Kashmir Hill took a look into AI chatbot companions. Specifically, people using ChatGPT to create companions. This read is well worth your time:

"She Is in Love With ChatGPT" (NYT)

My take:

1/ As an AI practitioner, I can't help but consider the risk/reward tradeoffs. Especially when I see how AI companies often release products before they are ready. And how there is so little regulation to hold them to account. That said ...

... 2/ As someone who grew up with tech, I wonder how soon "talk to your AI companion" will be normalized the same as "text-based chat with online, pseudonymous randos." Because that, too, was much maligned in its early days. (Maligned for very different reasons, but maligned just the same.)