On this week’s Hard Fork there were two talking points about AI that I think are being talked about the wrong way.
At one point, someone says that in the future most of our friends will be AI because AI will listen to you better than any human friend does. Later on, they’re talking about ways for Claude to subtly report on a child’s chatbot conversations topics back to the parents, like, “your daughter has been looking into eating disorder stuff” or something.
What baffles me about this is that no one is asking the real question: Why are human beings so shitty to each other that we’d rather talk to AI than to real humans? Why are human parents so shitty that their children would rather talk to AI than have a real and close relationship with their parents? Why is it that, when faced with an AI that shows how flawed humans are and how bad they are at interpersonal relationships, do we see it as a problem with AI and not a problem with humans that we could solve if we really wanted to do so?
Instead of forcing rote memorization on children for a decade or more of their life under the guise of education, maybe we should consider teaching them interpersonal skills, and not the kind that is being taught in schools currently. We’re releasing people into the world who have no real idea how to live with other people, no idea how to talk to other people, and a deep revulsion to any sort of sincerity or vulnerability. AI is giving us an opportunity to reflect on this and instead we’re talking about how potentially dangerous it is.