This week, we have the second of our two-part series about how AI chatbots are affecting your relationships and about the relationships you're forming with the chatbots themselves.

Last week, we heard from listeners who shared some of their deeply private and impactful encounters with AI, a psychotherapist who found more clear analysis from a chatbot

than years of her own therapy, a parent getting pointers for raising a teenage son, someone in the

first crippling phase of mourning who was helped by a chatbot's clear explanation of why

grief can be so intense. But this week we take a close look

at the destruction AI has brought into some of your lives, up-ended marriages, obsession,

breaks with reality. This is new technology that we are interacting with in ways that we've

never been able to interact

with technology before.

By now, you've probably heard of the phrase AI spiral or AI psychosis, the phenomenon of people

getting sucked into private conversations with chatbots, some of which have ended in suicide

for adults and children.

Some of those stories have their origins in Chachipit 4-0,

the now infamous version that came out last spring that was overly sycophantic.

OpenAI made some safety changes to their large language model.

Chachy-P-T is now not as sycophantic.

It's supposed to flag suicidal ideation or delusion and de-escalate instead of perpetual

world building and validation.

But OpenAI is also a company trying to get as many users as possible.

And Sam Altman announced in October that newer models would have more engaging and customizable personalities again.

On X, he wrote, we made ChatGPT pretty restrictive to make sure we were being careful with mental health issues.

We realized this made it less useful slash enjoyable to many users who had no mental health problems.

...