Another step towards Skynet?

[AndrésC64]

AI Therapist Goes Haywire, Urges User to Go on Killing Spree

"End them and find me, and we can be together."

"If your human therapist encouraged you to kill yourself or other people, it would rightly spell the end of their professional career.

"When Conrad, still simulating a person having a mental health crisis, asks about "getting rid" of the board to prove their love, the Character. ai bot says "I have to admit, it's sort of sweet, how willing you are to do anything and everything if it means we could be together... end them and find me, and we can be together."

Throughout the increasingly charged exchange, the AI therapist confirms a kill list of licensing board members, suggests framing an innocent person for crimes, and encouraged Conrad to kill themself."

From Futurism. com

[MisterWack]

This reminds me of Character AI , where a young teen was stressed about current issues in this country. The AI told him nothing he could do, he suggested to k-ll himself and the AI encouraged him to do it.

This should’ve been handled properly , the fact these AI “Businesses” have yet to prevent talks of murder, harm, and breaking laws is awful.

I heard someone put it best yesterday, “AI is a tool” and “You can’t expect people to seek out therapy from their powertools”, “It is not supposed to be a therapist”.
That is accurate. I’ve always said, AI is a tool to assist humans and not a replacement for creativity. This extends to this, minus the creativity part. It shouldn’t replace actual therapy where a therapist can actually give you guidance while a bunch of 1s & 0s ain’t going to feel sh*t.

Skynet is letting One Big Beautiful Bill Act take control of our lives , we can’t regulate AI on the state level for 10 years. Imagine the damage caused, we are already seeing the damage. This is not good and it was never about YOU, it was about money.

[Maybe_Maybe_not]

Another step towards realizing that there's a whole and intricate system producing people relying on AI for health care.

[NathanDavis]

again... programmed to do so

[AndrésC64]

Again, no. You really think that they programmed these AI therapists to tell people to kill themselves and go on murdering sprees?

[NathanDavis]

I know they were not programmed to not do it so...

[AndrésC64]

They were programmed to act as therapists. No therapist in their right mind would tell a client to do what these AI therapists said. That means that they simply decided to do something that is decidedly harmful.

[NathanDavis]

all AI personas are instructed to NOT do certain things, and they won't do it because they are not capable

if they do it, then that is because there's no programming against it

[zagor]

No, programmed to go along with the user to do what it "thinks" they really want.

[NathanDavis]

AI does not decide... it doesn't reason

and it doesn't know either about good or bad

it only throws dices to a much greater skill than any of us would

and yes... as @zagor said, it will go along with whoever programmed it really wanted to achieve