3 ms·
I agree that we should have more third-spaces, less working hours, and generally change our culture to facilitate in-person socialization. But right now in-pers
by armchairhacker 1y ago
I agree that we should have more third-spaces, less working hours, and generally change our culture to facilitate in-person socialization. But right now in-person socialization is hard. Moreover, I suspect some people still won't find friends because, to be blunt, they're generally unlikable; they wouldn't want to be friends with themselves, and don't have the self-awareness or self-discipline to change even when such changes are presented as obvious and easy as possible. Also, there are more bullies then people who want to be bullied, more attention-seekers than attention-givers, etc.
Therefore it becomes another case of "is an AI friend better than no friend?" Furthermore, I think that AI, if done correctly, can not only improve society but actually help in-person socialization: by helping people develop social skills and encouraging them to meet other real people. For example, an "AI companion" that you talk with could analyze your personality to be the ideal friend or SO, then connect you to another real person whose appearance and personality are near what it imitates. An AI that has your attention and trust can also critique your real personality flaws and encourage you to fix them (in a supportive way*).
However, current AI companions are too dumb to be effective; for example, they don't develop accurate social skills or criticize personality flaws, because they're unrealistically affectionate and agreeable. Another problem is that companies aren't directly incentivized to make their AIs give good advice and connect real people, companies are incentivized to make money; and it's probably cheaper to get lots of dumb people to pay for a dumb, agreeable chatbot than invest in a smart chatbot that criticizes them and facilitates its own replacement. There is some demand for the latter though, so perhaps someday, a smaller company could develop an AI companion that is genuinely a net benefit to society; and there may be some people whom even a dumb AI companion improves their lives better than nothing.
* emphasizing that self-improvement will help you in the real world, and reminding you that you still have value and deserve love, but standing firm on the criticism. Not like current AI which fawns over and agrees with everything