I suggest that it is possible to add more context windows, like a plus button +(Tokens) where you can add more tokens, I know it is possible to delete and continue but in the case of large projects it would be interesting That the user could add more things, I know that the model engineering is already predefined for the maximum of 1 million, but I believe that with time in more sophisticated projects and prompts this window It should get bigger by attracting more users who don’t lose their interaction and projects. Or increase the context window to 2 million as it was some time ago with model 1.5.
Another thing when we create a prompt everything is thrown to Google Drive directly in the AI Studio folder, it would be interesting if each prompt created a subfolder within this with its respective files This provides more organization to the prompts/projects.
They could also save the prompts in json or csv as it was in the past, this makes it easier to port these files.
I know we used the model for testing but the voice reading ability of the model would be very good.
Sometimes the conversation is not saved even when the internet comes back, this causes us to lose interactions and responses. In OpenAI models, we have the continuation of these conversations as if there were a ghost cache which continues the serious conversation interesting to put this on the platform instead of resetting the answer.