Preface

Since the first programmable digital computer called ENIAC came to our world in 1946, computers have been so widely used and have become an integral part of our lives. It’s impossible to imagine a world without computers.

Entering the 21st century, the so-called ABC Triangle stands out in the computer world, and its three vertices represent today’s most advanced computer technologies – A for Artificial intelligence, B for Big data, and C for Cloud computing, as you can see from the following figure. These technologies are reshaping our world and changing our lives every day.

It is very interesting to look at these advanced computer technologies from a historical point of view, to understand what they are and how they have developed with each other:

• Artificial intelligence (AI) is a technology that enables a machine (computer) to simulate human behavior. Machine learning (ML) is a subset of AI that lets a machine automatically learn from past data and predict based on the data. AI was introduced to the world around 1956, shortly after the invention of ENIAC, but in recent years, AI has gained momentum because of the accumulation of big data and the development of cloud computing.
• Big data refers to the steady exponentially increasing data generated and stored in the past years. In 2018, the total amount of data created and consumed was about 33 zettabytes (1 ZB =8,000,000,000,000,000,000,000 bits) worldwide. This number grew to 59 ZB in 2020 and is predicted to reach a mind-boggling 175 ZB by 2025. To process these big data sets, huge amounts of computing power are needed. It is inconceivable to process these huge data sets on commodity computers, not to mention the time it takes for a company to deploy traditional data centers to place these computers. Big data processing calls for new ways to provision computing powers.
• Cloud computing came into our world in 2006, about half a century after the idea of AI. Cloud computing provides computing powers featuring elastic, self-provisioning, and on-demand services. In a traditional computing model, infrastructure is conceived as hardware. Hardware solutions are physical – they require space, staff, planning, physical security, and capital expenditure – thus they have a long hardware procurement cycle that involves acquiring, provisioning, and maintaining. The cloud computing model made the infrastructure as software – choose the cloud computing services that best match your business needs, provision and terminate those resources on-demand, scale the resources up and down elastically in an automated fashion based on demand, deploy the infrastructure/resources as immutable codes that are managed with version control, and pay for what you use. With the cloud computing model, computing resources are treated as temporary and disposable: they can be used much more quickly, easily, and cost-effectively. The cloud computing model made AI computing feasible.

AI, big data, and cloud computing work with each other and thrive – more data results in more AI/ML applications, more applications demand more cloud computing power, and more applications will generate more data.

Famous for its innovation-led mindsets and industry-trend-led products, Google is a leader in the ABC Triangle technologies. As an ML pioneer, Google developed AlphaGo in 2017, the first computer program that defeated a professional human Go world champion. AlphaGo was trained on thousands of human amateur and professional games to learn how to play Go. AlphaZero skips this step and learns to play against itself – it quickly surpassed the human level of play and defeated AlphaGo by 100 games to 0. In addition to the legendary AlphaGo and AlphaZero, Google has developed numerous ML models and applications in many areas, including vision, voice, and language processing. In the cloud computing arena, Google is one of the biggest cloud computing service providers in the world. Google Cloud Platform (GCP) provides the best cloud services on earth, especially in the areas of big data and ML. Many companies are keen to use Google Cloud and leverage the GCP ML services for their business use cases. And this is the purpose of our book. We aim to learn about and master the best of the best – ML in Google Cloud.