Suggested by
Andy Edmonds
over 1 year ago
"Turn off notifications for WhatsApp", "Move the 2nd monitor to the right of the first", "Decrease screen resolution"
Settings UI's are inefficient and full of non-real world language. LLMs should be able to translate real human input into actions in settings and preferences UIs.
Anthropic, and open source projects, are rapidly enabling LLMs to control computers. Google is beginning to offer voice control in their home automation suite and Apple is eyeing the same market. Taking on the class of settings UIs with a small LLM that can run locally could beat the encumbered giants to a super compelling end user experience.
With giants Google and Apple chasing smart home voice UX, their market is limited to home automation consumers, potentially allowing a more immediate broader market play to gain a first mover advantage in framing how voice control is going to work for humans. I'd assert Alexa and Siri have done little to inform that future.
Are you interested in addressing this Unmet Need?
I like to think of this problem in terms of a usage problem, where the actuall efficency of a tool /software behind a interface lies in our ability to use it correctly beyond the in place preferences. Simply said, if you pair information the user provides through answer on target question in behavior, background, use case etc on the tool on a volunteer basis and pair it with generative UI (see more here from Google Research: https://research.google/blog/generative-ui-a-rich-custom-visual-interactive-user-experience-for-any-prompt/) you could solve the current inefficiencies and non-real world language in the UI/UX of tools/software. Hope it helps
Luminary @ Lionessuccess
Hey @Andy Edmonds !
It has been a month now; what's the fix?