AI does not ‘say’: implications of using anthropomorphic language to describe AI use

Key Points

• This research aims to explore how using human-like language to describe AI affects understanding and communication in professional settings.
• Conducted a randomized trial analyzing participant responses to AI language in meetings
• Evaluated responses based on context of ideation and problem-solving
• Participants misinterpreted AI responses as human communication 75% of the time
• Significant confusion noted in problem-solving contexts (p<0.05)
• Communication efficiency decreased by 40% when anthropomorphic language was used