Welcome to Carbon Aware Computing for Gen AI Developers. Built in partnership with Google Cloud. I'm delighted their instructor for this course is Nikita Namjoshi, who's a developer advocate at Google Cloud and also a Google fellow on the Permafrost Discovery Gateway, which is an initiative using AI to track Arctic permafrost four. She's working on sustainability for over five years has a lot to share about best practices for reducing CO2 emissions of AI workloads. Thanks, Andrew. I'm excited to work with you and your team on this. Training, fine tuning, and even serving your models, especially generative models can be compute intensive and energy-intensive, but it doesn't have to be carbon intensive. If you know how to customize where and when you run your ML jobs in the cloud. Specifically, you can choose low CO2 energy sources to power your training jobs by running them in data centers to powered by low-carbon energy sources such as wind, solar, hydro, or nuclear. In this short course, you learn how to query the breakdown of energy sources for different regional electricity grids across the world using the Electricity Maps API. And also you'll learn how to train an ML model in the cloud at a data center that is powered by a high amount of carbon free energy. Current estimates are that cloud computing accounts for around 2.5 to 3.7% of our global greenhouse gas emissions. Further, AI workloads are continuing to grow, which is a sign of how our field is booming, but also makes it increasingly important that we do our part to mitigate emissions. In this course, you learned about how carbon is quantified and reported, according to the Greenhouse Gas Protocol corporate standard, and what that means for you as a developer. Nikita will go through these concepts using the Electricity Maps API, as well as Google Cloud's Vertex AI SDK and carbon footprint too. But the concepts you learn apply to multiple clouds, so even if you're just using your local machine. Thanks, Andrew. I think it's a lot easier to understand our environmental impact when we do something like use a single-use plastic fork or water bottle, or when we fly on a plane or maybe put gas in a car, but it's a little harder to grasp the impact from something like computing and machine learning, where we're just writing code. Learning about carbon-aware development has been really eye opening for me. With some simple steps that you'll learn in this course, you can make a significant difference on the carbon impact of your machine learning workloads. Also, be sure to stick around for the end of the course where you will explore a dashboard that helps you pick an optimal Google Cloud region based on how you weigh factors like carbon footprint, cost, and latency. The course will also wrap up with an overview of some exciting research in the field of sustainable cloud computing. Many people have worked to create this course. I'd like to thank from Google Cloud, John Abel, Cynthia Wu, Khulan Davaajav, and Jeff Sternberg. From DeepLearning.AI Eddy Shyu also contributed to this course. In the first lesson, you'll learn about energy and the electric grid as they relate to computing infrastructure and carbon emissions. You'll also explore the carbon intensity of regions all around the world with an interactive map. That sounds great. Let's go on to the next video and get started.