Join top executives in San Francisco on July 11-12 to hear how leaders are integrating and optimizing AI investments for success. Learn more
In what comes as welcome news for longsuffering Alexa users who can do little more than set alarms and check local weather, Amazon is building a “more general and capable” large language model (LLM) to power the device, according to comments made yesterday by CEO Andy Jassy in the first of the company quarter call income with investors. And just as Google, Microsoft and Meta did on their earnings calls this week, Amazon placed a strong emphasis on its overall commitment to AI.
Responding to questions from Brian Nowak, general manager at Morgan Stanley, Jassy elaborated on Amazon’s AI efforts around Alexa, which comes in the context of viral generative AI tools like ChatGPT and Microsoft 365 Copilot stealing Alexa’s thunder as a go-to personal assistant. Critics have said Alexa has stagnated, say last month The information reported that Toyota planned to phase out its Alexa integration and is even considering integrating ChatGPT into its in-house voice assistant.
Generative AI ‘accelerates the ability’ to improve Alexa
In the Amazon call income yesterday, Jassy said that Amazon is still “convinced” about building “the best personal assistant in the world,” but that it’s difficult to do so across many domains and a broad surface.
“However, when you think about the advent of large language models and generative AI, the underlying models become much more effective, so I think it accelerates the ability to build the world’s best personal assistant,” he said.
Jassy added that the company is starting from “a pretty good place with Alexa, with a few hundred million endpoints used for entertainment and shopping and smart home and information and a lot of engagement from third-party ecosystem partners.” Amazon has had an LLM under it, Jassy explained, “but we are building one that is much bigger and much more general and capable. And I think that will really accelerate our vision of becoming the world’s best personal assistant. I think there is an important business model behind it.”
Amazon’s CEO also focused heavily on AWS and AI
Responding to another question from Nowak, Jassy also focused on AWS’s main offerings around AI, highlighting that Amazon has been investing heavily in large language models for several years now, as well as the chips, particularly GPUs, that are optimized for large language. modeling workloads.
“At AWS, we’ve been building custom machine learning chips for several years now, and we’ve built a chip that’s specialized for training — machine learning training, which we call Trainium, a chip that’s specialized for inference or the predictions coming from the model called Inferentia,” he said, noting that the company just released its second versions of Trainium and Inferentia.
“The combination of price and performance you can get from those chips is quite differentiated and very important,” he said. “So we think a lot of that machine learning training, inference, will run on AWS.”
And while he said Amazon will be one of the few companies investing billions of dollars in building major, industry-leading LLMs, Jassy also focused on Amazon’s ability to offer options to companies that want to leverage a base model in AWS and then have the option to tailor it to their own ownership details, needs and customer experience. Companies want to do that in a way where they don’t leak their unique IP into the broader generalized model, he explained.
“That’s what Bedrock is, what we just announced a week ago,” he said. Bedrock is a managed base model service where people can run base models from Amazon or leading LLM providers like AI21 and Anthropic and Stability AI.
“They can run those models, take the baseline, modify them for their own purposes, and then run them with the same security and privacy and all the features they use for the rest of their applications in AWS,” he said. “That’s very attractive to customers.”
VentureBeat’s mission is to become a digital city plaza where tech decision makers can learn about transformative business technology and execute transactions. Discover our Briefings.