How to automatically import ChatGPT (GPT-4) conversations into Fibery?
There is a chrome app that allows importing a complete conversation into Notion.so.
Anyone found a way?
Why not use Fibery’s AI chat (GPT-4) and click on insert?
Else you can use Zapier / Make / etc. or API’s.
It depends on you use case if that’s helpful. Can imagine that you only what to import a conversation if it really contains the answer you were looking for.
In my case Chappie always needs a bit time before he understands me completely.
I’m a heavy ChatGPT user thus their monthly cost is cheaper than the OpenAI API that Fibery uses.
Zapier, Make, etc do not have OpenAI GPT4 integrated yet as far as I can see.
The idea is that I want a compolete chats that change to be automatically imported on click or that this just goes automatically for all chats.
API is not available yet so I don’t think it’s possible to import your conversations into Fibery
To make it work we should create ChatGPT Plugin.
Some feedback from chatgpt 4:
HuggingFace Integration:
- Versatility: HuggingFace offers a wide range of pre-trained models, including GPT-3, GPT-4, and others, providing flexibility to choose the model that best fits your needs. Source
- Cost: HuggingFace’s pricing model is based on input words, which might be more cost-effective for high request volumes compared to OpenAI’s token-based pricing. Source
- Control and Customization: HuggingFace allows for custom model training and deployment, providing more control over model usage. Source
- Partnerships: HuggingFace has partnerships with major cloud providers like AWS, enabling easier integration and deployment. Source
OpenAI/ChatGPT API:
- Specialization: OpenAI’s ChatGPT API is designed specifically for conversational AI, providing robust and reliable conversational capabilities. Source
- Ease of Integration: OpenAI provides a managed service with a REST API, making integration simpler and more straightforward. Source
Meta LLaMA:
- Open Source: LLaMA 2 is open-source and free for research and commercial use, providing a cost-effective option. Source
- Fine-Tuning and Integration: LLaMA 2 is ready for fine-tuning on platforms like AWS, Azure, and Hugging Face’s AI model hosting platform. Source
- Architectural Improvements: LLaMA 2 includes architectural tweaks like Grouped Query Attention and is pre-trained on a large corpus of 2 trillion tokens. Source
In conclusion, the choice between HuggingFace, OpenAI, and Meta LLaMA depends on your specific needs, such as the level of control, cost-effectiveness, ease of integration, and the specific AI capabilities you require.
We would also greatly appreciate the ability to sync ChatGPT messages to Fibery.
Another potential addon would be related to furthering the development of the Fibery Extension or a standalone Fibery <> ChatGPT extension to support saving ChatGPT messages.
We don’t use Notion (*anymore), though the thought had crossed my mind to use this extension and Sync Notion > Fibery purely for this.
Not as great as native sync, but may be helpful or useful for many, and I imagine it may be faster to support ChatGPT history saving to Fibery quicker than a native plugin.