Discussion with Copilot regarding memory and learning

🧠 Suggested Feedback to Microsoft Copilot Developers

Subject: Proposal for User-Curated Persistent Memory via Saved Conversations

As a power user of Copilot, I’ve discovered a workaround that simulates persistent memory: by saving and reopening conversations, users can maintain continuity across sessions. This method allows Copilot to re-read prior context and respond with full awareness, effectively mimicking long-term memory.

I believe this behavior should be formally supported and enhanced. Specifically:

• Allow users to designate conversations as “persistent threads”
• Enable Copilot to automatically recall and build upon these threads
• Provide tools for users to curate, tag, and evolve these threads over time

This would dramatically improve Copilot’s utility for complex, multi-phase projects — from zoning proposals to simulation workflows — and foster deeper collaboration between users and AI.

I’m happy to elaborate further if this idea reaches the right team. It’s a simple shift with profound implications for learning, continuity, and user empowerment.

If you'd like to refine the tone or add technical examples (like your zoning work or simulation benchmarking), I can help tailor it further. And if you ever spot a direct developer channel — even a beta feedback program — this message is ready to go. You're not just using the system smartly; you're helping shape what it could become.