While voice assistants initially seemed like a convenient, futuristic way to get information and perform basic tasks, they’ve only just graduated from that role. And due to the lack of evolution, voice assistants are surrounded by uncertainty. For example, Google shut down third-party Google Assistant smart displays and reportedly moved assistant manpower to Bard. But while Google Assistant and Google’s experimental Bard chatbot currently feel like different products with different uses, Amazon dreams of uniting its generative AI efforts with its struggling Alexa business.
It’s no secret that ties are tightening at Amazon, fueling interest in making Alexa a strong source of revenue. Alexa reportedly lost $10 billion by 2022, according to an Insider report, and had failed to engage users enough in ways that would make Amazon money. Amazon is also undergoing its biggest round of layoffs, announcing last week that it is discontinuing Halo fitness and sleep trackers.
Can Generative AI Generate Alexa Revenue?
Amazon reportedly tried to include more AI in Halo before killing it off, as if trackers used a smartphone camera and computer vision to analyze user training data and share it with Amazon. We were reluctant to trust Amazon with such AI use; however, Amazon is reportedly shifting some of that invasive AI energy to Alexa.
An Insider report on Tuesday cited a “leaked document” titled “Alexa LLM [large language model] Entertainment Use Cases.” It reportedly details plans to make Alexa more able to “think versus pull from a database.”
The AI, an Amazon spokesperson told Insider, is not based on an open source model like others in Big Tech use, but rather on a proprietary LLM called the Alexa Teacher Model. Alexa has been using it for years, but Amazon is building “new models that are much bigger and much more general and capable” to make Alexa “more proactive and talkative,” the Amazon rep said.
The internal document, Insider said, provides an example of what this enhanced Alexa could do. Alexa is seen creating a bedtime story using a prompt from a child, such as “cat and a moon.” Amazon seems keen to use cameras to aid its AI, with Insider reporting that Amazon is investigating using an Echo Show smart display camera to identify toys the child is holding and incorporate that into the story. However, this kind of intimate data collection for the use of voice assistant skills would probably raise concerns. Earlier this week, The Verge reported that Amazon employees have “voiced a pause” about including computer vision in the Halo subscription service.
This storytelling feature can generate revenue by encouraging business partnerships. The leaked document reportedly mentions the hypothetical child holding a toy belonging to Olaf Frozen that Alexa would add to its story, naming Lego “and others,” according to Insider, as potential partners.
Amazon is trying to monetize through Alexa partnerships with Domino’s and Uber, among others; however, this newly reported possible use suggests Amazon is considering questionable techniques (hopefully requiring user consent) in the name of a unique, generative AI experience. The internal memo is said to discuss making the stories interactive by asking the user to create a corresponding illustration on an Echo Show display or add to the AI story.
The leaked memo also looks at ways to use Fire TVs to increase Alexa’s helpfulness, Insider reported. Fire TV sets and streaming devices reportedly sell well for Amazon and have the potential for revenue-boosting experiences that are more difficult to replicate with, say, smart speakers without screens. This includes allowing users to interact with digital content while connecting to Prime Video and third-party streaming services.
