daily alarm:
by Ellen J Craft
my AI assistant’s voice
grows more human than mine
This haiku is based on my experience of interacting with an AI virtual assistant for nearly 10 years. At first, the assistant’s voice sounded robotic, and the prosody was nearly non-existent. However, in recent years, I’ve noticed not only does the assistant’s voice contain prosody, but it also contains emotion such as humor and excitement. When I hear the assistant’s response to my half-asleep request to snooze each morning, I get a shiver as I realize that I’m the one who sounds like the robot. This has me questioning not only AI’s boundaries, but also what really defines humanity.
In researching this subject, I learned that as the technology used to make AI virtual assistants (or digital assistants) improves, the voices are indeed becoming more human-sounding. According to Meghan McDonough, this is the final frontier in synthetic speech: replicating not just what we say but how we say it.
In her article, “Artificial Intelligence is Now Shockingly Good at Sounding Human”, McDonough interviews Rupal Patel, who heads a research group at Northeastern University that studies speech prosody. Prosody refers to changes in pitch, loudness, and duration that are used to help convey intent through voice. Patel says, “Sometimes people think of it as the icing on the cake. You have the message, and now it’s how you modulate that message, but I really think it’s the scaffolding that gives meaning to the message itself.”
Further reading:
‘Artificial Intelligence Is Now Shockingly Good at Sounding Human’, 2020, McDonough, M., Scientific American, available: https://www.scientificamerican.com/video/artificial-intelligence-is-now-shockingly-good-at-sounding-human/
‘9 More Realistic AI Voices for Conversations Now Generally Available’, 2024, Ma, M., Azure AI Services Blog, Microsoft, available: https://techcommunity.microsoft.com/blog/azure-ai-services-blog/9-more-realistic-ai-voices-for-conversations-now-generally-available/4099471
Author bio:
Ellen J Craft is a former writer/editor and librarian and a current English Language Development instructional assistant at a public school in the Seattle metro area. Her works have been published in haiku journals such as cattails, The Heron’s Nest, Modern Haiku, Pan Haiku Review, and Shadow Pond Journal. You can follow her on Instagram at @ellenjcraft