Considering AI Chat Design Implications #559
Unanswered
AllDaGearNoIdea
asked this question in
Q&A
Replies: 1 comment
-
|
Hey @AllDaGearNoIdea Thank you very much for pushing the discussion here. Thank you again. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
A continuation of a thread started here:
#500
macOS launchers traditionally have input on top, results on bottom.
Starting with Sherlock:

Then Spotlight:


Including third-party apps like LaunchBar and Alfred:





Windows history is a bit different, with Start Menus from Vista -> 10 having input at the bottom and results above:
The popularity of smart phones (eclipsing desktop) has seen an ongoing normalisation of placing search fields at the bottom (where they are easier to reach), with results displayed atop.
Starting with Apple Maps:
Then rolling out to to the rest of iOS:

Simultaneously, chat interfaces have become the most popular type of apps (e.g. WeChat in China), increasingly so with the explosion of LLMs. There's an interesting lineage here, with CLIs establishing the top-bottom chronological history standard, then coming full-circle with AI Chat, which is essentially a modern CLI.
From IRC:
To MSN:

To the original iPhone:

Chat interfaces consistently have entry at the bottom, with a top -> bottom chronological timeline of messages:


This forces a decision for a launcher app, if it is to include a chat interface.
The pragmatic (and easy to implement) solution is Raycast's approach, which accepts the two different design idioms, and incorporates the established one for each mode.
Launcher:

Chat:

However, looking at the direction of the industry, I think there's a likelihood that the next computer interface from an AI company, will be bottom focussed. Arguably, Claude Code is this already:
So a potential (and brave) choice, would be to make the entire app bottom focussed.
I'm not sure how I feel about this, but it does make sense the more I think about it.
Another choice, would be to invert the chat chronology (bottom -> top).
This isn't an unseen interface, as email, rss, and social media traditionally place new items (and where applicable, entry) at the top.




Which in practice would look like this:
Personally, I find this idea most interesting, and I think it feels more natural for important things to be at the top, plus it means that new responses can be scrolled through top-bottom, rather than having to see the bottom of the message come in, then scroll back up to read the beginning.
All of this is to say, the way the chat interface currently works in Asyar doesn't make sense to me and I can't think of a single example of it working this way, where the entry is at the top, then the new message enters at the bottom. It's a mixing of metaphors. I'd be willing to hear an argument and case made for it, but I'm unconvinced one can even be made.
I don't know what the answer is here, but I think the options are:
Are there any other options people can think of, or is that a fair summation?
Finally, it may not be true in all cases, but I believe the quality of an app's design to be inverse to the number of settings, because sensible defaults and an intuitive interface negate customisation.
A relevant "tweet" in regard to ChatGPT's recent "upgrade":
"The settings of an app are the clearest tell of taste & vision IMHO"
I think it's important for Asyar that we make a justified, well-informed, and concrete decision, rather than push this cognitive burden over to users.
Beta Was this translation helpful? Give feedback.
All reactions