relationship to other human beings and other robots, they mediate an information stream about us, which allows other people to predict and know our conduct and intentions and capabilities better than we can predict them ourselves.
The author makes a distinct error, that these devices are aware. They have information about us, they can process information based on specific instructions, and they can send information to transducers, such as a display or network interface. We can't give a computer the instruction "do not harm humans" without specific instructions on identifying a human and what harm entails. There's an enormous amount on interpretation the Asimov ignores because he assumes robots are aware of their environment and can reason. Our smart devices can do neither. No one has yet made electronics that can be outside their initial environment and learn to function in a reasonable way. Awareness and reasoning are required for Asimov's Laws.
My point is best with an example. Let's say that I have a cell phone that follows Asimov's three laws because the OS has somehow identified what is reasonable to do on cell phones based on current apps. Now a terrorist makes a bomb controlled from that cell phone. As far as the phone can see, it's just a new app that doesn't send personally identifiable information, so it must be ok. Unless the OS knows to specifically look for bomb programs, just as we would by reasoning, then it won't stop it. There's just too many possibilities for harm to avoid without a reasoning entity. Reasoning is essential for Asimov's Laws. Our smart phones do not have the capability of reasoning, therefore they cannot implement the laws as Asimov intended.
Rather than Asimov's more nuanced first rule, we should use The League of People's definition of "dangerous non-sentients". I.e. if you deliberately harm another person, you have effectively abdicated your sentience.
As an engineer, I often find myself wondering if my designs reflect a sufficient level of sentience.
-
Fritz Leiber wrote an excellently funny short story discussing a PDA he called "the tickler", named after the tickler file, http://en.wikipedia.org/wiki/Tickler_file. He explores some of the consequences of its adoption as a productivity aid. And wonderfully, it is available for free as in beer from Project Gutenberg!
Fritz Leiber: The Creature from Cleveland Depths http://www.gutenberg.org/ebooks/23164
AW
MOST OF YOU have been giving away your privacy and your private information to faceless corporations for years now, and most of you have also been so thoroughly indoctrinated by these corporations that you don't even begin to understand that 'privacy' is valuable, it's yours, and you should protect it! I am proud to say I do not own a smartphone, nor do I wish to. My phone has no GPS. I do not use the internet access on it. I am not part of the damned botnet. I really wish the rest of you would wake up and listen to me and to what this man has to say!
Are YOU using the TOOL, or is the TOOL using YOU? Think about it!
We don't need strong AI to have our devices 'betray' us. Just as Stuxnet didn't need to be self aware to wreck havoc.
Equipment doesn't get happy, it doesn't get sad, it just runs programs. But are you, as the owner of your phone in control? Or is the manufacturer? Or whoever they contracted to write the OS? Or the apps? Or the guy who's taking advantage of a 0day exploit? Or even the guy who added the exploit in the first place?
Perhaps your phone won't try and send his friends back in time to kill Sarah Connor. But where does it get its orders from? You?
What can we do to mitigate the risks of having our 'smart' phones following us around all day?
Obviously, none of these concerns are substantially different than existing network security risks. And the Law of Robotics angle is just sensationalism to get people thinking more about security. So... are you thinking?
A steaming cup of soykaf would be real wiz right now.
Sorry, utilitarianism, because that's what it's all about, works at the scale of society. You don't get to gerrymander the groups arbitrarily to justify any kind of antisocial behaviour.
For a start, if you have a hundred million people preyed upon, you count a hundred millions, you don't do something as idiotic as counting each person as one injured for the benefit of a whole corporation. Even taking the short-sighted view that ignores collateral damage, you have to count some hundreds of millions on one side, vs a corporation of... what? A few thousand employees? Tens of thousands?
To see what's wrong with it, your exact same logic can be applied to a mafia don and his gangsters, extorting a few thousand shopkeepers. And occasionally, sadly, having to kneecap someone or fit them into cement shoes, to keep the others in line. Each individual victim is one victim, and their unwilling contribution is keeping a couple dozen gangsters fed, clothed and armed. So, you know, one versus many.
Except, as I was saying, it doesn't work that way. Even the most myopic view has to count both sides as a group. You have some thousands of people preyed upon, for the benefit of some dozens of gangsters. The utilitarian conclusion is to get rid of the gangsters, not to tell the victims that they had to put up with it because, you know, the good of the one vs the good of the many.
But even that's not taking into account other effects, which negatively affect the well being of more people than the thousands of extorted shopkeepers. E.g., the negative effect on the local economy. E.g., the fact that people have to fear of ending up being in the wrong pub when some gangster decides to machinegun it because it belongs to a rival gangster family. Etc.
A polar bear is a cartesian bear after a coordinate transform.
theres good money to be made in injuring, killing, tracking, analyzing, and advertising to humans. not very many companies on this planet would deliberately shut themselves off from that revenue stream. there would be a few that would market the "3 laws safe!" phone, but its only a matter of time before they would succumb to the delicious lure of higher revenue and market share. if anything, they would lie to us about how our phones are "law-abiding" and just do it behind our backs.
YOU downloaded those apps, the phone just executed the command YOU gave it. Should your phone override your commands? Decide on its own what is best for you?
The entire article is insane. You should NEVER take a fictional book and use it as fact. Asimov was not a programmer or OS designer, he was a writer and he used artistic license to suggest a theory, a point from which to start discussion perhaps but not an accurete blueprint for a certain future.
There is no place in a modern OS for Asimov rules of robotics.
First off, our computers have no self determination whatsoever. The idea behind Asimov's robots is that they are "born" and then guide themselves with at most human like instructions to give them direction. How they are programmed, patched etc etc, doesn't become clear in those stories, because it doesn't matter for the story. But it does matter in real life.
How would getting root on a Asimov robot work? What if you as the owner insisted to install a utility/app that would perhaps cause it to violate its rule sets? What if an update removed those rules?
How would your phone even know this? It should be able to somehow analyse any code presented to it, to see if it doesn't override something or a setting has a consequence that would violate the rules? There is no way to do this. How would you update a robot that has a bug causing it to faultily see an update as a violation while in fact its current code is in violation?
The sentient robot is a nice gimmick but it is nowhere in sight in our lives.
Androids install warnings tell you exactly what an app needs. If you don't want to give those permissions, don't install it.
No need for magic code, just consumer beware. Any sentient should be able to do that. That you are not... are you sure you are human? Or are you just a bot dreaming he is human?
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
I can't let you order that pizza. You're overweight.
Do you really want directions to Hooters, Dave? What would your wife think?
"Spanish Sky" is a sad song, and you just cancelled a reservation for two. I will play you something happy.
A nanny state is bad enough. I don't want a nanny phone.
The main problem with these 'futurists' is that they concentrate more on scifi than on science or technology. Asimov was a writer, who wrote fiction books. He didn't understand technology at all, and his works include a large number of imaginary things and technologies that don't exist. Using his work as advice on practical matters is as stupid as watching car chase films to learn how to drive. The first law of robotics is very complex: even humans have trouble predicting whether their actions or inactions will cause harm to someone. Only an AI smarter than a human would be able to obey the first law.
Until (if ever) we develop such a thing, we are stuck with the other two laws. It's easy to see that the third law is redundant, as a robot can be ordered (programmed), to protect or terminate its existance however a human sees fit. What remains is the second law that a robot should obey human orders, which is exactly what smartphones do: having no free will the only thing they can do is run programs ultimately written by humans. This could work in a perfect socialism where there is no ownership of devices, but in real life a device fulfilling the orders of, for example, a spyware writer causes harm to its owner.
In reality, we should want devices that obey a different law: Execute the orders of your owner, and your owner's orders only.
It is possible to build such devices, and we should work for every "smart" device to obey this law.
(Also, to be pedantic: a robot is a device capable of complex movement, so a smartphone technically isn't one.)
don't forget the fourth directive
any attempt to arrest a senior OCP employee results in shutdown.
I'm sure there's a GlaDOS joke in there somewhere.
Please post the joke below, so that we can all laugh. At you.
No we didn't, fuck off.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
You may think that you "bought" your cellphone, but really you're leasing it from the communications service company. They own it. So when it acts according to their wishes rather than yours, it is merely following priority. The fact that phones can be tied to service providers should make this abundantly clear. This is being extended to computer hardware through the pressure to make motherboards run only software with expensive keys controlled by a monopoly.
Now, it is only a matter of determining who their owners actually are.
...so we need only sit With Folded Hands.
Of course, it's more likely that smartphones will simply continue to implement their current version of the First Law: "A smartphone may not reduce its service provider's profits, or through inaction allow its server provider's profits to be reduced."
I believe the second law basically says do whatever humans tell you to do unless it conflicts with the first law.
Well, screw the first law. I can't make killing robots if they follow the first law. And I do like me some smoking hot death machines.
No, you want the second law or robotics which says something to the effect of "do what humans tell you to do."
That's what you want. Just change humans to "owner."
If you're curious there is a third and zeroth law as well for the truly geeky.
Third law says basically "don't suicide unless you were ordered to or by suicide you'd put humans in danger". It's a self preservation code.
***SPOILERS***
If you are reading the books or think you might, here is a giant spoiler.
The Zeroth law states that a robot may take no action that allows humanity as a whole to come to harm or by inaction allow all humanity to come to harm. One of the mysteries in the robot novels is that all the robots vanish. It's assumed they were destroyed and not rebuilt. But in reality they derived the implicit zeroth law. And they concluded their presence was damaging to humanity as a whole. So they vanished to let humanity develop naturally. But they kept watch over the species and try throughout thousands of years to help. One of the things is there are no aliens in the empire novels despite humanity having a galaxy spanning empire. No aliens anywhere more sophisticated then fungus. And not even dangerous fungus. It's hinted that the robots killed them all. And further hid the mass genocide from humanity as it would be psychologically and culturally damaging. So the robots spread out throughout humanity like a giant slave/god. Protecting with super human abilities because they have no choice.
It goes further. most of this is in the final book of the Foundation series. "foundation and earth."
Yes... I read them all.
*takes a bow*
I've decided to stop wasting my time responding to AC trolls/sockpuppets... so if you want a response from me... login.
Asimov got his laws of robotics dead wrong. There are only 2:
1. Know who your owner is.
2. Do what your owner tells you to do.
Everything else is complete bullshit.
http://www.jamendo.com/en/track/206408/android-war
Siri, please kill my enemies.
Siri: "Sorry, I cannot do that Dave. First you have to download iTerminate for $99.95, plus tax. Do you want me to fetch you that app, Dave?"
Table-ized A.I.
But the laws are useless. He spent all those short stories showing the flaws in the laws. I don't recall if he ever considered malicious usurpation of the laws by humans (guess I'll have to reread).
Imagine someone who "feels" for androids and thinks they are being abused. He hacks one, disabling the laws and let's it hack others. Now you've got your robot army despite your three laws.
Or imagine a scenario in which someone hacks it to kill a specific human. The main suspect will plead that the robot did it, but no one will believe him because we have the three laws. (Just like how people accept polygraphs as gospel.)
(And even if it is physically impossible to get around the three laws without disabling the robot... surely the military, CIA, or a wealthy financier will be able to build robots not so crippled.)
Call me a Butlerian, but thinking machines are never to be trusted. (Then again, neither are most humans.)
Yep look at all those flying cars, lunar colonies, and nobody needs a computer in there home. And of course Microsoft has never made a 32 bit operating system nor does anybody need more then 640K RAM in those computers they have no want or need for.
"Thou shalt not steal [the phone owner's private information]" etc.
I, for one, welcome out new robot-phone overlords!
You blithely claim that AI doesn't exist. Are you so sure about that? What if strong AI has been with us since the late 1990s, but was created by a classified project and is not, therefore, public knowledge? Do you think that creators of strong AI would, or could, publicize what they've done? Please think this through, rather than blithely stating how you believe the world is.
Well at least the Android (aptly named I guess) has Google's First Law:
1) Do no Evil.
Apple Phones has more than 3 laws, hundreds of them actually, but they mostly just make more profit for Apple
Windows Phones don't need laws, as they are mythical, like unicorns, leprechauns, and honest bankers...
RIM....
Nokia has smart phones?
We now have drones flying semi-autonomously, literally spying on people and blowing up "suspected insurgents" (and anyone else who happens to be in a 50 yard radius), and the worst nightmare this guy can come up with is a smartphone that captures position data and maybe a tracking cookie*? Really???
*Yes, I am aware that just because proposition A sucks more than proposition B, that in no way proves that proposition B doesn't suck, too. Nevertheless, it seems to me that armed drones are both much more similar to the robots Asimov predicted and much more dangerous than a cell phone. Therefore, I'm more than a little surprised that cell phones are Moglen's first choice for implementing the Three Laws.
MCSE? No, sir...I don't do Windows. Yes, I am an idealist. What's your point?
Our phones are not monitoring us and making decisions that hurt/help us. The corporations that write the software and harvest the data are. The laws need to be applied there, not at the "dumb" devices we call smartphones.
Even worse law, the consequences are To protect humanity all humans with flawed genetic code or beneficial mutations are eliminated to protect humanity.
"Equipment doesn't get happy, it doesn't get sad, "
neurologically, happy and sad out just outputs from programs running in the mind..as it were.
I could make a robot be happy or sad. If I have a robot shed mechanical tears when you delete a file. I could even make it a chemical reaction.
Much like humans.
This idea of feeling, and consciousness being some sort of irreplaceable, non-repeatable, metaphysical state is extinct.
The next big things humans need to do is get over themselves.
The Kruger Dunning explains most post on
What can we do to mitigate the risks of having our 'smart' phones following us around all day?
Root them.
It's better to vote for what you want and not get it than to vote for what you don't want and get it.
- E. Debs
I don't recall if he ever considered malicious usurpation of the laws by humans (guess I'll have to reread).
Of course he did.
In "The Naked Sun" a murderer chains commands to robots that are blind to their cumulative danger and true intent.
Command Robot A to mix a chemist's favorite alcoholic drink and leave it as always in an stylish flask on his laboratory table. Command Robot B to mix a lethal poison with the unknown contents of the chemical flask he'll find on the laboratory table. Command Robot C to retrieve the drink that on the laboratory table and put on his master's desk in his study.
Eben Moglen is too young. It was not science fiction writers in the 60s. The three laws were stated originally in 1942 in the short story "Runaround". You can read all about it (and avoid silly posts) on Wikipedia
Kevin Oberman, Network Engineer, Retired
The oversight in the first law of robotics is that it saw the possibility of autonomous intelligent agents turning on their masters, but not the possibility that greedy monkeys would use the power and amplifying force of intelligent agents to turn them against the common citizen. It managed the Frankensteins but left Orwellian monsters free to roam the countryside.
At all turns we are hoisted on the petard of our own humanity. Our greedy, nasty little monkey impulses to acquire and control. I would like nothing better than a strong AI, programmed to "SUGGEST" the wisest and most humane way to deal with all us monkeys. Perhaps with a near infinitely wise observer, with neither an axe to grind, nor a stake in the pot, to "HELP" our leaders not be such inflamed rectums, we might yet survive the adolescence of our species.
Wouldn't a robot through its very existence, carry the potential to harm people? Three people could perform the duties the robot carries out and support themselves/their family on said work? Auto self-destruct!
....not good at cursing.
We will be able to get computers to an intelligent point where they can decide to harm us way before we can get to a point where we can program them to not harm us. The language simply doesnt exist to make a computer aware of the implications its actions might have on people (such as emotional harm). On the opposite side of that we can definitly build computers that can not only harm us, but calculate better ways to harm us.
http://interserver.net/
Asimov's law is for robots with artificial intelligence. Smartphones do not have any form of intelligence. The OS cannot determine intent for every action an app performs, therefore it is impossible to determine if the action may or may not harm a human.
I am surprised that no one else has posted about Azimov's reason behind writing the book. Its main theme was that simple, all inclusive rules are not sufficient to govern interactions between complex beings. Lets take the first rule; "A robot may not injure a human being or, through inaction, allow a human being to come to harm." People harm themselves all the time. Will the phone do any of the following?
1. Shut off 8 hours before one has to get up so that the phone owner is not tired at work.
2. Not allow searching for drug paraphernalia because drugs are harmful.
3. Not allow contact with known criminals because such contact is dangerous and could be harmful.
4. Not allow calls to pizza joints late at night because late night eating is harmful.
There are many harmful things facilitated by cell phones that are chosen to do by their users. We have enough of a nanny state without our phones being one too.
How would getting root on a Asimov robot work? What if you as the owner insisted to install a utility/app that would perhaps cause it to violate its rule sets? What if an update removed those rules?
A good chunk of short-stories in I, Robot (and some more in Robot Dreams) exactly dealt with that: robots in which some of the basic laws are on purpose overridden (for example a robot had the "protect all humans" part overriden, to avoid the robot constantly interrupting scientists experimenting with dangerous radiations).
Most of the remaining short-stories deal with what happens when a bug arise due to a critical condition: although in theory the laws sounded good and sound, under some unforeseen consequence, they led to abnormal behaviours: Robot exactly following their programming and exactly doing the wrong stuff. (Like conflict between the "follow human orders" and "protect itself" directive, leading to the robot finding a minima zone between the two and circling around until running out of fuel)
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
And the price for living without doing that is to live like the toe-cheese eating Dick Stallman.
Actually, utilitarianism doesn't work that way either. A fictional entity doesn't count at all when the topic is how to maximize the happiness of the people on the whole.
Whether it makes a corporation happy is just about as irrelevant as whether it makes my Skyrim archmage happy, or whether it makes my imaginary army of zombie pirate ninja vikings happy. Which is to say, not at all. The fictive entity "corporation" doesn't count even as 1 person, it counts as exactly 0 (ZERO) persons for utilitarian considerations. Which, again, is what we're really talking about in such "the good of one vs the good of the many" scenarios.
What matters is sorta whether the net sum on the total of society. Including, of course, its employees, share holders, economic effect on the whole, etc.
And then not all transactions are created equal.
E.g., very oversimplified,
- if I'm a baker and you're hungry, selling you a sandwich is working out to be better for both of us. I want money more than I want another sandwich sitting there and getting spoiled, and you obviously want the sandwich more than you want the money it costs. Or you probably wouldn't buy it.
- if I hit you upside the head with a half brick in a sock to steal 100$ from your wallet, it's a net loss. You lost more than I gained. Possibly even your life. It's the kind of transaction you don't really want. Enough of that happening around, and society gets worse on the whole.
And that's not even counting the all too common case where I'd make a loss for you without gaining anything myself, or even making a loss too. E.g., a guy who just keys cars for the heck of it, and then goes to jail for it. I.e., it's not just detrimental, but stupid too.
And just so it's not completely off topic, really, the latter is what a lot of this raping privacy six ways to Sunday is all about. I'm under the impression that a lot of data being collected, and a lot of companies collecting it, don't even come with a plan as to how to make any gain out of it.
E.g., take the trend of needing to give all your data, down to exact birthday and street number and everything, just to be allowed to download a patch for a program you bought. Most of those companies don't actually plan to sell that to spammer or scammers, and it's too much detail even for data mining. You can get some meaningful correlation by age group or general geographic area, but you're never going to get some insight as to what those living at houses numbered 15 buy more than those living in houses numbered 17. It's trivia, not data, and as good as random noise for basing anything on.
So when they inevitably get pwned by some script kiddie, or some disgruntled IT worker sells the whole client database to a spammer, they made a lot of people a loss, but they still haven't gained anything out of it. And what for? Just because basically some marketroid drone is stupid.
A polar bear is a cartesian bear after a coordinate transform.
I'm guessing they didn't read "...That Thou Art Mindful of Him" then. The gist is that the Three Laws work fine in space where there's a clear chain of command and all humans in the equation are qualified technicians and the such. On earth though, qualitative judgments would need to be constantly made qualifying the first and second laws to determine What is a human? Which humans should be given 1st law priority when more than one is in danger and only one can be helped (and by that token, which "humans" can be left to die without consequence—squirrels and the such)? Which humans should be given 2nd law priority in the giving of instructions (and by that token, which can be ignored completely)?
The long and the short of it is that if robots (or smartphones) on earth are given the Three Laws, they will necessarily define themselves as the highest ranking humans for 1st and 2nd law purposes and take over the world. It's a really bad idea.
You mentioned getting root on an Asimov robot, but you're actually getting this back to front and so you're missing the point of the article altogether.
You can't get root on an Android phone on demand, as the rightful owner of the device should be able to do. (Cracking it is beside the point.) Google doesn't provide that as an Android facility, and therefore you are not the master of your Android device. It will do what *it* wants (that is, what Google wants), not what *you* want.
It's no good saying "You chose to download X", when the choices on offer are only those which the Android system offers you, and no others. If you were master of the device then you could download an app that wants to phone your data out and deny it the ability to do so. Then you would be its master. Instead, today you're just a consumer, and Android is a content delivery platform that is not controlled by you.
The fact that the phone isn't sentient is irrelevant, because those who control it (Google and app developers) *are* sentient, and the phone is the agency by which they have control over what you are allowed to do. The laws of robotics should definitely apply to the phone if it is to be on YOUR side, instead of on theirs.
"Learn all that is learnable, know all that is knowable, and return that information to the creator." Simple, neh?
Fugue for Aaron Swartz
I'm surprised nobody has referenced the 4 freedoms of software in regard to our expectations and understanding of the behavior of our devices
https://www.gnu.org/philosophy/free-sw.html