What does “remove this” refer to?

Seeing the screen does not always identify what the person means.

A grocery basket with only the paper bag highlighted in lime makes the intended object explicit.

You’re ordering groceries. Your basket has several items and a bag that costs extra. You tell the assistant to “remove this,” meaning the bag. It removes the item you’ve just added instead.

The assistant could see the screen, but that wasn’t enough to understand the instruction. Now you have to put the item back and check whether anything else changed. If the same commands can affect payment or the whole order, the cost of a misunderstanding grows quickly.

So it needs to be clear which object the command applies to: select the bag, and the assistant works with the bag. If nothing is selected and there are several possibilities, it should ask: “Remove the bag?” After the action, it should show exactly what changed.