This really splinters into lots of sub-questions. The main thing is whether the robots are sentient, and then it depends on what one means by ‘having sex’, which covers a multitude of sins, as it were.
The only difference between that and human love is the fact that the robot is artificial. Then there’s the question of whether the human partner in the relationship was under some delusion that the robot was sentient, in which case issues of dignity would come up.
If you consider a robot that is genuinely sentient, then you are transporting yourself into a world very different to the one we are currently in, in terms of what technology can produce. Nobody really has any idea what could be meant by genuine sentience, when you’re talking about experiences around sex and love.
Thandie Newton and Angela Sarafyan as the ‘host’ prostitutes whose consciousness – and humanity? – awakens in Westworld (HBO, 2016)
"If the sentient robot that we’re imagining was created in a lab, or by some corporation, and then was sold as a product you could own as a domestic appliance, then that would be highly ethically questionable situation"
My view is that in the case of a completely sentient robot, you would have to operate on the same standards of consent as you would a human being. I would think that the only reason for not doing so is that they’re artificial, rather than natural. But if the experience is the same – if you’re comparing like with like at the level of experience – then to discriminate would be a form of racism, or speciesism.
If you’re talking about robots which no scientist would consider to have any degree of sentience, I have talked to AI people who have said that they have a programme in their laptop which runs as a model of consciousness, which they think is conscious. When you press them on that, they mean ‘conscious’ in a very vestigial kind of way. They don’t mean that it’s capable of feeling pain, love, passion or anything like that.
There is a controversy within the machine consciousness community as to whether consciousness is just a form of cognition, to put a complicated argument very crudely. There are some people say that it might be possible to create computational archictectures which are very much like, or analogous, to states of human consciousness, in that they have properties which philosophers think human consciousness has – ideas of privacy, ineffability, intrinsicness, which have technical definitions. Even then, it’s very highly hedged, and it may be that only some of the range of human conscious experiences could be expected in robots.
If you think for a moment about the experiences we have with food, you can see that most of them are very intimately bound up with physical processes – eating and digestion and so on – and the neural process which subserve the experience of, say, the taste of chocolate ice cream, are bound up with other processes in the body – in the endocrine system which creates various hormonal secretions which produce the feeling of pleasure we get from ice cream. So it would be very much a full-body process, and the problem is that robot technology is all about trying to reproduce neural processes, which are implemented in our brain, rather than in other parts of the body. Nobody’s thinking very much about creating robots which can eat.
If you can imagine a situation where we think it’s morally okay to have sex with other humans, then that would apply to sentient robots. The problem would be that if the human subject were to be ignorant of the true nature of the robot’s constitution – and, for example, that the robot was a real human – then the human would be being exploited. If the robot were owned by the human, then the robot would be being exploited. If the sentient robot that we’re imagining was created in a lab, or by some corporation, and then was sold as a product you could own as a domestic appliance, then that would be highly ethically questionable situation.