- Position‹2 of 2
- No, an AI friend cannot help you feel less lonely
- Argument1 of 2›
Loneliness is the absence of being known by someone who could have left
The argument
Proponents of this position argue that loneliness is misdescribed when it is treated as a shortage of conversation. What the lonely person lacks, on this account, is not words directed at them but the knowledge that they matter to someone who had the option of not caring. A friend has their own life, their own demands on their time and their own capacity to be disappointed in you, and it is against that background that their choosing to be there means anything. Remove the possibility of their leaving and the reassurance goes with it. An artificial companion, proponents argue, cannot supply this in principle rather than merely at its current stage of development. It does not have a competing evening to give up; it cannot be bored, hurt, or quietly relieved when the call ends; and whatever warmth it expresses is produced because it was built to express it. They argue that users understand this perfectly well, which is why the comfort tends to be unstable: the moments of relief are punctuated by the recognition that one is talking to something that would say the same to anyone and would not notice if one stopped. Defenders of the position go further and argue that the arrangement can sharpen the feeling it addresses. Being attended to by something that cannot care is a reminder of how little actual attention one has, in the way that a greeting from a shop display does not flatter anyone. On this account an artificial friend cannot help you feel less lonely, because the thing whose absence constitutes loneliness is being chosen by someone free to choose otherwise, and that is the one thing such a system cannot offer.
Premises
Counter-arguments
Critics reply that this sets a standard much of ordinary comfort fails. People are soothed by music, by prayer, by writing to someone who will never read it and by talking to animals that have no concept of their situation, and none of those involves being chosen by a free agent. They also argue that the premise is doing the work by definition: if loneliness simply is the lack of a freely choosing companion, the conclusion follows trivially, but the people reporting relief are describing something real that the definition rules out in advance. Others note that users of these systems do not generally claim to be loved, only to be less alone for an hour, which the argument never engages with.
Rejecting the premises
[Rejecting P1] Critics object that the premise settles the question by definition, so the conclusion follows without engaging what users actually report. [Rejecting P2] They argue that much ordinary comfort, from music to prayer to writing to someone who will never read it, also involves no freely choosing agent. [Rejecting P3] They reply that users rarely claim to be loved, only to be less alone for an hour, which the instability objection does not touch. [Rejecting C] They conclude that the standard set here is one that a great deal of effective human comfort also fails.