Yeah, we already started migration to it 
1 Like
@mdubakov
Got this email from OpenAI
So after migration the Fibery in data will not be used outside Fibery, correct? 
2 Likes
What do you mean by “after migration”?
That was regarding this post. It wasn’t clear for me if the first set-up did or did not shared data for model training. But if I understand the email of Open AI correctly this isn’t the case. Correct? 
We already migrated to ChatGPT API so this agreement is already in place. BTW, this is also applicable for other models as I understand.
It probably means that you can be more confident that if you use data in your AI request, this data will not leak out and be available to other people using OpenAI.
Kinda answers this question.
2 Likes
Yes! Thanks
Here is how we are going to change Fibery AI Assistant to chat-like experience. Any feedback is welcome!
4 Likes
Side question: Is that a callout block I see?
It is, but development is not started yet (but will be soon)
1 Like
Looks great! Would be great if I can do everything what I do know in Chat GPT in Fibery so I don’t have to leave the app.
I mostly start with a new chat, so with no input at all. Based on the answers that AI gives, it would be awesome to select a text and then create an entity or link an entity.
Can imagine that this is out of scope but that would be a real game changes since my clients & myself use it a lot for research purposes.
Sounds interesting, but can you expand the case to fully grasp it? So you opened a chat, got some text and then what? You want to create an entity from part of this text right in chat?
It’s usually the case when I do research for content. I have two databases: content ideas and published content items.
In the first database I write down the ‘content idea’. Usually I do some research based on keywords etc.
So for example output of AI assistant for that research = 5 possible blog post headlines.
Then I can create multiple ‘content items’ entities based on that outcome.
Normally I do this manually which also works fine. But maybe there are more use cases where it can be handy to create entities
Edit: On the other hand… I think it’s more logical to insert the useful output because you also want to keep that. And then create the entities like you normally would. So for my own use case it wouldn’t be necessary I think.
Some progress
5 Likes
Official release today
2 Likes
A post was split to a new topic: Ask Fibery AI Assistant questions about data in Fibery (databases content, docs content)
Thank you. We are working on context indexing already. As for long docs, it maybe will be solved as a part of index solution, or with larger limits for GPT API soon
5 Likes