Amazon CEO Andy Jassy recently shared the company’s plans to compete in the AI race against Microsoft and Google. In a letter to shareholders, Jassy revealed that Amazon will be investing heavily in large language models (LLMs) and generative AI, the same technology that powers ChatGPT and other similar AI chatbots.
Amazon has been using machine learning for more than 20 years in a wide array of products and services, from recommending products on Amazon.com to training the Alexa voice assistant. However, the onset of generative AI “promises to significantly accelerate machine learning adoption,” Jassy wrote.
Amazon’s approach to the AI race is different from the more cutthroat stance of other tech companies, who have invested billions in their own projects and AI startups to gain a proprietary edge over competitors. While Amazon already has its own patented AI projects, such as CodeWhisperer, an AI-powered code generator, its take on the AI race is tamer than that of Google and Microsoft. Amazon’s generative AI strategy will first focus on corporate customers through the Amazon Web Services, its most profitable division. In addition to developing its own model, Amazon will serve as a platform for other AI companies to run their models on.
Jassy said that Amazon will be focusing on strongly investing in LLMs and generative AI, which is based on LLMs across a wide range of datasets. Amazon’s goal is to offer less costly machine learning chips so that “small and large companies can afford to train and run their LLMs in production”. Large language models are trained on vast troves of data in order to generate responses to user prompts. Jassy believes that LLMs and generative AI will be transformative and accelerate machine learning adoption.
Amazon’s generative AI foray will also see it offer new AWS instances optimized for running language models and other generative AI using custom silicon. Amazon is making its AI tool for coding, Code Whisperer, free for all individual users through AWS, something that may also help get users hooked on its cloud.