As one of the largest cloud computing providers in the world, Amazon's (AMZN -1.27%) Amazon Web Services (AWS) has been at the forefront of innovation since its inception. In recent years, the company has been doubling down on its efforts to integrate artificial intelligence (AI) into its services, and the results have been nothing short of remarkable.
With a renewed focus on AI, AWS is rejuvenating its cloud business and solidifying its position as a leader in the industry. Let's explore how AWS is leveraging AI to drive growth and create new opportunities.
The three layers of opportunity in generative AI
When most people hear the term generative AI (GenAI), they immediately think of ChatGPT, the app released by OpenAI in late November 2022 that stunned the world. Yet, ChatGPT is only one application on the broad surface of the GenAI opportunity, and many more companies will benefit from the adoption of GenAI than OpenAI and its partner Microsoft.
One huge beneficiary of the proliferation of this new technology will be AWS. On its second-quarter 2023 conference call, Amazon's management explained that the GenAI opportunity has three layers, and the company's AWS unit has heavily invested in each. The first layer is the computing power to teach and run an AI model. Most companies rely on Nvidia's graphics processing units (GPUs) for computing. The problem is Nvidia can barely keep up with the market's demand for them.
A few years back, Amazon started developing its own AI chips, Trainium and Inferentia, for training and running AI systems. These chips have already undergone a second version, and management thinks they offer an enticing price-performance option for those developing and running AI models. These chips could provide an attractive alternative to Nvidia's chips, although it's still unclear whether the company plans to sell them to others or only use them within its own data centers.
The middle layer of the opportunity is the AI model level. Think of the AI model as the engine for an application. For instance, ChatGPT is a consumer application that runs on top of an AI model called GPT-4. People may interface with ChatGPT, but the AI model GPT-4 creates responses to inquiries. These AI models take a massive amount of time and money to develop, discouraging many companies from developing their own.
Therefore, there is another business opportunity for AWS to offer AI models as a service, allowing customers to build their applications on top. The company has an AI model called Bedrock, which customers can customize with their proprietary data and have everything managed by AWS. In addition, Bedrock can access AI models from companies like Anthropic, Stability AI, AI21 Labs, Cohere, and Amazon's Titan model. Bedrock could be a huge opportunity.
The top layer is the application layer, and the company plans to take advantage of it by building valuable applications for its customers. One such application is Amazon CodeWhisperer, a coding assistant that uses AI to provide coding recommendations to help developers increase their productivity.
It has been successful so far, and Amazon is working on creating more AI applications to improve its customer experience. Just think of all the different applications it could build for personalized product recommendations, customer service, new product ideas, improvement to existing product designs, fraud detection, and content creation. The sky is the limit.
Enormous competition in all three layers
AWS faces significant competition on all three levels of the GenAI opportunity. On the AI chip level, in addition to chips made by traditional chipmakers like Nvidia and Advanced Micro Devices, it faces competition from AI chips made by Alphabet's Google AI Hardware team and Microsoft's Azure.
On the AI model level, there are many competitors to Amazon's Bedrock. A few of the most notable ones are OpenAI's GPT-4, DeepMind's Gato, Google PaLM (Pathways Language Model), and Microsoft's Megatron-Turing Natural Language.
GenAI is rapidly evolving, and many companies are continuously developing new models. Competition in this space will likely intensify in the future. Whichever company can produce the most accurate, creative, flexible, scalable, and affordable AI model will have a significant advantage in the market.
Lastly, AWS also has significant competition on the application level. For instance, Amazon CodeWhisperer faces substantial competition from GitHub Copilot, a collaboration between Microsoft-backed GitHub and OpenAI that uses an AI model to suggest code completions and solutions for various programming languages and frameworks. Whatever new AI application Amazon develops could compete with better-known or better-performing applications from other companies.
GenAI should help AWS results rebound
Over the last two years, high inflation and rising interest rates have increased uncertainty in the economy, and this has caused many customers to look for ways to cut costs. As a result, AWS revenue growth has declined over the last seven quarters as its customers sought to conserve cash.
However, management believes rising customer interest in GenAI should bring more business to AWS. With recent comments from Amazon's CEO suggesting there may be a stabilization in AWS' revenue growth rate, we could soon see an upswing in the cloud unit's growth, a prime opportunity for investors to consider investing in Amazon and potentially reap the benefits of a rebound in its most lucrative segment.
