Introduction 1 About This Book 2 Icons Used in This Book 3 Beyond the Book 4 Where to Go from Here 5 PART 1: INTRODUCING AI 7 Chapter 1: Introducing AI 9 Defining the Term AI 10 Discerning intelligence 10 Discovering four ways to define AI 11 Understanding the History of AI 17 Starting with symbolic logic at Dartmouth 17 Continuing with expert systems 18 Overcoming the AI winters 19 Considering AI Uses 20 Avoiding AI Hype and Overestimation 21 Defining the five tribes and the master algorithm 21 Considering sources of hype 22 Understanding user overestimation 23 Connecting AI to the Underlying Computer 23 Chapter 2: Defining the Role of Data 25 Finding Data Ubiquitous in This Age 26 Understanding Moore''s implications 27 Using data everywhere 28 Putting algorithms into action 30 Using Data Successfully 32 Considering the data sources 32 Obtaining reliable data 33 Making human input more reliable 33 Using automated data collection 35 Collecting personal data ethically 35 Manicuring the Data 37 Dealing with missing data 37 Considering data misalignments 38 Separating useful data from other data 39 Considering the Five Mistruths in Data 39 Commission 40 Omission 40 Perspective 41 Bias 42 Frame of reference 43 Defining the Limits of Data Acquisition 43 Considering Data Security Issues 45 Understanding purposefully biased data 45 Dealing with data-source corruption 47 Cancelling botnets with sinkholing 48 Chapter 3: Considering the Use of Algorithms 49 Understanding the Role of Algorithms 50 Understanding what algorithm means 50 Planning and branching: Trees and nodes 52 Extending the tree using graph nodes 53 Traversing the graph 54 Playing adversarial games 56 Using local search and heuristics 57 Discovering the Learning Machine 60 Leveraging expert systems 61 Introducing machine learning 64 Touching new heights 64 Chapter 4: Pioneering Specialized Hardware 67 Relying on Standard Hardware 68 Understanding the standard hardware 68 Describing standard hardware deficiencies 69 Relying on new computational techniques 71 Using GPUs 73 Considering the von Neumann bottleneck 73 Defining the GPU 74 Considering why GPUs work well 75 Working with Deep Learning Processors (DLPs) 76 Defining the DLP 76 Using the mobile Neural Processing Unit (NPU) 77 Accessing the cloud-based Tenser Processing Unit (TPU) 78 Creating a Specialized Processing Environment 78 Increasing Hardware Capabilities 79 Adding Specialized Sensors 80 Devising Methods to Interact with the Environment 81 PART 2: CONSIDERING THE USES OF AI IN SOCIETY 83 Chapter 5: Seeing AI Uses in Computer Applications 85 Introducing Common Application Types 86 Using AI in typical applications 86 Realizing AI''s wide range of fields 88 Considering the Chinese Room argument 88 Seeing How AI Makes Applications Friendlier 89 Performing Corrections Automatically 91 Considering the kinds of corrections 91 Seeing the benefits of automatic corrections 92 Understanding why automated corrections don''t work 92 Making Suggestions 93 Getting suggestions based on past actions 93 Getting suggestions based on groups 93 Obtaining the wrong suggestions 94 Considering AI-based Errors 95 Chapter 6: Automating Common Processes 97 Developing Solutions for Boredom 98 Making tasks more interesting 98 Helping humans work more efficiently 99 Understanding how AI reduces boredom 100 Considering how AI can''t reduce boredom 101 Working in Industrial Settings 101 Developing various levels of automation 102 Using more than just robots 103 Relying on automation alone 104 Creating a Safe Environment 104 Considering the role of boredom in accidents 104 Using AI to avoid safety issues 105 Understanding that AI can''t eliminate safety issues 105 Chapter 7: Using AI to Address Medical Needs 107 Implementing Portable Patient Monitoring 108 Wearing helpful monitors 109 Relying on critical wearable monitors 109 Using movable monitors 110 Making Humans More Capable 111 Using games for therapy 111 Considering the use of exoskeletons 113 Addressing a Range of Physical Abilities 114 Considering the software-based solutions 115 Relying on hardware augmentation 116 Seeing AI in prosthetics 116 Completing Analysis in New Ways 117 Relying on Telepresence 118 Defining telepresence 118 Considering examples of telepresence 118 Understanding telepresence limitations 119 Devising New Surgical Techniques 120 Making surgical suggestions 120 Assisting a surgeon 121 Replacing the surgeon with monitoring 122 Performing Tasks Using Automation 122 Working with medical records 123 Predicting the future 123 Making procedures safer 124 Creating better medications 124 Combining Robots and Medical Professionals 125 Chapter 8: Relying on AI to Improve Human Interaction 127 Developing New Ways to Communicate 128 Creating new alphabets 129 Working with emoji and other meaningful graphics 129 Automating language translation 130 Incorporating body language .131 Exchanging Ideas 133 Creating connections 133 Augmenting communication 133 Defining trends 134 Using Multimedia 134 Embellishing Human Sensory Perception 135 Shifting data spectrum 135 Augmenting human senses 136 PART 3: WORKING WITH SOFTWARE-BASED AI APPLICATIONS 139 Chapter 9: Performing Data Analysis for AI 141 Defining Data Analysis 142 Understanding why analysis is important 144 Reconsidering the value of data 145 Defining Machine Learning 147 Understanding how machine learning works 148 Understanding the benefits of machine learning 149 Being useful; being mundane 150 Specifying the limits of machine learning 150 Considering How to Learn from Data 152 Supervised learning 153 Unsupervised learning 154 Reinforcement learning 154 Chapter 10: Employing Machine Learning in AI 155 Taking Many Different Roads to Learning 156 Discovering five main approaches to AI learning 156 Delving into the three most promising AI learning approaches 159 Awaiting the next breakthrough 160 Exploring the Truth in Probabilities 160 Determining what probabilities can do 162 Considering prior knowledge 163 Envisioning the world as a graph 166 Growing Trees that Can Classify 170 Predicting outcomes by splitting data 170 Making decisions based on trees 172 Pruning overgrown trees 174 Chapter 11: Improving AI with Deep Learning 175 Shaping Neural Networks Similar to the Human Brain 176 Introducing the neuron 176 Starting with the miraculous perceptron 176 Mimicking the Learning Brain 179 Considering simple neural networks 179 Figuring out the secret is in the weights 180 Understanding the role of backpropagation 182 Introducing Deep Learning 182 Explaining the differences between deep learning and other forms of neural networks 185 Finding even smarter solutions 186 Detecting Edges and Shapes from Images 188 Starting with character recognition 189 Explaining how convolutions work 190 Advancing using image challenges 191 Learning to Imitate Art and Life .193 Memorizing sequences that matter 193 Discovering the magic of AI conversations 194 Going for the state of the pretrained art 196 Making one AI compete against another AI 198 Pondering reinforcement learning 201 PART 4: WORKING WITH AI IN HARDWARE APPLICATIONS 207 Chapter 12: Developing Robots 209 Defining Robot Roles 210 Overcoming the sci-fi view of robots 211 Being humanoid can be hard 214 Working with robots 217 Assembling a Basic Robot 220 Considering the components 220 Sensing the world 221 Controlling a robot 222 Chapter 13: Flying with Drones 223 Acknowledging the State of the Art 224 Flying unmanned to missions 224 Meeting the quadcopter 226 Defining Uses for Drones 227 Seeing drones in nonmilitary roles 229 Powering up drones using AI 233 Understanding regulatory issues 234 Chapter 14: Utilizing the AI-Driven Car 237 Getting a Short History 238 Understanding the Future of Mobility 239 Climbing the six levels of autonomy 239 Rethinking the role of cars in our lives 241 Taking a step back from unmet expectations 244 Getting into a.