Considering what kind of data people are giving to AI, security is an important issue
Cloudflare worker for building AI plugins. https://github.com/team-openpm/cloudflare-basics-ai-plugin
WorkGPT - framework for working with GPT functions. https://github.com/team-openpm/workgpt
Everything is open source.
You make a good point though. We should write a langchain plugin to our API.
I'm really not sure...
EDIT - they are APIs with the neccesary metadata to be used as GPT plugins.
I was confused by the fact they aren't listed in GPT when you're in plugin mode - but that's just the officially vetted list.
(Before that I was confused by the interchangable use of "plugin" and "api" and the the use of "AI" when OP specifically seems to mean "ChatGPT"...)
Are you sure you're not confusing OpenAPI and OpenAI? OpenAPI is a spec for HTTP APIs. Many AI applications are giving language models the capability to interact with OpenAPI documented APIs.
Thanks for the heads up by the way!
* closed
* pay per usage api to make money and have ability to shut down users
* hide their research results
* quickly build up data
* quickly advance with Microsoft in bed
* lobby law makers to regulate it
* ...so entry barrier is higher for competition
* ...so they secure as much pie for themselves as possible
* exploit any other means to increase their dominant position and keep throwing rocks at others if possible
They're open-for-business, not open-source.
The best we can do is ignore them and support alternatives.
Aka:
OpenAI: we welcome regulation
EU: https://softwarecrisis.dev/letters/the-truth-about-the-eu-ac...
OpenAI: no-no-no, not like that
