Advanced Computing in the Age of AI | Thursday, February 22, 2024

AWS Expands Amazon Bedrock with Additional Foundation Models, New Model Provider, and Advanced Capabilities 

NEW YORK, July 27, 2023 -- Amazon Web Services, Inc. (AWS) at AWS Summit New York announced the expansion of its fully managed foundation model (FM) service Amazon Bedrock to include the addition of Cohere as an FM provider and the latest FMs from Anthropic and Stability AI, as well as a new capability for creating fully managed agents in just a few clicks—a game-changing feature for builders, with no expertise required.

Credit: Shutterstock

Customers can use Amazon Bedrock to build and scale generative AI applications with a selection of industry-leading FMs by accessing a simple application programming interface (API), all in a secure environment and without managing any infrastructure. The expansion of Amazon Bedrock to include a new model provider, leading FMs, and the ability to easily create managed agents offers customers the broadest and most comprehensive toolset to leverage generative AI for any use case.

Many customers are already using Amazon Bedrock to transform their businesses with generative AI, including Coda, Lonely Planet, Ryanair, Showpad, and Travelers. Today’s announcement reinforces how Amazon Bedrock gives customers greater choice and flexibility to find the right FM for the job, while also providing a new capability that helps customers build generative AI applications that can take actions in the real world and deliver personalized answers to users based on their company’s data.

“Generative AI has the potential to transform every application, business, and industry. Advancements across data processing, compute, and machine learning are expediting the shift from experimentation to deployment for AWS customers of all sizes,” said Swami Sivasubramanian, vice president of Database, Analytics, and Machine Learning at AWS. “Through services like Amazon Bedrock and collaborations with industry leaders, we are democratizing access to generative AI, so wherever customers are on their machine learning journey, they can use AWS to reimagine experiences and bring new products to life.”

With the massive proliferation of data, the availability of highly scalable compute capacity, and the advancement of machine learning (ML) technologies, generative AI is now capable of transforming every industry, and customers want to understand how they can quickly take advantage of this new technology.

However, selecting the right model, securely customizing it using sensitive intellectual property or company data, and integrating it into an application is a complex process requiring significant time and highly specialized expertise. Amazon Bedrock is already simplifying this for customers, making it easier to build and scale generative AI-based applications, with access to a broad selection of best-in-class FMs from companies including AI21 Labs, Anthropic, Cohere, Stability AI, and Amazon, via a simple API. As a fully managed service, Amazon Bedrock allows customers to find and test different models, get started quickly, securely customize a model to fit their needs, and then integrate and deploy it to production.

New Model Provider Cohere Offers Even More FM Selection

AWS offers the broadest and deepest set of ML services and supporting cloud infrastructure to put ML and AI capabilities in the hands of builders of all levels of experience. Amazon Bedrock will add Cohere as the newest FM provider to give customers an even greater choice of best-in-class FMs to build their generative AI applications. Cohere is a leading developer of enterprise AI platforms and state-of-the-art FMs, and its FMs help to unlock more intuitive ways to generate, search, and summarize information. Command, Cohere’s flagship text generation model, is trained to follow user commands and be useful instantly in practical business applications such as summarization, copywriting, dialogue, extraction, and question answering. Cohere’s text understanding model, Embed, can be used for search, clustering, or classification tasks across 100+ languages, allowing organizations to easily search by meaning or categorize text.

Saurabh Baji, senior vice president of engineering at Cohere, said of the launch on Amazon Bedrock, “We are excited to expand our collaboration with AWS and offer our FMs on Amazon Bedrock. Our top-performing models are trained for business applications such as summarization, copywriting, dialog, extraction, and question answering. This collaboration will provide our joint customers with the greatest possible flexibility in ways to access our cutting-edge technology. Bringing Amazon Bedrock and Cohere together offers customers intelligent systems so that developers and enterprises alike can innovate faster and more easily with AI.”

Claude 2 from Anthropic and Stable Diffusion XL 1.0 from Stability AI Become Available on Amazon Bedrock

With Amazon Bedrock, customers can drive rapid innovation as new versions of FMs become available. Amazon Bedrock offers a range of text and image FMs from leading providers that span applications such as analyzing text for sentiment, classifying images, and forecasting trends.

  • Anthropic Claude 2 on Amazon Bedrock: Anthropic, an AI safety and research company that builds reliable, interpretable, and steerable AI systems, has brought Claude 2, the latest version of their language model, to Amazon Bedrock. Claude 2 can take up to 100,000 tokens in each conversational task prompt, meaning it can work over hundreds of pages of text, or even an entire book.Claude 2 can also write longer documents—like memos and stories on the order of a few thousand tokens—compared to its prior version, giving Amazon Bedrock customers even greater ways to develop generative AI applications.
  • Stability AI Stable Diffusion XL 1.0 on Amazon Bedrock: Stability AI is a community-driven, open AI company, offering FMs that create text, images, audio, video, code, and more from simple text instructions and will release the latest iteration of its text-to-image suite of models, Stable Diffusion XL 1.0 (SDXL 1.0), on Amazon Bedrock. SDXL 1.0 produces improved image and composition detail over its predecessor with its ability to generate more realistic creations for films, television, music, and instructional videos. SDXL 1.0 is also being released on Amazon SageMaker JumpStart, an ML hub that provides access to algorithms, models, and solutions.

Daniela Amodei, president and co-founder of Anthropic, said of today’s launch, "We're excited that our latest model Claude 2, which has made significant gains in performance and safety, is now accessible on Amazon Bedrock. This collaboration opens up new opportunities to deploy Claude to more enterprises using Amazon Bedrock's secure infrastructure and familiar AWS tooling. We are happy to work with AWS because they share our commitment to creating safer, more reliable AI. Together, we can help our customers unlock the power of generative AI and accomplish their goals."

Emad Mostaque, founder and CEO of Stability AI, said of their news, “Stability AI thrives on innovation, and our foundation model Stable Diffusion on Amazon Bedrock reflects the most pioneering developments across generative AI. Offering the latest release of our foundation model SDXL 1.0, with improved image and composition detail over its predecessor, gives Amazon Bedrock customers access to cutting-edge resources and advances our goal to activate humanity’s potential with AI.”

Agents for Amazon Bedrock

With a robust understanding of natural language and the ability to be further customized to domain-specific tasks, FMs are incredibly powerful for a wide range of use cases, such as summarization and question answering. However, FMs on their own are limited because they cannot complete complex tasks that require them to interact with external systems and don’t have up-to-date knowledge sources, like inventory data for an ecommerce company.

To make these kinds of capabilities possible, developers must go through a multistep process to accomplish even simple tasks, like booking a flight or returning a purchased item. This process includes breaking the task into multiple steps, by providing specific definitions and instructions, configuring access to current and relevant knowledge sources, and writing code so the generative AI application can take action according to the plan through a series of API calls. Developers must then provision and manage the requisite infrastructure, along with setting up policies for data security and privacy, which is tedious and time consuming. The complexity of this process for a single task means many companies are not able to unlock the full potential of generative AI.

Agents for Amazon Bedrock is a new, fully managed capability that makes it easier for developers to create generative-AI based applications that can complete complex tasks for a wide range of use cases and deliver up-to-date answers based on proprietary knowledge sources. With just a few clicks, agents for Amazon Bedrock automatically break down tasks and create an orchestration plan—without any manual coding. The agent securely connects to company data through a simple API, automatically converting data into a machine-readable format, and augmenting the request with relevant information to generate the most accurate response. Agents can then automatically call APIs to fulfill a user’s request. For example, an insurance agency might want to develop a generative AI application to help employees automate tasks like processing insurance claims or managing pending paperwork. As a fully managed capability, agents for Amazon Bedrock remove the undifferentiated lifting of managing system integration and infrastructure provisioning, allowing developers to leverage generative AI to its full extent throughout their organization.

To learn more about Amazon Bedrock, visit aws.amazon.com/bedrock.

About Amazon Web Services

Since 2006, Amazon Web Services has been the world’s most comprehensive and broadly adopted cloud. AWS has been continually expanding its services to support virtually any workload, and it now has more than 240 fully featured services for compute, storage, databases, networking, analytics, machine learning and artificial intelligence (AI), Internet of Things (IoT), mobile, security, hybrid, virtual and augmented reality (VR and AR), media, and application development, deployment, and management from 99 Availability Zones within 31 geographic regions.


Source: AWS

EnterpriseAI