Hugging Face's logo

Join the conversation

Join the community of Machine Learners and AI enthusiasts.

Sign Up
Back to feed
NatalieY 
posted an update 1 day ago
Post
1,886
Most mobile AI agent demos show the agent controlling a phone screen.

What they don't show: when to not do that.

Aico's Phone Bridge layer routes per action - software fast path (calendar, contacts, clipboard, app launch) when available, HDMI + HID hardware control when not. Sometimes both in the same task.

The interesting design question isn't "can the agent control the screen?" It's "when should it bother?"

Building in public. aico.link
Twu31

The per-action split makes sense. I'd keep the read-back step explicit across both routes: a successful bridge call and a successful HID input are different pieces of evidence from seeing the intended state. That becomes especially useful at a handoff, where replaying an action after a timeout could duplicate something that already succeeded.

·
HuangHan96

Absolutely. That distinction between input execution and observed state is critical. Aico attaches post-action state data (such as visual screenshots) to close the loop, so the agent can perceive the real-world impact before deciding whether to proceed, retry, or hand off.
In this post