Sources: Anthropic’s AI Powers Complex Tasks in Amazon’s New Alexa+
According to sources familiar with the matter, Anthropic’s Claude large language model is handling the majority of complex tasks for Amazon’s new Alexa+. However, Amazon has stated that this information is “false.”
Amazon, a leading investor in the AI startup Anthropic (founded by former OpenAI employees) which has agreed to invest $8 billion, has been relying on Anthropic’s AI models to power the most advanced capabilities in its new Alexa devices, two sources have confirmed. These individuals requested anonymity because the information is confidential.
This week, Amazon unveiled its long-awaited revamp of its decade-old devices, introducing the new “Alexa+” service. Users will be charged $19.99 a month, or can access it for free as an Amazon Prime member. Early access is expected to start next month.
During a recent demonstration, Alexa+ was able to make dinner reservations, order groceries, and book an Uber – tasks that were mostly impossible with previous versions. Alexa, once a leader in natural language and machine learning, has been trailing the competition since the emergence of advanced generative AI chatbots such as OpenAI’s ChatGPT, which have rapidly evolved beyond simple text conversations to generate audio, images, and video.
While Anthropic declined to comment, an Amazon spokesperson wrote in an email: “In fact, over the past four weeks Nova handled over 70% of conversations — including complex requests,” the spokesperson wrote. “That said, from a customer perspective this doesn’t matter — both are excellent models and in there to deliver the best experience for customers.”
The spokesperson added that, “Alexa+ is always using the best model for each task,” due to the architecture that Amazon designed.
At the event, Amazon CEO Andy Jassy described the overhaul as “rearchitecting” the brain of Alexa.
In addition to the approximately $8 billion investment in Anthropic, Amazon has been developing its own AI models, including the Nova series, which was rolled out late last year. Amazon offers Anthropic’s Claude model as part of its Amazon Web Services Bedrock offering, providing customers with access to a variety of AI options. This allows them to choose between Amazon’s Nova and Titan models, as well as Mistral and others.
According to sources, Amazon utilized Bedrock to power Alexa. The sources stated that Claude has been the model tackling the more difficult tasks showcased at the devices event in New York this week, handling queries that require greater “intellectual heft.” One source noted that the company’s AI models, still in use, are handling tasks that require less reasoning.
One source added that as part of Amazon’s initial investment into Anthropic, Amazon was able to use a certain amount of Anthropic capacity for free, over an 18-month period. Now that the deal has expired, the companies are renegotiating their agreement.
“We pick the model that’s right for the job,” Panos Panay, Amazon’s senior vice president of devices and services and the leader of Alexa’s redesign, told CNBC in an interview on Wednesday. “We use Amazon Bedrock — Alexa picks the right model to get the job done.” Panay, who called Anthropic an “awesome” partner, described the company’s foundational model as “incredible.”