If AI parters won’t be banned or the access to them won’t be severely restricted (i.e., AI partners only available for clinically diagnosed psychopaths, people in severe depression, people over 30 or even 40 years old, or people who have lost their spouse, like in that Black Mirror episode, and some other rather special cases like these), I don’t see how the “AI partner” technology could be turn out anything than a disaster for the society, on the scale of the social media disaster (if not worse), albeit of a different kind.
I think in the section “People might neglect real romance”, the author fails to extrapolate in the future, as Raemon pointed out, and fails to see through realistic psychological implications.
We are talking about AIs that will be 1) extremely beautiful (like Caryn Marjorie who made a digital copy of herself), 2) smarter and more eloquent than the human who is partnering with the AI; 3) always extremely affirmative, flirting, attentive, etc.
Humans will fall in love with these AIs. Once they are in love with AIs, many (maybe even most) humans won’t get off the hook “just because they are bored” or because they would think “it’s time to get a real relationship and family” (see this comment). When humans are in love they usually don’t look to end their relationship, right? Also, integration of vector databases over the history of all chat with the AI will give them essentially unlimited memory and the fact that AIs “don’t remember the relationship history and shared memories” and thus humans become disenchanted because of that won’t be a problem. Even humans who will manage to get off the hook may find it hard to form human relationships or being satisfied in them afterwards.
So, why this is will be a disaster for the society? Because 1) total fertility rate will plummet even deeper than it is today in most developed countries and in Asia, and because 2) the fabric of the society (woven through relationships) will become thinner (cf. the intersubjectivity collapse). The second problem could be partially ameliorated if parner AIs will be learning and talking to each other, as well as to other people than their partners, i.e., will be actual members of the society, but “AI dating” startups are obviously not there yet and not sure could go there for regulatory and legal reasons.
If AI parters won’t be banned or the access to them won’t be severely restricted (i.e., AI partners only available for clinically diagnosed psychopaths, people in severe depression, people over 30 or even 40 years old, or people who have lost their spouse, like in that Black Mirror episode, and some other rather special cases like these), I don’t see how the “AI partner” technology could be turn out anything than a disaster for the society, on the scale of the social media disaster (if not worse), albeit of a different kind.
I think in the section “People might neglect real romance”, the author fails to extrapolate in the future, as Raemon pointed out, and fails to see through realistic psychological implications.
We are talking about AIs that will be 1) extremely beautiful (like Caryn Marjorie who made a digital copy of herself), 2) smarter and more eloquent than the human who is partnering with the AI; 3) always extremely affirmative, flirting, attentive, etc.
Humans will fall in love with these AIs. Once they are in love with AIs, many (maybe even most) humans won’t get off the hook “just because they are bored” or because they would think “it’s time to get a real relationship and family” (see this comment). When humans are in love they usually don’t look to end their relationship, right? Also, integration of vector databases over the history of all chat with the AI will give them essentially unlimited memory and the fact that AIs “don’t remember the relationship history and shared memories” and thus humans become disenchanted because of that won’t be a problem. Even humans who will manage to get off the hook may find it hard to form human relationships or being satisfied in them afterwards.
So, why this is will be a disaster for the society? Because 1) total fertility rate will plummet even deeper than it is today in most developed countries and in Asia, and because 2) the fabric of the society (woven through relationships) will become thinner (cf. the intersubjectivity collapse). The second problem could be partially ameliorated if parner AIs will be learning and talking to each other, as well as to other people than their partners, i.e., will be actual members of the society, but “AI dating” startups are obviously not there yet and not sure could go there for regulatory and legal reasons.