Let's begin our journey by peering into the future and envision how we'll see ourselves interacting with computers. Unlike today's computers, where we are required to continuously type in our emails and passwords to access information, the computers of the future will easily be able to recognize us by our face, voice, or activity. Unlike today's computers, which require step-by-step instructions to perform an action, the computer of the future will anticipate our intent and provide a natural way for us to converse with it, similar to how we engage with other people, and then proceed to help us achieve our goal. Our computer will not only assist us but also be our friend, our doctor, and so on. It could deliver our groceries at the door and be our interface with an increasingly complex and information-rich physical world.

What is exciting about this vision is that it is no longer in the realm of science fiction but an emergent reality. One of the major drivers of this is the progress and adoption of machine learning (ML) techniques, a discipline that gives computers the perceptual power of humans, thus giving them the ability to see, hear, and make sense of the world—physical and digital.

But despite all the great progress over the last 3-4 years, most of the ideas and potential are locked away in research projects and papers rather than being in the hands of the user. So it's the aim of this book to help developers understand these concepts better. It will enable you to put them into practice so that we can arrive at this future—a future where computers augment us, rather than enslave us due to their inability to understand our world.

Because of the constraint of Core ML—it being only able to perform inference—this book differs vastly from other ML books, in the sense that the core focus is on the application of ML. Specifically we'll focus on computer vision applications rather than the details of ML. But in order to better enable you to take full advantage of ML, we will spend some time introducing the associated concepts with each example. 

And before jumping into the hands-on examples, let's start from the beginning and build an appreciation for what ML is and how it can be applied. In this chapter we will:

• Start by introducing ML. We'll learn how it differs from classical programming and why you might choose it.
• Look at some examples of how ML is being used today, along with the type of data and ML algorithm being used.
• Finally, present the typical workflow for ML projects.

Let's kick off by first discussing what ML is and why everyone is talking about it.