Foreword
AI and machine learning are revolutionary technologies that can change the world, but they can only do that if there are developers using good APIs to take advantage of the advancements these technologies bring.
One such advancement is the ability to run machine learning models in the browser, empowering applications that act intelligently.
The rise of TensorFlow.js tells me that AI has arrived. It’s no longer exclusively in the realm of data scientists with supercomputers; it’s now accessible to the millions of developers who code in JavaScript daily. But there’s a gap. The tools and techniques for building models are still very much in the hands of those who know the mysteries of Python, NumPy, graphics processing units (GPUs), data science, feature modeling, supervised learning, tensors, and many more weird and wonderful terms that you probably aren’t familiar with!
What Gant has done with this book is to cut to the chase, teaching you the important stuff you need to know while keeping you firmly within the web developer role, using JavaScript and the browser. He’ll introduce you to the concepts of AI and machine learning with a clear focus on how they can be used in the platform you care about.
Often, I hear developers ask, when wanting to use machine learning, “Where can I find stuff that I can reuse? I don’t want to learn to be an ML engineer just to figure out if this stuff will work for me!”
Gant answers that question in this book. You’ll discover premade models that you can take lock, stock, and barrel from TensorFlow Hub. You will also learn how to stand on the shoulders of giants by taking selected portions of models built using millions of items of data and many thousands of hours of training, and see how you can transfer learn from them to your own model. Then, just drop it into your page and have JavaScript do the rest!
Developers ask, “How can I use machine learning on the platform I care about without extensive retraining?”
This book goes deeply into that—showing you how to bridge the gap between JavaScript and models that were trained using TensorFlow. From data conversion between primitives and tensors to parsing output probabilities into text, this book guides you through the steps to integrate tightly with your site.
Developers ask me, “I want to go beyond other people’s work and simple prototypes. Can I do that as a web developer?”
Again, yes. By the time you’ve finished this book, not only will you be familiar with using models, but Gant will give you all the details you need to create them yourself. You’ll learn how to train complex models such as convolutional neural networks to recognize the contents of images, and you’ll do it all in JavaScript.
A survey in October 2020 showed that there were 12.4 million JavaScript developers in the world. Other surveys showed that there are about 300,000 AI practitioners globally. With the technology of TensorFlow.js and the skills in this book, you, dear JavaScript developer, can be a part of making AI matter. And this book is a wonderful, wonderful place to start.
Enjoy the journey!