There's little question that the advent of cloud computing was a game changer. The ability to store data in the cloud changed the technological landscape by providing secure access to mission-critical software and systems anytime, anywhere. It also catapulted Amazon (AMZN -1.27%) into the next stage of its growth, expanding beyond its e-commerce roots.
As a pioneer in the cloud computing space, Amazon has long been the leader in the industry it helped to pioneer, but it's also not the only game in town. Microsoft's Azure and Alphabet's Google Cloud have joined Amazon Web Services (AWS), forming the "big three" in cloud infrastructure services. While AWS is still the leader by a fair margin, its two rivals have been growing more quickly in recent years, threatening to unseat the leader.
Amazon recently revealed plans to maintain its dominance and reignite its cloud growth.
Image source: Getty Images.
Late to the AI party?
This year's market recovery has come in stark contrast to its predecessor, and the primary driver has been advances in artificial intelligence (AI).
Microsoft helped ignite the fervor around generative AI, announcing a $13 billion investment in OpenAI shortly after the debut of ChatGPT. Microsoft quickly integrated ChatGPT into its Bing search and enhanced a broad cross-section of its enterprise software with AI functionality.
Alphabet was quick to respond, debuting its own chatbot and laying out plans to integrate new AI-powered tools across its vast empire.
While its rivals were swift to adopt AI, the perception was that Amazon was caught flatfooted, scrambling to play catch-up. The company is trying to put that idea to rest.
A three-pronged plan
When Amazon released its second-quarter results, much of the ensuing conference call was spent discussing AI. CEO Andy Jassy said Amazon is "reimagining" generative AI since consumer applications are just "one layer of the opportunity."
Amazon views generative AI as three potential opportunities, each of which is substantial:
- The company is leveraging Nvidia H100 AI chips to help customers "train large models and develop generative AI applications." This is an optimal solution for companies with the know-how that merely need access to the computational horsepower to get it done.
- Over the past couple of years, Amazon has developed its own custom AI processors -- dubbed Trainium and Inferentia -- and believes that in the future, many of the large language models (LLMs) that underpin generative AI will be trained and run on these chips. Jassy also noted that most companies don't want to expend the "billions of dollars and multiple years to develop" their own LLMs from scratch. Furthermore, they want to protect their proprietary data from leaking to competitors. Amazon views this opportunity as offering LLMs-as-a-Service, leveraging its custom AI processors. The company has already developed its own LLM and partnered with leading AI companies to make these LLMs available to AWS customers.
- The third layer of this trifecta is conversational models, similar to ChatGPT, which are the applications that the LLMs enable. Amazon's Bedrock allows customers to create their own chatbots or "conversational agents" that can be customized to specific applications. One example would be a customer service chatbot that could handle simple requests before sending more complex issues to human operators.
Jassy summarized Amazon's position, saying, "The core of AI is data," and Amazon "has the broadest array of storage, database, analytics, and data management services for customers," but also more users and data stores than its rivals.
Early days
It's important to remember that while Amazon hasn't been at the forefront of the conversation about advances in generative AI, the company has long used AI algorithms to suggest products on its e-commerce website, plan shipping routes, track packages, run warehouses, maintain inventory levels, and choose items for cross-docking.
There's no question that Amazon will take the same careful and thoughtful approach to integrating generative AI tools into its cloud products and services. Amazon has a long track record of success, and there's no evidence to suggest that generative AI will be any different.
