I often struggle to find the correct cooking time for pasta and rice. I’d love to just take a picture of the package and ask R1 to set a timer.
Or another use case I shared here; using ChatGPT I let it figure out the remaining charge time for my EV to charge to 80%, based on a picture of the dashboard (works perfectly).
It would be cool if R1 could do that but then actually set a timer for it.
So my suggestion is to support timers from vision mode to start timers based on the visual context.