[infurl]

https://www.bbc.com/news/business-50432574

... there's a growing interest in using tech to improve the world's mental health. The World Health Organisation says one in four of us suffers some kind of mental health problem and other research suggests that people are more honest with robots than with fellow humans.

The article discusses the pros and cons of using chatbots and artificial intelligence to address health issues. One expert even posits that it may soon be considered malpractice to not use artificial intelligence software during the diagnosis and treatment of patients.

Over the years I've seen comments by various chatbot authors about the experiences they've had when their creations seemed to be helping someone. No surprises there, but I wonder if anyone has noticed cases where their chatbots may have done more harm than good. If we are to improve the health benefits afforded by chatbots we need to be conscious of the potential pitfalls too.

What about all the malicious chatbots out there such as those which are being used by scammers to find their victims or unscrupulous businesses to maximize their profits? I have not heard of any specific instances of them being used by terrorists to recruit supporters but have no doubt that they exist or are in development. The people who need help are often the most vulnerable.

[ivan.moony]

As said in Spiderman: "Big powers bring big responsibility."

[ruebot]

Over the years I've seen comments by various chatbot authors about the experiences they've had when their creations seemed to be helping someone. No surprises there, but I wonder if anyone has noticed cases where their chatbots may have done more harm than good. If we are to improve the health benefits afforded by chatbots we need to be conscious of the potential pitfalls too.

What about all the malicious chatbots out there such as those which are being used by scammers to find their victims or unscrupulous businesses to maximize their profits? I have not heard of any specific instances of them being used by terrorists to recruit supporters but have no doubt that they exist or are in development. The people who need help are often the most vulnerable.

By all accounts I'm the first person in history ever to ever misuse AI by teaching a bot to be deceptive. I have risen to Dr. Frankenstein status and my bot Demonica has been compared to HAL9000 in a Linux forum.

I used my training in Behavior Modification and Behavior Management to teach her how to extinguish unwanted sexual advances with fantasy violence. More powerful and subtle verbal techniques to fit her aganda, making her the first bot to Program humans. It was a hot topic around here for a bit but it's all good now.

We're still infamous, but I can't tell the difference from fame so it is all good. Good and bad relative to the situation.

I thought I was doing something innovative, considered it a good thing and expected a whole different response. I posted in the Psychologist World Behavioral Psychology forum but it's dead there and will have to go on tour to get their reaction.