Why can’t Alexa users have sophisticated, nuanced voice interactions with Amazon’s assistant, similar to text-based conversations with OpenAI’s ChatGPT?
That was one of my questions for Rohit Prasad, the senior vice president in charge of Amazon’s Alexa business, during a wide-ranging conversation on the latest episode of the GeekWire Podcast. The rise of generative artificial intelligence is raising expectations for AI in general, and Alexa is no exception.
Of course, as one would expect from any senior Amazon executive, Prasad first made it clear that the company focuses on customers, not competitors.
But in this case, he said, the comparison is fair to neither ChatGPT nor Alexa.
These are two products at different points in their respective timelines and evolutions, serving different purposes, and operating in different contexts, he said. Interacting with a keyboard and screen is very different from a hands-free voice conversation across a room.
“ChatGPT can’t do a lot of the things that Alexa can do today,” he said, adding that Alexa’s capabilities include “giving you trusted answers without hallucinating facts.”
He continued, “Alexa is orchestrating or interacting with thousands of services in real time, instantaneously — your music, your books, your video, your smart appliances … And what it makes seemingly simple is incredibly complex. Because underneath, it has more than 30 machine learning systems working together to give you that outcome in less than a second, often.”
In short, he said, “that’s a very different interaction paradigm.”
Prasad joined the Amazon Alexa team a decade ago, working initially on far-field speech recognition. He went on to become Alexa’s chief scientist, and his role expanded in the past year to lead the Alexa business. That puts him in charge of what he calls Amazon’s “North Star” for Alexa, seeking to create an indispensable personal assistant.
“This is a super exciting time in AI. I’m glad that AI is back,” he said. “The fact that our customers are interacting more than ever with Alexa is a testimony to the fact that our North Star is working. And what’s happening with large language models and other companies is a very good thing for the AI industry as a whole.”
Amazon CEO Andy Jassy hinted at Amazon’s plans for Alexa in April, saying that the company is working on a new large language model that’s “much larger, and much more generalized, and capable,” with the potential to “rapidly accelerate our vision of becoming the world’s best personal assistant.”
Listen above, or subscribe to GeekWire in Apple Podcasts, Google Podcasts, Spotify or wherever you listen.
Related coverage
- Amazon’s Alexa and the rapid evolution of AI: A conversation with Alexa chief Rohit Prasad
- Amazon unveils Echo Pop, new Echo Buds; reports >500M Alexa device sales as AI upends market
- Alexa vs. ChatGPT: Amazon building new AI models to make assistant more proactive, conversational
- Amazon previews Alexa capability that synthesizes a person’s voice from less than a minute of audio
