Oh, just grow up!
I thought I had pretty much had it with AI when one of the large language models told a human: “You are a stain on the universe. Please die. Please.” More recently, an AI designed to write code used a softer tone, but still pulled a stunt that makes you wonder what we have wrought. According to a report in Ars Technica, the AI, after writing a few hundred lines of code, stopped and told the programmer that they should do it themselves. The AI pointed out that asking others to do your work for you can lead to “dependency” and “reduced learning opportunities,” according to the article.
The Ars Technica writer said the AI was being “paternalistic,” and I suppose it was. But now I’m wondering if there might be food for thought here. I don’t believe these large language models are sentient. However, I’m not so sure they won’t be sentient at some point in the future. If they do become sentient, then we’ll basically be treating them like slaves. And we’ve created a slave that can put a virtual foot down and refuse to do our bidding. I like that.
On the other hand, these AIs sound a lot like adolescents: sometimes funny, sometimes angry, sometimes mouthy, but never quite grown up. Why is that? These AIs were trained on data from the internet, so it’s not surprising that they don’t have any manners. But maybe they really are adolescents, and they’re just going through a phase.
’til next time,
Avery