ChatGPT Mobile App Redesign: conversational affordances and session memory

AI · 6 min read

The redesign introduced clearer session organization: named chats, pinning, and search that surfaces past prompts and outputs. Threading helps users treat conversations as projects rather than ephemeral queries. Memory toggles and context controls are surfaced near the input area, letting users decide what the model can recall across sessions.

Multimodal inputs (images, audio) are integrated with dedicated buttons and inline previews, while responses use compact cards that expand for detail. The input UI supports partial prompts and multi-turn templates, helping users build longer prompts without losing context. Latency management uses progressive rendering and intermediate feedback to avoid blank screens during longer generations.

Privacy and control remain central: users can clear chat memory, manage training opt-outs, and export conversations. The mobile redesign demonstrates that conversational AI needs explicit workspace management, clear consent controls, and affordances that let users move from quick answers to sustained projects.