Robots are chattering at us all the time these days, from TikTok reels, to Siri, Alexa and beyond. For some of us, this new kind of communion has become seamless, like second nature. And it’s true that some robot narration is almost uncannily close to human, though much remains so bad it’s nearly comical, as if the prospect of AI sentience is a total fantasy.
Despite how we might feel about these now commonplace encounters, the fact remains that robot narration can facilitate a world of tasks, and not just for people with their hands full. For those among us with disabilities, or for people tasked with learning a new language like the millions of ELLs in our schools, narration can be an essential tool to access written content they might not be able to otherwise.
The quality of the robot narration we encounter might not seem important. A word is a word, a sentence a sentence. If a narration tool communicates the material terms of the content, it must be working, right?
Well, not exactly.
All narration is not created equal.
A recent study showed that very young students who listened to a story narrated by two different robot voices, one with a flat affect which mimicked the classic speech-to-text narration more closely associated with our idea of robots, and one with a more warm, expressive affect that closely followed human speech patterns, had very different outcomes in vocabulary and narrative retention.
The students, surprisingly, seemed to favor neither voice. But, when tested later, those who were narrated to by the expressive voice showed more engagement, comprehension and better recall of vocabulary words and narrative arc than their counterparts. They also showed more physical engagement based on body language and facial expression.
It makes sense! We have an intuitive affinity for human warmth; speech patterns that mimic what we know. Expression and speech are multidimensional, and robot narration is a valuable tool. But when it comes to leveraging narrative technology for education tools, quality remains a key component.