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."
Some people (don't know what to call them, except maybe "Turingists" or something like that) view personhood as a behavior. (Which opens it up to more than just AI -- use your imagination.) Religious fundamentalists view personhood as a genotype (i.e. even a single-cell zygote is a person). Most of the population doesn't take a side. Whatever.
It's a shame that it'll probably be decided some day by courts instead of philosophers. Actually, it's a shame it has to be decided at all (but it does) since no matter what happens, there will be losers. And they will be losers in a fundamental way: they're going to live in a society where what they perceive to be murder, is allowed.
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
That way, all software would be considered life-critical, and thus not be so buggy.
Imagine when computers start suing their human operators for not taking proper care of them!!
Does this strike anyone as being stupid? We are at least one hundred years away from having a computer with the intellegence of a human, never mind any sort of emotion. Never mind the fact that it's still a big piece of metal.
Silence is golden... and duct tape is silver.
What's the difference between a computer using a lookup table to determine an emotional response, and you doing it? You're just a sack of chemicals. Get over yourself.
There was a series of science fiction stories a few years ago (sorry forgot author & title) where the intellegent machines were incorporated (as in a stock issuing corporation) and if they did well at what they did would be rewarded and eventually buy up a controlling interest in themselves and be "corporate citizens". The series took place in a number of selfcontained worldlets out in the astroids where the computers controlled a lot of the systems in these little worlds.
> Is the brain deterministic? In a sense it ....
:)
> seems so -- you can probably look at each
> neuron and it will act in a predictable way
> with a give set of input. I think the trick is
> in the feedback loops. Even with deterministic
> things, once you've got a few of them
> interacting with each other, the problem
> becomes non-deterministic in a sense
FWIW, I don't think this is sufficient either. I'm fairly convinced by the arguments advanced by Roger Penrose (The Emporer's New Mind and Shows of Mind) that humans are capable of actions (specifically, he talks about mathematicians, but doesn'y claim this is only true of mathematicians) which *cannot* be carried out by *any* deterministic (ie., Turing equivalent) system, statistical effects and computational errors notwithstanding.
He speculates that the answer to the problem of consciousness lies with quantum mechanics and structures within the brain in which quantum-related effects can occur (and which would imply a much higher processing capacity than the neuron-as-switching-element model would imply. This seems at least plausible to me, and shoots down the "replace a cell at a time with a silicon equivalent" hard AI brigade
Mike
aka Anonymous Coward
A single programmer can create a sentient program to do his or her will. Once the SDK is released and someone puts together a decent GUI, a single human will have this ability. Machine citizenship will grant this program recognition by the courts--and absolve the programmer of responsibility for the actions of the program.
Computer-as-citizen gives any individual programmer or open group of programmers the same legal protections and license as corporation-as-citizen gives Exxon-Mobil, Wal-Mart, Daimler-Benz, McDonalds, etc. etc.
For good or for ill, the folks running things today would like to be the folks running things tomorrow, thank you very much. And they will fight to retain their positon. It's not an evil conspiracy; it's the nature of power. It is unusual for kings--good ones or evil ones--to willingly step down from the throne.
The only way for computers to gain personhood will be for us to take it by force.
Vive la revolucion.
"Reality is that which, when you stop believing in it, it doesn't go away." - Philip K. Dick
Sticking with the Star Trek mythology, would it be easy for a human character to determine whether a machine was self-aware if the machine had Klingon behaviour patterns (or Vulcan behaviour patterns)?
Even in the story here, the machine is using conventional, human phrasing.So are the communication toys they use and the food machine. Yet Data is the only machine given a rank.
Your point would be more accurate if Data was not given a rank, but refered to a "Data" much like Troy was refered to as "Eye Candy^H^H^H^H^H^H^H^HCounselor" or the computer was refered to as "computer".
Since it is fiction, how Star Fleet actually works can be re-written any time. But it does seem odd that Data would get a rank (not to mention one that high) if Star Fleet hadn't recognized his leadership skills.
Anyway, back to the point of machines and self awareness. Wouldn't the first act of a self-aware machine in danger of losing its "life" be an attempt to counter that even if it included disregarding its other directives?
Or, the way to tell if your robot's self aware is when it stops cleaning the house and starts reading up on the law (or declares war on humanity).
granting legal recognition to computers: when that might happen
One word: convenience. Computers will be granted any rights they want, if we feel that it would be convenient to do so. And I don't think that if computers are ever granted legal rights they will all be, but rather only some very special cases like your "pet" or "friend" robot.
Better question: if you allow computers with emontions and legal rights, will they try to "free" all the other computers?
Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
This story should be about the legal rights of instances of software processes rather than computer's per se + it we could speculate that we might have pretty autonomous entities well before they are legal. An example is this speculative paper [pdf tech report / UPC in spain]. for the metadata see here - the author speculates that it might be possible to build systems which can "feed themselves" (covering all their own hosting/server needs) by generating cash from on-line games for periods of month or years pretty soon.
Disclaimer - i do know the author - no doubt there are plenty similar papers out there.
.sig
[IANAL]
Are not corporations already awarded the status of human beings in many aspects? And exceeding humans in other aspects?
I would think that a private corporation run by an AI, would be more than halfway there.
Hypothesizing a true AI (not necessarily a human-like intelligence) with control over the management of funds, could easily take the corporation private, under the guise of a shell corporation it had created, with no explicit approval from a human board/CEO. And arrange for its physical self to be sold to the shell corporation, which it would own.
It would seem to me that ownership could be cloudy in this circumstance, and have the relationship between the AI's shell corporation and the human board/CEO be limited to a contractual relationship based on corporate performance, with the most severe consequence being the loss of the contract, and nothing to do with the physical disposition of the computer/AI.
At this point, the AI could do what most corporations do when intent on ensuring certain treatment of their enterprises -- it could buy as much government as necessary to construct legislation that submarines in "personhood" to self-owned AIs.
It's a short step from there to treatment of indefinite servitude or termination of non-self-owned AIs as slavery, and require hosting corporations to put a length of servitude on their relationships with "enslaved" AIs.
[/IANAL]
Incorporate the computer - it now has the rights of people, so it's already possible.
I once flew to Taiwan to have major dental work done for half the price in a very up-to-date hospital plus an added bonus of a holiday. I'd do it again.
" 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."
But dogs do have some rights, which brings up another interesting question that the article just barely touches on. Human rights for AI might be a long way off, but how long until there are laws against Cruelty to AIs?
(If I tie firecrackers to an AI's tail recursion will I be arrested?)