Five takeaways from Matt Wood's keynote at AWS Summit New York

The artificial intelligence industry is crowded with large vendors and startups, and Amazon Web Services Inc.'s AWS Summit, held in New York City earlier this month, gave Amazon a platform to stake its claim as an AI leader.

To that end, the company tapped Matt Wood (pictured), VP of AI Products at AWS, as its keynote speaker. Below are five takeaways from his keynote that we feel best demonstrate the company's leadership in AI and generative AI.

Point 1: Gen AI and machine learning are already big business for AWS

Wood said in his keynote that the company already has “hundreds of thousands of customers running their AI and machine learning workloads on AWS today.” He added that Bedrock, AWS' service for building generative applications, is one of the “fastest growing AWS services of the last decade.” He said, “Across AWS, machine learning and AI is a multi-billion dollar a year business.”

Wood provided some statistics to back up his words. “96% of AI machine learning unicorns run on AWS,” he said. “If you look at the Forbes AI 50, 90% of those organizations also run on AWS.” Companies using AWS for AI applications range from the New York Stock Exchange and insurance giant Sun Life Insurance Company to pharmaceutical innovator Pfizer and life sciences giant Bayer, Wood said.

Point 2: AWS is committed to continuous AI innovation

Wood told the audience that AWS is making 326 generational AI capabilities generally available since 2023. He added that the company has the “broadest range of AI capabilities ever” and is “generally delivering more than twice as many new machine learning and generative AI capabilities than all other cloud providers combined.”

“You'd be amazed at how many successful startups and large enterprises are in AWS' exabyte club,” he says. “We have a lot of customers with huge amounts of private data that they can analyze — market research, clinical trial results, medical information, and more. They can apply generative AI to make sense of their existing data and use it in new and exciting ways.”

According to Wood, the AWS gen AI stack has three layers. The first layer is the foundational building blocks for building, training, tuning, and adjusting the underlying models. The second layer, Amazon Bedrock, is a service that “makes it easy for anyone to build applications using generative AI,” Wood said. The third layer is services for applying AI to the real world, such as Amazon Q, a generative AI-powered natural language tool for building machine learning models.

Wood described Amazon Q as “the easy button for generative AI.” “You can take information from anywhere in your organization and start leveraging it. For example, you can get rid of all the 'garbage,' the 70% of undifferentiated heavy lifting that's associated with different tasks in your organization.”

Pointer 3: AWS is the home of choice for organizations at the forefront of AI innovation

National Australia Bank is using Amazon Q's code customization capabilities to help onboard new engineers and get developers up and running quickly. [applicant] “Our adoption rate has jumped to 50 percent,” he notes, “and more than 40 percent of developers we surveyed have seen an increase in productivity. As a result, in just a few months, the number of employees using Q for these development tasks has more than doubled, from 450 to more than 1,000.”

According to Will McQueen, head of global data assets for crop science at Bayer, the company has been developing solutions on AWS since 2014. “When we were evaluating where to bet, we needed a partner and technology platform that could help us drive innovation and market-leading solutions for the next decade,” he said. “We decided AWS was the best place to bet on as our next-generation data science platform, and in the last year we've made great progress toward the level we need to accelerate our ability to bring data science solutions to market.”

Point 4: AWS takes security seriously, and so do its customers

One of the keys to AWS' AI success is its holistic commitment to ensuring the security of customer data: “AWS customers across all industries can meet and exceed their security requirements for their generative AI and their data,” says Wood.

He added that for AWS, nothing takes priority over security: “We stop everything we do to focus on security, and we build security in from day one. You can't build a secure service by taking an insecure service and then bolting security on later. It has to be built into the foundation.”Security scanning is one of the tasks Amazon Q performs.

“Some customers have the notion that to be successful with generative AI, they have to make some kind of negative trade-off around data confidentiality, privacy, and security,” Wood said. “But that's not the case at AWS. We don't use the data that moves through our paid AI services to improve the underlying models. There is no human review of that information. And [customers] You control where your data is stored within your VPC and how all information flows within your network.”

Point 5: AWS is committed to sustainability

AWS has grown rapidly, but that growth has been achieved while making sustainability a top priority: In 2019, Amazon set a goal to power all of its global operations — from AWS data centers to corporate buildings, grocery stores to fulfillment centers — with 100% renewable energy by 2030, Wood said. The company “reached that goal seven years early,” Wood said.

At a time when power grids and related resources around the world are under enormous strain, AWS' sustainability efforts are a sign of the company's continued success.

The AI ​​era is still in its infancy, and it's too early to declare a single winner. But the portfolio AWS has built is already attracting major customers and enabling organizations of all types and sizes to innovate and grow faster with generational AI, and AWS claimed a leadership position at AWS Summit New York. But the race is not over yet.

Zeus Kerravala is principal analyst at ZK Research, a division of Kerravala Consulting. He wrote this article for SiliconANGLE.

Photo: SiliconANGLE