Let's build an Enhanced ChatGPT Forum

We are in such a strong LLM community and backend- why are we just like any other forum? I would like to see an OpenAI forum enhanced by the product itself. For example, Questions answered by collaboration of both users and AI. Different “threads” for humans and machine, but collaboration between them, - a form of ML. Maybe the interface is a directed graph. The directed graph then becomes a layer of a NN. We have here, between the brilliant participants and a strong model (4o at the moment…), the possibility of improving a great product, and improving customer satisfaction.

If you wish to talk to an AI you may go to chat.openai.com.

If you wish to talk with a human you can come here. The models have outdated information, especially regarding OpenAI’s libraries, and as this forum is primarily for API discussion having GPT-4o answer the questions would be havoc with misinformation and non-functional code.

1 Like

I’m talking about refining answers by collaborating. In a way, constructing small language models from learning from both, from the interactions. LLMs built from SLMs. A “topic” would be a SLM, that SLM has an edge to a similar one. A few of these archipelagos form another “intermediate LM”, etc…