ラスベガス万歳！：AWS REINVENT 2023の概要

OK, I’m stretching the subtitle a bit, but last week I had the opportunity to attend AWS ReInvent. Many thanks to AWS for the invitation to meet, learn, and share insights with AWS execs, banks, and vendors from around the world, over the course of an intense four days.

I’ve seen estimates that state between 50,000 and 70,000 people descended on Las Vegas for ReInvent 2023, which was hosted across multiple hotel conference centers. The Formula 1 street circuit snakes around the hotels and the iconic “Sphere” and, with the F1 race held only the weekend before, the overhead lighting and grandstands were still being deconstructed. It’s amazing I was able to find anyone (never mind everyone) as planned!

As expected, the event provided a proverbial firehose of information about new products and capabilities across the entire technology stack. Not surprisingly AI and GenAI capabilities were dominant themes. GenAI grabbed the media headlines in 2023 and Celent has covered this extensively in our Celent GenAI.

With so much attention being paid to Meta, Google, Microsoft, OpenAI, and NVIDIA, in some ways the focus on AI by AWS has been overlooked in the mainstream and business media. AWS ReInvent was an opportunity to make up for some lost media perception and highlight the AWS capabilities for building AI-based applications and the use of GenAI. The emphasis was clearly on industrial AI at scale, built from the ground up. Indeed, you could be forgiven for thinking that this was an AI conference! However, that only underscores the speed and scale of AI development in the cloud - and how the cloud value prop is shifting. Some of the key AI announcements include:

• Two significant partnerships were highlighted in the keynote:
  • NVIDIA’s Jensen Huang joined the keynote to discuss how NVIDIA and AWS are working together to build supercomputers for GenAI
  • Anthropic selecting AWS as its primary cloud provider, and a recent AWS investment of up to $4bn for collaboration in AI development
• There is no “one size fits all” approach to using models, you need the right model (tool) for the job. There is a choice of foundation models from AWS and their partners, including Anthropic.
• Guardrails for Amazon Bedrock that provide developers with a framework within which to build GenAI applications in line with a firm’s responsible AI policies.
• “Amazon Q” – a GenAI-powered assistant that is apparently inspired by the quartermaster, “Q” from James Bond. Currently in preview, this GenAI copilot is intended to be tailored (and trained) to meet the needs of a customer’s business.

At the Financial Services Industry Keynote, Scott Mullins, GM AWS Worldwide Financial Services was joined by speakers from Standard and Poors, JP Morgan, and Nasdaq - who all talked about their cloud, data, and AI journeys. What stood out to me – and is a theme that I’ve written about before – is that their journeys began with data strategy (and spending some time on data management hygiene) before moving to the cloud. Once in place, financial institutions can better start to realize the benefits of AI.

Banks must be aware of (and evaluate) the capabilities that mega-tech platforms like AWS and Microsoft are building to support developers across the financial services community. Furthermore, AI capabilities are now being embedded in these platforms rather than offered merely as a bolt-on. These are important considerations for banks and fintechs as they update data, applications, and AI strategies.