Multi-Tab Chat Interface, and ability to access archived chats from account

Hello OpenAI Team,

I’ve been an enthusiastic user of GPT for various purposes: co-writing stories, engaging in RPG games with Starfinder mechanics, and seeking help in fleshing out my own game book. Given my experiences, I’d like to propose a few features that I believe could significantly enhance the user experience:

  1. Multi-Tab Chat Interface: Similar to how Notepad++ allows multiple tabs, it would be immensely helpful if GPT could support a tabbed interface within a single session. This would allow users to seamlessly switch between tabs, all while GPT retains the context of each. For instance, when using GPT in a roleplaying scenario, one tab could be dedicated to scene development, another for character sheets, a third for roleplaying, and maybe even one for dice-rolling. The overarching goal is to enable GPT to cross-reference information across these tabs, enhancing its ability to provide cohesive responses.

  2. Access to Archived Chats: If GPT could tap into and reference older chat sessions stored in a user’s account, it would allow for greater continuity. This feature would be particularly beneficial for those wanting to expand on previous story arcs or discussions.

  3. Multimedia Interpretation: This might seem a bit ambitious, but considering the rapid advancements in AI, I wonder if there’s a possibility for GPT to interpret and derive context from images and music in the future? When I’m building lore and stories, I often draw inspiration from visuals and soundscapes. Having GPT understand the essence of a particular image or musical piece (rather than me trying to describe it, which can sometimes be inaccurate) would be a game-changer.

Thank you for considering these suggestions. I truly believe these enhancements could elevate the user experience to new heights!

Warm regards, Danny089

Interesting ideas!

  1. You have a tabbed browser?

  2. The AI doesn’t have the ability for lengthy understanding. You might have noticed it doesn’t even reflect well on long single conversations.
    2a. You can edit a conversation at any point, even one you recall, and this will send the chat in a new threaded direction (look for a button on the right that only appears when you hover). The prior path can still be selected.

  3. AI Multimodality is not currently presented in ChatGPT. You can see image upload and recognition being used in Bing AI chat, but the usefulness is describing what you already see.
    3a. A third-party plugin within ChatGPT plus that could listen to music or other multimedia features would be interesting (like getting the transcription for a video link), but you couldn’t directly give the AI music for such a feature, as it only takes text.

1 Like

:wave: Hello and thank you for your insights! :slightly_smiling_face:

Indeed, I am familiar with the capabilities of a tabbed browser. My suggestions are more about the potential for GPT to evolve in the future and incorporate native multi-tab functionalities to better manage complex narratives or tasks.

I do recognize the current limitations of the model in maintaining lengthy conversations, which is precisely why I’m hopeful for improvements in this area with subsequent versions.

As for AI multimodality, I understand it’s not a feature of ChatGPT at the moment. The mention was indeed an ambitious outlook on how the platform could eventually interact with various mediums. It’s always exciting to think about where technology might head!

I genuinely appreciate your feedback, and my intention was simply to share ideas for potential future growth, rather than seeking immediate solutions. Let’s continue to envision and discuss the endless possibilities of this technology together!