Amazon Web Services has announced that its large-scale simulation system is now available for government customers. AWS Worldwide Public Sector Vice President Max Peterson sits down with Yahoo Finance’s Dan Howley to learn more about how AWS leverages generative AI.
video transcript
Diane King Hall: A big announcement from Amazon Web Services this morning. The company’s large-scale simulation system is now available to government customers following compliance requirements introduced last year. AWS SimSpace Weaver allows you to test scenarios at scale and use the results to plan and train real-world responses.
Various levels of government can now conduct tests with sensitive data. Big news for disaster and emergency scenario planning. Our own Dan Howley spoke to Max Peterson, Vice President of Global Public Sector at AWS about the system. Please try to hear.
Max Peterson: Everyone is looking forward to attending the World Public Sector DC Summit here. Part of that is due to generative AI. Amazon makes various announcements. Customers are so used to using Amazon’s machine learning and generative AI that they may be unaware of where they shop personalized, ask Alexa, checkout through Amazon Go, and more. yeah.
That’s why we’re here today at the summit to discuss the latest AWS services that help customers build and deploy generative AI solutions. Bedrock provides a simple and easy way to deploy generative AI solutions using pre-trained models. AWS cloud infrastructure is the most cost-effective way to train and run this kind of model using Inferentia and Trainium.
And finally, we have customers who are not just using it, but actually building foundational models at scale on AWS. So, I am pleased to tell you about the Technology Innovation Institute, the technology and research arm of the Government of Abu Dhabi, which trained a Falcon 40B large scale language model on AWS SageMaker.
And they have overcome performance, training and cost barriers that many others have failed to achieve. And we want to acknowledge and celebrate the fact that they were recognized.
Hugging Face is currently rated as the #1 LLM in the world with a large language model index.
Dan Howley: In other words, talk about aspects of generative AI and training and deployment. As you know, it’s obviously very costly. We have discussed this with OpenAI Google and Microsoft. So what does this mean for companies other than their main competitors? So what does it do for companies when we are talking about training and/or deploying generative AI for these large scale language models?
Max Peterson: Many customers simply don’t have the time, resources, or inclination to train their own large scale language and underlying models. And we believe Amazon Bedrock will be the way more customers can actually use enterprise and government-class AI and generative AI applications.
Bedrock makes it easy for customers to use pre-trained models such as Amazon’s Titan One and Titan Two models, or models from Stability AI. And these pre-trained models have known data provenance. They have known uses.
These are ready for use on AWS within the customer’s private virtual cloud, using the customer’s sensitive and often personal information, to train and enhance these already good models to achieve their intended purpose. can be used for Enables better interactive discussions with citizens about government services.
Or, in the case of a helper or co-pilot, to help that person, whether a law enforcement officer or a defense attorney, to make better and faster informed decisions by retrieving vast amounts of information. can be quickly summarized. Analyst.
Dan Howley: One of the things you guys announced is called SlimSpace Weaver. It’s basically a way of providing security and compliance requirements to the government. it satisfies them. But it’s meant to be a kind of simulation and training platform. Can you explain what it is and the idea of a digital twin involving government and the public sector? And what does it mean for them?
Max Peterson: yes. One of the most dynamic areas is government, which recognizes the need for more comprehensive planning.
For example, using digital twins and SimSpace Weaver, governments can model thousands or millions of different variables and create highly complex simulations in 2D or 3D to simulate complex traffic patterns in cities around the world. You will be able to run flow etc. It also makes it easier to model what happens with construction changes and road closures. And how are you going to use the simulated digital twin to produce better results in the real world?
Many things like emergency planning can be done more efficiently with a digital twin, a simulated environment. Because this kind of large-scale disaster scenario is never created in real life. But you still need to be ready.
What’s interesting about SimSpace Weaver’s announcement at the Summit is that it’s now available on GovCloud in the US. GovCloud is the cloud used for security and compliance by national, state and local government customers and companies building solutions for these types. Customers. So AWS SimSpace Weaver is now available on his AWS GovCloud.
Dan Howley: As a final question, I think you have SimSpace Weaver and generative AI. We talk a lot about generative AI, but it’s also about business and creativity. How far will we go in the public sector? What benefits does it bring to the public sector?
Max Peterson: As I just talked about, governments, healthcare, and educational institutions are using generative AI to, for example, provide better services to their citizens, provide more immersive conversational environments, and deliver better outcomes to their citizens. There are many applications that we can offer. Students can quickly summarize vast amounts of information for clinicians and healthcare professionals, helping them quickly find the right resources to help improve research and medical outcomes, and shift focus away from patients. increase. It’s all about the information gathering and evaluation part where tools like machine learning AI and generative AI help customers.
Diane King Hall: have understood. That’s our own Dan Howley, speaking with Max Peterson, Vice President of Worldwide Public Sector at AWS, about systems.
