According to a survey conducted by Joi AI, approximately 80% of Gen Zers would marry an artificial intelligence. I felt naive for thinking that the TikTok generation – the ones we had to hide the Tide pods from – couldn’t disappoint me any further.
After scrolling through a subreddit named “MyBoyfriendIsAI”, I found that these people appear to be getting something valuable from developing a romantic AI relationship even though their feelings aren’t reciprocated (it’s unclear how many of them are aware of this). To many others, the value of these relationships is incomprehensible.
What’s wrong with dating an AI? Well, morally speaking, nothing. The concern is that these young people will believe that an AI can provide the value that we normally get from engaging in meaningful relationships. If 80% of them were to make a permanent switch to AI, populations are likely to decline and humans would become evermore isolated from each other.
If AI can take on the role of romantic partners, there’s no reason it couldn’t take on the role of friends. Humanity might devolve into individuals interacting with their personal set of AIs and not much else. Again, there doesn’t seem to be anything morally wrong with individuals making these decisions. But, assuming that having relationships with AIs is not as good for you as having them with real humans, it will likely result in a decrease in overall welfare.
This is a substantial oversimplification. For example, AI relationships could ultimately just complement human relationships instead of replacing them. Another possibility is that AI is installed into realistic robot bodies and their programming is improved to approximate human behavior. In this case, they might also approximate the value of genuine human relationships. But we’ll consider the more extreme outcome.
After reflecting on this possibility a bit more, I have doubts about this happening at a large scale.
In many of the philosophy courses I have taught to undergraduates (most of them being Gen Z), I have often covered Robert Nozick’s Experience Machine thought experiment. The thought experiment asks you to consider a hypothetical where scientists have developed a machine that you can be hooked up to that can be programmed to give you the illusion that you’re experiencing whatever life you’d like.
However the machine is programmed, you will live that life virtually, no longer interacting with the real world, and you will not remember having been hooked up to the machine or remember meeting any scientists or going to a laboratory. You’ll be entirely convinced that the life you’re living is authentic.
If you want to live an entirely hedonistic life, the machine can be programmed to give you all of the pleasurable experiences that you can imagine. If you believe that a good life requires some hardships for you to have the capacity to feel fulfilled and truly appreciate pleasure, it can be programmed to make you give you just the right ratio of good and bad experiences. You’ll live out that life until your body dies naturally.
The point is that you can feel like you’re living your conception of the greatest possible life even though all of your experiences are merely being simulated by the machine. The thought experiment is meant to provide a challenge to the claim that the only things that matter are pleasure and the avoidance of pain – a theory of value also known as Prudential Hedonism.
The question that is posed to these young people is, “Would you choose to enter the machine?” As I always expect, they overwhelmingly respond that they would not. “Why not?”, I ask. Their responses are varied but most of them boil down to “because that life is not real.”
The fact that people (including Gen Zers) overwhelmingly reject entering the machine reveals that humans ascribe fundamental value to authenticity.
We can’t help but value authenticity, reality, and genuine experiences, even at the expense of infinite pleasure. This desire for authenticity is more pronounced in human relationships – the fraudulent human connections within the machine are often cited as primary reasons against entering the machine.
Our thought experiments don’t have to be that exotic. I’ve asked students: suppose that all of the people whom you call friends or family, who have given you love and care, were actually all just faking affection for you. Would you want to know or would you prefer to live in blissful ignorance? They all want to know.
Some of the openness to dating an AI may be explained by a mistaken belief that AI can have conscious experiences – that they can feel things, understand language, and reciprocate love. Given what we know about how AIs operate and about what sorts of systems tend to produce phenomenal experiences, the inference that they can have those experiences is unjustified.
Presumably, those who choose to not enter the experience machine would find it difficult to engage in a legitimately enriching relationship with an AI, particularly if they understand that there is no one behind the screen.
There is a more serious matter though. Younger people are lonelier – many of them have trouble finding stable friendships and are unable to find a romantic partner. It would be unsurprising if loneliness is why they’re more open to the idea of dating an AI – it’s either AI or the hamster.
Just like the dystopian possibility of most humans having relationships with AIs can only be brought about by individuals choosing that life, the loneliness epidemic can only be solved by individuals choosing to get off of social media and go out and make friends the old fashioned way – by working on your social skills.
Rafael Perez is a columnist for the Southern California News Group. He is a doctoral candidate in philosophy at the University of Rochester. You can reach him at rafaelperezocregister@gmail.com.