Major cloud players are moving swiftly to take control of the generative AI narrative.

This dynamic could be seen in November when Microsoft Corp. acted rapidly to assist OpenAI in getting back on track following days of boardroom turmoil. It was also reinforced last week at AWS re:Invent in Las Vegas as the cloud giant unveiled the “Generative AI Stack,” a platform designed to give customers applications for leveraging AI, tools to build large language models and the infrastructure to drive model training and inference.

How Amazon Web Services Inc. intends to capitalize on the wave of interest in generative AI was one of several key themes to surface during the “Supercloud 5: The Battle for AI Supremacy” event, an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. Over the course of four days, theCUBE was on the ground in Las Vegas, broadcasting interviews with AWS executives, industry thought leaders and cloud practitioners to capture how the world’s market leader in public cloud unveiled its gen AI story.

“The undertone was, we are not going to let you or anyone else say that we’re behind in AI — we’re going to flex it,” said theCUBE industry analyst John Furrier, during analysis of the keynote address by AWS Chief Executive Office Adam Selipsky. “I think there’s a game-changing shift happening where the cloud players are the new global infrastructure, like hardware.”

Furrier was joined in the segment by fellow analyst Dave Vellante.

Here is theCUBE’s analyst discussion during Day 2 of Supercloud 5:

Here are three key insights you might have missed during the event:

1. Bedrock and Amazon Q emerge as the central focus of AWS’ gene AI strategy.

Amazon Bedrock, the cloud giant’s fully managed service for generative AI, became generally available at the end of September. Bedrock was launched with immediate access to a number of high-performing AI models, and AWS unveiled new additions to the service during re:Invent. The message was that AWS’ approach would be to provide choice and innovation as part of its gen AI offerings.

“Then you have customers who say, ‘No, I just want to use models that are provided by Amazon or others’ — then we have Bedrock,” said Bratin Saha, vice president and general manager of AI and machine learning at AWS, in an interview with theCUBE. “That provides you the most choice that there is out there of state-of-the-art models along with other capabilities we talked about today, like [retrieval augmented generation] knowledge bases and so on.”

Here’s theCUBE’s complete video interview with Bratin Saha:

Last week, AWS announced Amazon Q, an intelligent digital assistant built to guide developers and IT professionals in key programming tasks. Q can provide assistance in building applications on AWS, troubleshoot networking issues, and provide recommendations for instance types. The company envisions its use for a broad array of enterprise needs.

“The big trend is that you are getting Q to be this intelligent virtual assistant for everything,” said Tom Soderstrom, enterprise strategist at AWS, in conversation with theCUBE. “It’s at the code, how to design it, how to code it, build test cases and deploy it.”

Here’s theCUBE’s complete video interview with Tom Soderstrom:

The launch of Amazon Q also pulled the curtain back on AWS’ strategic approach in the AI marketplace and demonstrated the changes it has made to build consistency in its portfolio of products.

“I do think having a single generative AI client there that works across all your products [provides consistency],” according to Zeus Kerravala (pictured, center), founder and principal analyst at ZK Research, in an interview with theCUBE. “One of the things Amazon’s doing better now is they’re having better integration across all their building blocks, and I think Q’s a good example of that.”

Here’s theCUBE’s complete video interview with Zeus Kerravala, who was joined by Shelly Kramer (left), managing director and principal analyst with theCUBE Research:

2. Generative AI is already making an impact across a wide range of industries.

Although gen AI is not new, the technology did not explode into the public consciousness until OpenAI’s ChatGPT became generally available approximately one year ago. It didn’t take long before industries began to embrace the technology. One of these is the telecommunications industry, where gen AI has become a catalyst for telcos in data management.

“Generative AI has been in the news for a while now, but telcos are starting to figure out how to use it, and for that they needed access to some key capabilities, like protecting the data,” said Chivas Nambiar, general manager for the Telco Business Unit at AWS, during an appearance on theCUBE. “It’s not just a generative AI story. It’s AI and data. I think they’re moving faster than you think.”

Here’s theCUBE’s complete video interview with Chivas Nambiar:

AI is also making inroads in the healthcare industry, where organizations are spending less time analyzing and managing data and more time in research and patient care. This is allowing healthcare organizations to personalize the experience.

“We’re excited to work with healthcare customers who are saying, ‘I now have those disparate data points coming together, and I’m able to use it to know exactly what I need to do for clinical efficiency,’” said Angela Shippy, senior physician executive at AWS, in an interview with theCUBE. “That clinical efficiency is important. It’s eliminating waste. It’s giving those patients that personalized experience that they want, and it’s reinforcing the workforce so that resilience is there.”

Here’s theCUBE’s complete video interview with Angela Shippy:

Along with telcos and the healthcare industry, the impact of AI is also being felt in the world of sports. AWS has a high-profile relationship with the National Football League through provision of its Next Gen Stats capability, and it is also collaborating with the National Hockey League to provide advanced data analytics.

“We’re really analyzing all of the trends and areas of opportunity for innovation in sport,” said Julie Souza, head of sports and global professional services at AWS, in her discussion with theCUBE. “Data … is a huge part of that. We’re collecting data in copious volumes that we never collected in sports before.”

Here’s theCUBE’s complete video interview with Julie Souza:

3. Vector and RAG could be the next big things in AI.

Two of the technologies that could make AI even more powerful in the near future are vector search and retrieval-augmented generation, or RAG.

Last week, MongoDB Inc. announced the integration of its Atlas Vector Search with Amazon Bedrock. The move by MongoDB was designed to address a need for developers to enable more sophisticated analytics as part of building intelligent applications.

“As [AI] moves left away from the machine learning engineers to developers, really embedding AI to make apps smarter, it’s having a real profound effect,” said Dev Ittycheria, MongoDB’s president and CEO, in an exclusive interview with theCUBE. “The reason they gravitated to MongoDB is that the usability of all these capabilities is very well-crafted so that the developers can just get the technology out of the way, that they can just do their work. There’s other [tools] out there, but it’s very clunky for a developer to connect their vector data to their metadata, to their core data, to be able to orchestrate all that, then figure out how to do the embeddings and all that. It just becomes a very convoluted process.”

Here’s theCUBE’s complete video interview with Dev Ittycheria:

What companies such as MongoDB and AWS are realizing is that customers want choice and the ability to put vectors where they want them to be.

“As customers are looking at generative AI, they don’t want to be taking their data out of their data lake and shipping it to some external model,” said Andy Warfield, vice president and distinguished engineer at Amazon.com Inc., in an appearance on theCUBE. “They really want to be bringing the model to the data.”

Here’s theCUBE’s complete video interview with Andy Warfield:

This access to data becomes especially important for fine-tuning the accuracy of large language models. This will lead organizations to seek additional context from outside data sources to enable more accurate query responses, according to Jerry Chen, general partner at Greylock Partners, in a conversation on theCUBE. This will require RAG tools that will lead to the development of new software stacks.

“Think about all the software you need to train, fine-tune, run the models and build these applications,” Chen said. “[It’s] how to build these data pipelines of your enterprise data with these new models inside a public cloud or private cloud. I think you’re going to see new software stacks be created.”

Here’s theCUBE’s complete video interview with Jerry Chen:

To watch more of theCUBE’s Supercloud 5 event, here’s our complete event video playlist:

Photo: SiliconANGLE