J
Joshua Rammacher
Two related issues that slow down the Ask AI user experience:
  1. The text input is locked while the AI is responding. You can't type your next message until the response is fully complete. If something comes to mind while you're watching the response stream in — a follow-up instruction, a correction, additional context — you have to wait, hope you remember it, and then type it after. In longer responses this can be a minute plus wait just to start typing. Every other major AI chat interface lets you draft your next message in the input field while the current response is still generating.
  2. There's no way to stop or cancel a response mid-generation. If you accidentally hit Enter or Send before your prompt is ready, or realize mid-response that you sent something wrong, you're stuck waiting for the full response to finish before you can do anything. Depending on the length of the response, that's wasted time and wasted context — especially if the incomplete or accidental prompt causes the AI to go in the wrong direction, which then affects the rest of your conversation.
Both of these are standard features in Claude, ChatGPT, Gemini, and pretty much every modern AI chat interface. The expected behavior would be:
• Keep the input field active during responses so users can draft their next message while waiting
• Add a Stop / Cancel button (where the Send button is) that immediately halts the response mid-stream
• After stopping, let the user send a new message right away without needing to wait for anything to finish
These are small UX improvements but they make a huge difference when you're doing real work inside Ask AI and moving quickly between prompts.