Do you remember getting annoyed every time Clippy, Microsoft Office’s friendly animated paper-clip, tried to help you draft a document? Or do you now feel like Siri is just a little bit condescending? You’re not alone. According to a study published in the Journal of Consumer Research, most of us really don’t like receiving help from anthropomorphized digital assistants.

Researchers decided to investigate how consumers respond to digital assistants with human or non-human features. They recruited several hundred students to play a series of puzzle games on a computer. During each game, when participants got stuck or confused, they could click on a help icon which would either reveal a smiling face or a faceless interface to give them advice.

The team found that participants felt less autonomous, or less in control, when given help by the anthropomorphized interface, and enjoyed the game less as a whole. Even though the help given was the same regardless of interface, participants couldn’t help but feel a sense of unpleasant dependence when their helper seemed more human. On the flip side, researchers found that the faceless interface was seen as a tool and so did not detract from participants’ sense of autonomy.

The researchers wrote, “In the current research, consumers enjoyed a computer game less when they received assistance from a computerized helper imbued with humanlike features than from a helper construed as a mindless entity. We offer a novel mechanism that the presence of an anthropomorphized helper can undermine individuals’ perceived autonomy during a computer game.”

That is, while we all knew that Clippy wasn’t functionally any different from using the help menu in Microsoft Office, it was hard not to feel a bit annoyed (or, in the study’s terms, a bit like our “autonomy was being undermined”) when an overly cheerful paperclip popped up and started giving us unsolicited advice.