There's this small friction I've been running into with AI tools lately. Whenever I want to use an agent or assistant, I have to @ mention it every single time. It sounds trivial, and maybe it is, but after the tenth @ in a day it starts to feel tedious.
It's not that these tools don't work. They do. But the mental overhead of remembering to summon them explicitly adds up. I catch myself thinking "I should've just done this myself" not because the AI was slow or wrong, but because the interaction pattern felt heavier than it needed to be.
The interesting part is that this isn't a hard problem to solve. Ambient AI that knows when to jump in without being poked exists in some contexts. Maybe the real question is why more tools haven't adopted it, or whether the tradeoff (less control, more distraction) is actually worth it for most workflows.
Anyone else feel this friction, or have you found tools that handle this more gracefully?
Top comments (0)