Adding a "Regenerate with Different LLM Model" button to every response. This button would allow users to request a new response generated by an alternative LLM model. The feature would be integrated to the existing interface, allowing users to use different LLM for the same prompt
- Enhanced UX: Users could easily switch between models to find responses that best suit their needs, without having to start a new thread and switching between them.
- Increased Efficiency: Users can quickly compare responses from different models, saving time in finding the best answer.
- Potential Learning Opportunity: Users could learn from the differences in responses, gaining insights which is the best model for them/their task
What are the exact ways that you see a user using your proposed feature?
- User: prompts something
- LLM: I don't know/can't do that/garbage answer
- User: Clicks "regenerate with x LLM model" button
- LLM: Better answer
- User: continues using the LLM model that was used to start the thread / If the other LLM was much better, user would just use the button under to the prompt as you currently can
See mockup below
ChatGPT UI below
