OpenAI Changed TOS

[anthony]

Just FYI, the chatGPT integration may disappear here in the coming week or two, as it seems OpenAI have decided to change their TOS on health related sites, now requiring AI integrated answers to have a psychologist checking them to comply with their new TOS. That renders it useless for this sites use if they enforce this, which they have notified me already that this implementation may be suspended in the coming weeks.

 

[arfie]

TOS = terms of service?

 

[Sideways]

Clinical psychologist?

 

[Ecdysis]

Oh no... Will we be able to "keep" our previous work with it so far?

 

[anthony]

Yes, yes and yes.

 

[anthony]

There are a couple of options I am looking into... hoping one will solve the issue and check content prior to sending to GPT for violations in order to comply with their TOS.

 

[Charbella]

There are a couple of options I am looking into... hoping one will solve the issue and check content prior to sending to GPT for violations in order to comply with their TOS.

it works. I’ve really enjoyed using it.

 

[anthony]

Worst case scenario, I implement Googles Perspective AI system, which is a pre-moderation system that I can implement on a topic by topic basis. Example, I can implement pre-moderation on the AI topics where anything discussed about suicide, self-harm, etc, is discussed, then it makes you change such things prior to allow the content to be submitted.

The obvious is why I haven't used that system and why I use humans here - trauma contains lots of distressing shit that is upsetting in many contexts, but here, its just normal normal, part of our lives. Adding AI moderation system like Perspective AI here, well... it would cause a lot of false flags, BUT, I can limit it on a topic by topic basis and can limit within the topic only certain elements within a set of pre-defined contexts.

 

[Dadzilla]

I just paid for a year of it, not knowing it was disabled. May I get a refund?

 

[anthony]

I just paid for a year of it, not knowing it was disabled. May I get a refund?

Done.

 

[Dadzilla]

Done.

Thank you!