Artificial Intelligence (AI) may seem ubiquitous these days, but it is, and it has underpinned our daily lives for some time. Whether you ask Alexa to play a song, stride (with a sandwich) from a store with Amazon Just Walk Out, or press Play a movie recommendation from Amazon Prime, it’s powered by AI. It means that More specifically, we are working with machine learning (ML) models.
You’ve probably seen a lot of focus and attention on generative AI in recent months. Generative AI is a subset of machine learning that uses machine learning at massive scale. ML-modelThis includes large scale language models (LLM) and multimodal models (text, image, video, audio, etc.). Applications like ChatGPT and Stable Diffusion are capturing everyone’s attention and imagination, and with good reason. Generative AI is poised to have a major impact across industries such as healthcare and life sciences, media and entertainment, education, and financial services.
At Amazon, we believe AI and ML are among the most transformative technologies of our time, capable of tackling some of humanity’s toughest problems. Therefore, over the past 25 years, Amazon has invested heavily in AI and ML development, infusing these capabilities into every business unit.
Today, our ML models are working on behalf of hundreds of millions of Amazon customers around the world to remove friction from the supply chain, personalize digital experiences, and make goods and services more accessible and affordable. By doing so, we provide tangible value. Over 100,000 customers use dozens of Amazon Web Services (AWS) AI and ML services Delivered in the cloud.
Amazon still has a lot of features and is moving fast. Below are four of his innovations just announced to support generative AI applications.
Amazon Bedrock: Easily Build Generative AI Applications
Amazon Bedrock is a new service for building and scaling generative AI applications, applications that can generate text, images, speech, and synthetic data in response to prompts. Amazon Bedrock gives you easy access to Foundation Models (FM), the ultra-large scale ML models that generative AI relies on, from top AI startup model providers such as AI21, Anthropic, Stability AI, and exclusive access to the Titan family becomes possible. Foundation model developed by AWS. One model doesn’t do it all. Amazon Bedrock opens up a range of FMs from major providers, giving AWS customers the flexibility and choice to use the model that best fits their particular needs.
result: Amazon Bedrock is the easiest way to build and scale generative AI applications with FM.
Amazon EC2 Inf2 Instances Powered by AWS Inferentia2 Chips Generally Available: Reduce the Cost of Running Generative AI Workloads
Running super-large ML models in production requires massive compute. AWS Inferentia chips offer the highest energy efficiency and lowest cost for running demanding generative AI inference workloads (such as answering text queries and generating images) on AWS at scale.
result: Reduced costs and energy consumption make generative AI more accessible to a wider range of customers.
New Trn1n Instances Powered by AWS Trainium Chips: Custom Silicon for Faster Model Training
Generative AI models must be trained to provide the correct answer, image, insight, or other focus on which the model works, and performing these trainings requires extensive computing resources and is expensive. increase. The new Trn1n instances (the server resources where the compute takes place) run on custom Trainium chips from AWS, providing key large-scale network capabilities for training these models quickly and cost-effectively. To do.
result: Developers will be able to train models faster and at lower cost, ultimately leading to more services powered by generative AI models.
Free access to Amazon CodeWhisperer for individual developers: real-time coding assistance
Imagine software developers making coding faster and easier with an AI-powered coding companion. Amazon CodeWhisperer does just that. It uses generative AI under the hood to provide real-time code suggestions based on user comments and previous code. Individual developers can access Amazon CodeWhisperer for free with no usage limits (a professional paid tier with features such as additional enterprise security and management features is also available).
result: What are you waiting for? It’s free! get the coding.
These offerings are just the beginning. One of Amazon’s core principles is its commitment to long-term thinking. We believe they are in the early stages of a technological revolution that will last for decades to come. That’s why we continue to shape the future of this technology, based on our customers’ needs, towards the most promising, beneficial, and trusted AI applications.
Learn more about Generative AI with Swami Sivasubramanian, VP of Databases, Analytics, and Machine Learning at AWS and Dr. Werner Vogels, Amazon CTO.
Introducing Generative AI | Amazon Web Services
To dive deeper and get all the technical details in this exciting news, be sure to read on. Sivasbramanian’s Blog.
