Ask HN: Any way to have inbuild second screen for AI to control
TL;DR
Hey HN! I was thinking in windows and linux having running computer use, but rather than full control, a way agent sees the screen like layer of application like z index but then also send the click and insert text there too ? its possible on mac i can see chatgpt has done it but how about windows or linux ? ycombinator.
Nauti's Take
For small teams, this is first an integration question: verify whether the target system officially supports screen capture, overlays, and simulated input, and how permissions are managed. With no substantive replies yet, it remains unclear whether Windows and Linux can support a stable, secure workflow or whether each desktop stack requires its own workaround.
Sources