Alexa just went from a somewhat ordinary, although reliable digital assistant, of a powerful enigma. Alexa + replaces the section of the original with conversational intelligence, proactivity and real AI. It is a lot to approach and raises a lot of questions. I managed to get answers from Daniel Rausch, vice-president of Alexa and Echo from Amazon and people who helped build Alexa Plus, to obtain answers.
Even if I caught Rausch at the end of a long day, he seemed energized and clearly very proud of the creation of Amazon.
I started by asking him questions about something that has bored me since the launch: how many Alexa + is Claude? Anthropic, Claude’s developer, was presented as a partner in the development of Alexa Plus and listed during the presentation alongside Amazon Nova, the Amazon Great Language model.
Rausch, however, quickly disillusioned from this notion.
“You have seen Amazon Nova models up there and they are certainly where we start on Amazon,” started Rausch, “we are still starting with our own technology.”
Raushe explained that there are “incredible price performance in these models, incredible latency, incredible precision”.
Principles of foundation
It goes further than that, however. Rausch explained that the rocky substratum of Amazon is a kind of basic based on the cloud for all its generative work of AI. Anthropic is a “really important partner,” said Rausch, but the system built by Amazon is “agnostic model”.
“Bedrock’s objective is to serve very competent and ultramodern models,” he said. This means that the system will choose the best models for work.
It is logical that it was also that Rausch is not to identify which models, including Claude, are used when. “With Alexa,” he added, “we have access to the full suite.”
It goes further than that, however. Large language models are not the end or destination. Rausch told me that the Alexa + experience extends and relies on them. He discussed the idea of ”information experts” introduced during the launch event. The models use these experts to collect relevant and factual information. Rausch used the example of his constant requests on baseball, especially the Yankees, but added that the system is intelligent enough to know that, in his house, he only likes to speak endlessly of baseball when his daughter has no interest.
“I would say that the models help orchestrate the global experience and are the foundation and help us build the rest,” said Rausch.
This art
During the children’s children’s section, or the part that described the new “Explore and Stories with Alexa”, I noticed what looked like an art of generative AI.
Rausch confirmed that Alexa + generates this art on the fly on the basis of children’s ideas. During development, Rausch put the tools in front of the children of certain relatives to assess their reactions: “It is super fun. A child describes history, Alexa helps them to explore:” Hey, what would you like to write a story? “” For example, the kid describes a bearded dragon playing a saxophone. “Alexa is drawing creative works of art, asking questions on the path of history:” Where does the bearded dragon live? “Or” what city does the bearded dragon visit? “Of course, children are unlimited in their imagination.
It seems fun. I noticed that the images I saw looked like a generative AI, and Rausch told me that it was exactly what they were. However, he would not reveal which Alexa + generative image model uses. Everything would only say that “it is from the models of the rocky substratum”.
Security
As with any generative, the key to the utility of Alexa + is the data or rather your data. Generative tasks will be treated in the cloud, but Raush told me that everything will be encrypted “in transit”.
“It is incredibly safe and responds to our standard trust practices in general, which include security and deep private life,” he added.
“We are always starting with our own technology.”
Daniel Rausch, Vice-President of Amazon Alexa & Echo
Naturally, this led me to ask questions about children’s technology and the guarantees that Amazon built around the creation of generating images.
Rausch described it as “incredibly sure” and said that there were many guarantees in place to ensure that “children are still safe”.
I know that many companies say that their generative image platforms are safe, but Amazon has history in the development of platforms and systems adapted to children. The “Explore and Stores With Alexa” is an extension of all this work.
Not always a screen
It was difficult to ignore the omnipresent use of Echo 21 Smart Display throughout the Alexa +demo. I naturally wondered about all the Smart Echo speakers, at least those new enough to support Alexa Plus. What will experience with them look like?
During the demos, Amazon actually increased the length of the responses for Echo 21 screens to show what Alexa knows. But nobody wants to stay to look at a speaker while waiting for these answers. Rausch told me that they had designed Alexa Plus to provide more concise responses to the speakers. All of this is customizable.
“The idea is that we have chosen the right types of answers and the right interactions for the device and the modality in which you are,” he added.
Is Amazon ready
When Amazon book Alexa + in March, this could mean millions of Echo owners suddenly having generative AI conversations with the most recent chatbot on the block. It is potentially a large computer lifting. Is Amazon ready?
“Yes,” smiles Rausch, “it’s very nice to have AWS on Amazon.”
Amazon’s massive cloud computing platform supports countless websites and services and will now provide bandwidth for Alexa +. However, it will not be all everywhere, at the same time. Amazon plans to deploy Alexa + in waves.
And now, I feel like I understand Alexa + a little better when these waves have hit the digital shore.
