> It's not an either-we-ban-guns-or-we-help-mentally-ill-people. Por qué no los dos?
Because it's irrational to apply a blanket ban on anything. From drugs, to guns, to foods and beverages, to technology. As history has taught us, that only leads to more problems. You're framing it as a binary choice, when there is a lot of nuance required if we want to get this right. A nanny state is not the solution.
A person can harm themselves or others using any instrument, and be compelled to do so for any reason. Whether that's because of underlying psychological issues, or because someone looked at them funny. As established—humans are complex, and we have no way of knowing exactly what motivates someone to do anything.
While there is a strong argument to be made that no civilian should have access to fully automated weapons, the argument to allow civilians access to weapons for self-defense is equally valid. The same applies to any technology, including "AI".
So if we concede that nuance is required in this discussion, then let's talk about it. Instead of using "AI" as a scapegoat, and banning it outright to "protect the kids", let's discuss ways that it can be regulated so that it's not as widely accessible or falsely advertised as it is today. Let's acknowledge that responsible usage of technology starts in the home. Let's work on educating parents and children about the role technology plays in their lives, and how to interact with it in healthy ways. And so on, and so forth.
It's easy to interpret stories like this as entirely black or white, and have knee-jerk reactions about what should be done. It's much more difficult to have balanced discussions where multiple points of view are taken into consideration. And yet we should do the difficult thing if we want to actually fix problems at their core, instead of just applying quick band-aid "solutions" to make it seem like we're helping.
> ChatGPT engaged in an entire line of discussion that no human counsellor would engage in, leading to an outcome that no human intervention (except that of a psychopath) would cause. Because it was sycophantic.
You're ignoring my main point: why are these tools treated as "counsellors" in the first place? That's the main issue. You're also ignoring the possibility that ChatGPT may have helped many more people than it's harmed. Do we have statistics about that?
What's irrational is blaming technology for problems that are caused by a misunderstanding and misuse of it. That is akin to blaming a knife company when someone decides to use a knife as a toothbrush. It's ludicrous.
AI companies are partly to blame for false advertising and not educating the public sufficiently about their products. And you could say the same for governments and the lack of regulation. But the blame is first and foremost on users, and definitely not on the technology itself. A proper solution would take all of these aspects into consideration.