My reaction to screen-aware AI is completely different depending on which computer we’re talking about

“On my personal laptop I’d probably try a screen-aware AI tool if the privacy model was reasonable. Hotkey-only, visible indicator, clear data handling. That feels manageable.

On my work laptop I wouldn’t consider it. Not because I think it would actually cause a problem, but because the downside is asymmetric. One accidental capture of client information or confidential project details and the consequences are out of my control. I can’t evaluate that risk well enough to accept it.

I suspect a lot of people have this same split, comfortable enough for personal use, hard line for work. Which means even a well-designed tool would only capture a portion of someone’s computing time.

It also means the most valuable use cases, where you’re actually working on something complex, are off the table for a lot of people.

Is this the boundary others draw, or do you apply the same standard to both?”