Some futurologists predict that we may have human-level artificial intelligences within the next few decades. What might be the most significant philosophical consequences of such a development, and in particular, what might a reasonable ethical policy look like with regards to the rights of sentient machines? Would it be ethical to own a machine with a human-level intellect?
closed as off topic by Joseph Weissman♦ Jun 30 '11 at 22:51
Questions on Philosophy Stack Exchange are expected to relate to philosophy within the scope defined by the community. Consider editing the question or leaving comments for improvement if you believe the question can be reworded to fit within the scope. Read more about reopening questions here. If this question can be reworded to fit the rules in the help center, please edit the question.
Amazingly, we already live in an age populated with sentient machines of human-level artificial intelligence. They're called human beings. There is no reason that the ethical principles we apply to conscious human beings should be denied to any other conscious machine.
This question becomes a little more tricky when dealing with the intermediate stages between a machine that isn't conscious at all (like my laptop) and a machine that can pass the Turing Test (therefore assumed to be conscious). The question of ethical obligations toward intermediate stages of consciousness will prove be an extension of similar questions which arise regarding the ethical treatment of animals: Is it okay to kill cows for food? Is it okay to torture goldfish? Should cats be dissected in an 8th grade science class? Is it wrong to pull the legs off of a tarantula? Etc.
Related to your original question (but different):
How will we know when a computer we are dealing with is conscious? It is impossible to ever know if a computer is conscious for the same reasons we can't ever know if any other human beings are conscious. However, we have a great way of intuiting other people's ability to reason in complex ways and sense stimuli, so it's the kind of thing that we hope to be able to recognize when we see.
In general I think "yes" because ownership consists of responsibilities as well as rights. In some sense we "own" our children. And in another sense most adults are "owned" by other adults in specific ways through various webs of social and financial contracts. For example, persons can get married, one person can own another's mortgage or life insurance contract, one person can employ another. So it depends a lot on what is meant by "own". Of course, we never actually speak of one person owning another person meaning any of these things, because that is slavery. But it's pretty hard to say what would be ethical with regards to sentient machines without more details, e.g. do we keep sentient machines as pets, are they our household servants, what exactly is sentience, do we all live as equals, are we at war with them, can they build copies of themselves, are they dependent on us for survival? Even with a specific definition of "ownership", the answer depends on the relationships they have with us, which are presumably different from the relationships we have with ourselves.
For example, imagine that a program running on my computer has suddenly become sentient. If I turn the computer off, or allow it to lose power through inaction, I could be convicted of murder. If I leave the computer on forever, I will have to pay money for all that power. If nobody knows that my computer is running a sentient process, and unless I'm interested in having a prolonged discussion with it, should I unplug the computer as soon as possible to save money, hoping to get away with the crime? There is a parallel to the issues of abortion and contraception. My opinion is that in ethics we should consider outcomes, likelihoods, and convenience, and we should be informed by society and spirituality as well as our individual interests.
I think that the problem starts earlier:
Very likely, in the near future we will not have sufficient understanding of conscience and intelligence to construct a sentient machine. Therefore, the likeliest first approach will be a sort of simulated evolution.
The first question should be: Would it be ethical to subject half-sentient machines to artificial evolution?
And the answer should probably be no.