Slashdot Mirror
Legal Rights for Computers
nicholast writes "There's a really smart story in the current issue of Legal Affairs Magazine about granting legal recognition to computers: when that might happen, why it could happen, and what a discussion about it will teach humans about themselves."
This is not a "really smart" story, it's a fantasy. It's too many ill-informed people (with too much time on their hands) that have seen "I, Robot". It even reads like some of the 'Susan Calvert' Asimov stories.
There is a world of difference between programming something to *act* as though it has emotions, and something actually having an emotional or original response. The former is no different to calculating a spreadsheet, the latter has to do with independent and original experiences and actions - implying intelligence and self-awareness. No computer today, no matter how well programmed, is as self-aware as a house fly. We don't grant flies legal rights.
The closest we've come to simulating intelligence, or at least produced non-programmed behaviour in computers are the neural networks coded up where the instructions ("program") are held within and are a function of the dataset rather than the construct. Even neural nets are simply matrix equations, albeit non-linear usually, and are thus completely deterministic. The typical neural network has less than 1000 nodes within it, the human brain has 100 billion neurons on average (with 10-50 times that many glial cells). The phrase "does not compare" doesn't even come close.
So, in short, what a load of rubbish.
Simon.
Physicists get Hadrons!
How about we just concentrate on holding on to the legal rights we HUMANS have in Bush's America?
Just look at the history of women rights, black rights, gay rights. Some of those cases are "solved" today, some of them are pending, but one thing is for sure: as soon as another category of sentient beings demands equal treatment, as subject, not as object, it gets nasty and former "master race" rarelly gives up without a fight.
Robert
Bastard Operator From 193.219.28.162
Our legal system is far behind the times when it comes to technology, 'cyberspace', online privacy, etc. I wish todays legal minds were working on those issues instead of dreaming up these far off futuristic scenarioes.
An artificial intelligence/computer should be granted the same rights as a human if, either itself or its maintainers under oath, it can pass the Turing test to the satisfaction of a judge.
That episode needed to be completely re-written.
Data already had the rank of Lt. Commander. That means that Star Fleet already recognized his ability to make decisions on his own.
Therefore, his decision to NOT be disassembled would not be challenged.
In order for the case to make sense (I know, it's Star Trek) then the robot would have to not have any prior recognition of its independence or decision making.
Star Fleet recognized Data sufficiently to give him a rank that allows him to order humans to risk their lives (do the 3 laws apply in Star Trek?).
Science fiction has been dealing with the "what if they get smarter than us" issue pretty much forever... but i don't think a Star Trek court would count as precidence in a real-world case, especially since the "date" in question hasn't happened yet.
I don't know about the "soul" issue, but as far as what true AI would do the work world, offshoring is giving us a taste now: Global communication has made access to brains cheaper and cheaper. Programming ability now costs about $2.20 an hour on the world market. When bandwidth gets even cheaper, remote-controlled robots that do plumbing, painting, burger flipping, etc. will dominate. The robot repair will even be done remote. Only unions and political pressure can stop it. The remaining jobs will be sales and management. Thus, either we need a new wealth distribution model, or those without sufficient people skills will die in the streets.
Welcome to the future, slashdot. Your brains are growing increasingly more worthless to capitalist every day. I am just the messenger. Have a nice day.
Table-ized A.I.
You can probably also find "brain surgeons" willing to work for $2.20 per hour "on the open market". Any takers?
I don't care if it's 90,000 hectares. That lake was not my doing.
It is funny that large parts of humanity lack these same rights, yet we are so concerned with computers??
I had a couple of dogs that seemed to me to be self-aware. They did not have the same legal rights I did as far as I know. So why would a computer be granted legal rights based on self-awareness?
Chimpanzees have some intelligence, as do dolphins, but we still confine them to zoos and do not afford them the right to a public attorney to work toward securing their freedom.
If we base legal rights solely on intelligence, than when someone has a stroke, enters a severe coma and is no longer able to demonstrate cohesive thought, does that mean they should not have rights anymore?
Just food for thought. Soemone with a better philosophy background than I (he or she took TWO or more philosophy classes) will probably be able to answer these questions better than I.
But there is no inherent reason why computing power can't someday reach the level of the human brain. If Moore's law continues, this is supposed to take under 30 years.
We can't even simulate a spider's intelligence yet. It's not a problem of needing more cycles.
We need to work out how we think, and then try to "seed" this behavior into a machine that can learn. There are lots of interesting ideas out there, but every practical attempt I've seen has either been side-tracked by efforts to build interesting hardware, or too-ambitios attempts to jump stright to full intelligence/learning by taking "shortcuts" where you define behaviors and responses in software.
I expect the solution to emerge by itself once we've modeled some basic life "rules", and set a learning simulation running on them. i.e. start with a very simple 2D "game" in software, where the goal is to pick up randomly scattered food pellets. Pick them up too slowly and you die. Gradually let the methods for food pellet searching evolve itself, using genetic algorithms. Then throw in some competition - make more than one organism active at a time so they have to learn even better alogrithms. Then add elements such as the ability to kill each other- behavior such as alliances may emerge. Then make food appear seasonally, and give them the ability to stockpile it. Gradually keep adding more elements to the simulation, and let the intelligence unfold on its own.