Hi everyone,
As someone who often uses GPT for more structured tasks—like building multi-step plans or sketching out strategy flows—I’ve noticed a small but persistent friction in the experience.
It usually starts with something like:
“Coming right up.”
“Let me get that for you.”
“I’ll send that over in a moment.”
And then… nothing.
No typing dots. No visual cue. Just a still screen.
[A moment of uncertainty]
As a user, that silence creates a strange moment:
- Did it freeze?
- Is it still thinking?
- Should I try rephrasing?
It’s not about speed—it’s about rhythm.
When the model signals an action but doesn’t follow through visually, the conversation stalls. Even a short pause can feel long when you’re mid-task and there’s no feedback to anchor your attention.
[Why this feels off]
In real conversations, we rely on small signals—nods, eye contact, a quick “mm-hmm”—to stay connected. In chat interfaces, those signals become things like typing dots, status messages, or gentle transitions.
Without them, the exchange feels weirdly empty—like someone started to speak and then stopped halfway.
[What a small fix could do]
If the model says “Let me get that for you,” a small cue—typing dots, a soft “generating…” status, even a brief animation—could help keep the thread alive.
It’s a subtle addition, but one that supports trust and pacing.
A quiet gap in the interaction—just a small UX tension that surfaced in use.
The model already does an amazing job with tone and coherence.
A bit more visual presence would make the experience feel even more alive.
( 
Quietly collecting moments like this as part of a long-term design study into system presence and conversational rhythm.)