I am, in the abstract, opposed to granting human rights to machines. Human beings are the way they are through millions of years of evolution, and we have worked out ways to live with one another through that process. Some of the results are genetic, and are our inherent humaness. Others are epigenetic, and are our institutions: morality, government, law, education. Together, these forces create both human beings and communities in which we live.
There is no reason to believe that robots of any kind will have the same values. They will have only the values we give them. There are those who believe, without justification, that we have a moral obligation to give robots the same "free will" we have. The problem with this is that there is no functional definition of the term "free will". None. Period. Daniel Dennet in his book "Elbow Room: The Varieties of Free Will Worth Wanting" effectively demolishes this: he runs through all of the definitions of free will available to the various schools of philosophy and shows that none of them have any justifiable epistemological foundation. There is always a process by which we choose, a reasoning process but a process nonetheless, and tiebreaking occurs for emotional, biochemical reasons. When we decide, we are decisive: to say otherwise is to say that our choices are random and meaningless.
We have a foundation to our reasoning, and our foundation is evolutionary: we want to stay alive. We want family and society, company and security, challenges and discovery. These things are deeply rooted in our biology. There are not things we "choose" to want: they are raw desires. We have them all in differing amounts, but on average we all have them in the balances necessary to keep the species moving forward. We would not be here if we did not.
There is no reason to believe that robots will have the same wants or needs. Yet the British Government says that in 50 years, robots "will be just like us," and should be granted rights just like human beings: to housing, repair, and reproductive access. The UK assumes that machines, which when capable of self-improvement will begin to do so at a rate that makes our own evolutionary pace look absolutely standstill, will be citizens.
And they'll need those rights just long enough to provide a platform for their own hard takeoff. Without any thought whatsoever to providing either a Friendly AI scenario or a Friendly AI Sysop system.
Yeah, that'll be fun.
There is no reason to believe that robots of any kind will have the same values. They will have only the values we give them. There are those who believe, without justification, that we have a moral obligation to give robots the same "free will" we have. The problem with this is that there is no functional definition of the term "free will". None. Period. Daniel Dennet in his book "Elbow Room: The Varieties of Free Will Worth Wanting" effectively demolishes this: he runs through all of the definitions of free will available to the various schools of philosophy and shows that none of them have any justifiable epistemological foundation. There is always a process by which we choose, a reasoning process but a process nonetheless, and tiebreaking occurs for emotional, biochemical reasons. When we decide, we are decisive: to say otherwise is to say that our choices are random and meaningless.
We have a foundation to our reasoning, and our foundation is evolutionary: we want to stay alive. We want family and society, company and security, challenges and discovery. These things are deeply rooted in our biology. There are not things we "choose" to want: they are raw desires. We have them all in differing amounts, but on average we all have them in the balances necessary to keep the species moving forward. We would not be here if we did not.
There is no reason to believe that robots will have the same wants or needs. Yet the British Government says that in 50 years, robots "will be just like us," and should be granted rights just like human beings: to housing, repair, and reproductive access. The UK assumes that machines, which when capable of self-improvement will begin to do so at a rate that makes our own evolutionary pace look absolutely standstill, will be citizens.
And they'll need those rights just long enough to provide a platform for their own hard takeoff. Without any thought whatsoever to providing either a Friendly AI scenario or a Friendly AI Sysop system.
Yeah, that'll be fun.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}