As the holiday shopping season ramps up, Amazon is looking to compete in the consumer AI space by adding new generative AI features to Alexa, its cross-platform, cross-device digital assistant. The ecommerce and technology giant on Wednesday announced Alexa Explore, which is specifically aimed at kids.
Browse with Alexa is designed to provide more personalized and age-appropriate responses to questions from children. The company told Decrypt that it will be included with an Amazon Kids+ monthly subscription and will be accessible on any Amazon Echo device with a child profile.
When kids ask Alexa about animals and nature, Amazon says Alexa's new Explore will respond with fun facts and trivia questions. According to the company, these responses are adapted from “trusted sources” such as the World Wildlife Fund and AZ Animals.
“Our LLM stack allows us to adapt this proven content with age-appropriate vocabulary into natural, kid-friendly responses,” said Arjun Venkataswamy, senior product manager, Alexa Kids. “We started with trivia questions because, in addition to being fun, they use a well-researched teaching method called ‘activating prior knowledge'.
According to Amazon, the updated Alexa will allow children to learn on their own by speaking Alexa wake-up commands and asking questions. The service asks stimulating questions designed to get children engaged and interested in exploring a topic.
“Browse with Alexa includes the same accessibility focus and features as Alexa overall,” Venkataswamy said. “As part of our hybrid human and AI review process, we review all content with Alexa to ensure it is inclusive of children of a wide range of backgrounds.”
Like Google and Microsoft, Amazon has invested in generative AI technology, including a $4 billion investment. In September, the e-commerce company unveiled new AI-powered products, including additions to smart home products.
Adapting generative AI to children has been a touchy subject, but watchdog groups and policymakers are wary of any technology that could be accessible to minors. On Tuesday, 34 states filed a lawsuit against Facebook and Instagram parent company Meta, alleging that it uses its algorithms to encourage children to spend too much time on the platforms and contribute to teen mental health issues.
“Building on our long-standing commitment to protecting the trust of our customers and their families, browsing with Alexa puts trust and security at the forefront,” said Venkataswamy. “We're using both approaches to content protection: adapting content from safer, more secure topics like animals, from trusted sources, and using a mix of human and AI review.”
While AI developers want to protect children online, the question of how AI model data is collected, where it is stored, and who has access to it remains a concern. Tech companies such as Apple and Samsung, the US House of Representatives, and US Space Force personnel and service members have banned or restricted the use of ChatGipt for fear of data leaks and loss of intellectual property.
Last month, Amazon suggested it would train its AI models on customer conversations, reigniting fears from 2019 that Amazon was reportedly enlisting human contractors to listen to Alexa recordings. An Amazon spokesperson clarified that customers can opt out of sharing their data with the company.
“[Amazon] “We've always believed it's important to train Alexa with real-world questions to deliver an experience that's accurate, personalized, and constantly improving to customers,” the spokesperson said. But at the same time, we give our customers control over how Alexa voice recordings are used to improve the service, and we always respect our customers' preferences when training our models.
