There are 641 plugins in the store today. Each of them has a brief description only and that makes it very difficult for a user to know which plugins might be useful.
I suggest OpenAI make a plugin to recommend other plugins. Technically, it would be very easy because the ChatGPT Retrieve Plugin can be easily customized to serve that purpose. I could even volunteer to make that happen. Unfortunately, however, the current OpenAI plugin policy would make it almost impossible because the would-be plugin would have to carry words like “ChatGPT”, “Plugins”. Otherwise, nobody would know what the plugin is about. But those words are banned for developers. So, OpenAI itself should develop a plugin like that.
1 Like
Hello, in fact, there is already such a plugin called “pluginpedia” that can meet your needs.
Thanks for your info. Tried it and it was not good. It behaves like an early 90s keyword-based search engine. I am pretty much sure it is not built on the ChatGPT Retrieval Plugin, which supports semantic search.
Just imagine the entire web will be transformed by plugins. Semantic search for them will become imperative.
I just notice the “pluginpedia” has both banned words in its name and description. I am glad that was not deemed to have violated the rules by OpenAI reviewers, because it was approved and published.
So, I might as well build a better plugin for plugin semantic search.
A friendly suggestion for plugin developers: you should perhaps attach a hyperlink in your plugin description to point to your website where you can have a detailed description of your plugins. The more detailed your descriptions are , the more accurate the semantic search for your plugins is.