Index

• Advertising, targeted/discriminatory, 100–101, 115–116
• Allegheny County Family Screening Tool (AFST), 141–144
• AlphaGo, xix
• AlphaZero, 72
• Amazon, 3, 47, 115, 153–154
• Artificial general intelligence (AGI), xviii–xix
• Automation bias, 79–91, 138
• Autonomous vehicles, 73–76, 86, 168, 171–173
• no-fault compensation schemes and, 75–76
• strict liability and, 74
• Autonomy, 64, 103, 107–126
• human vs. machine, 71–72
• intrinsic vs. instrumental value of, 112–113
• liberty, actual vs. potential, 111–112
• privacy and, 113, 116, 118

• Babbage, Charles, xi
• Analytical Engine, xi
• Bias, 43–60. See also Privacy
• in algorithms, 45–46, 47–48, 52–54, 55–56, 57–60
• in humans, 34–36, 43–45
• sampling/selection, xv

• Cambridge Analytica, 103
• Classification tasks, xiv–xv
• human performance in, xiv–xv
• classifiers, 11–12
• Clinical judgment, xiii–xiv
• COMPAS, 57–60, 82–83
• Computer science, relation to AI, 1
• Computers vs. humans. See Humans vs. computers
• Confusion matrices, 15
• Copernicus, Nicolaus, xviii
• COVID-19, xvi, xxii, 175–176

• Darwin, Charles, xviii
• Data protection. See Privacy
• Decision trees, 10–12
• random forests, 11
• Domain-specific AI, xviii–xix
• Dworkin, Gerald, 109
• Dynamic pricing, 55, 103

• Echo chambers, 115
• Employment, 149–157
• COVID-19 and, 176
• dignity of, 154–157
• gig economy, 153–154
• privacy and, 103, 153–154
• women and, 151–152
• Engels, Friedrich, xii
• Communist Manifesto, xii
• Evaluation methods, 15–16, 140–141
• Ex Machina (film), xix
• Explainability, 24
• double standards, 39–41
• Explainable AI (XAI), 37–39
• human reasons and, 30–33
• neural networks and, 28–29
• practical reason and, 30–33
• reasons, duty to provide, 26–28
• reasons for decisions and, 26
• right to explanation, 26–28, 106

• Facebook, 51, 97n, 100–101, 102n, 104, 116, 124
• Filtering algorithms, 115–116
• Freud, Sigmund, xviii

• General Data Protection Regulation (GDPR), 30, 33, 106, 125
• Gig economy, 153–154
• Gompertz, Benjamin, 7
• Gompertz curve, 7
• Google, 3, 39, 51, 54, 57, 115, 161–162

• Hobbes, Thomas, 111
• House of Lords Select Committee on AI (UK), 33–34, 163
• Human dignity, xvii–xix, 141
• Human factors, 79–91, 138
• Humans vs. computers, xviii–xix, 71–72, 87–88, 89
• rationality of machine learning, xx–xi, 140

• Internet of Things, 97

• Keynes, John Maynard, 150

• Lethal Autonomous Weapons Systems (LAWS), 79–81, 162–163
• Liability, 61–78. See also Responsibility
• strict, 65–66, 74
• strict vs. fault-based, 65–66
• Liberty. See Autonomy
• Locke, John, 111

• Machiavelli, Niccolò, 111
• Marx, Karl, xii
• Communist Manifesto, xii
• Microsoft, 51, 55

• Neural networks, 12–14
• error backpropagation, 14
• deep networks, 14
• feedforward networks, 12–13
• opacity of, 14, 28–29
• Nudging, 118–119, 121

• Pettit, Philip, 111
• PredPol, 49–50
• Privacy, 93–106, 113, 116, 118
• advertising, targeted/discriminatory, 100–101, 115–116
• access to data, 99, 106
• anonymization, 99
• by design, 105–106
• collected vs. inferred data, 98
• consent to collection of data, elusiveness of, 97–98
• correction of data, 99, 106
• data brokerage, 97
• data protection, as distinct from, 95
• definitional challenges, 93–95
• deletion of data, 100, 106
• derived data, 98, 106
• employment and, 103, 153–154
• inferred data, 98, 106
• informational, 95, 96–100
• observed vs. inferred data, 98
• purpose limitation, 96
• retention of data, limits to, 100
• varieties of, 93–96, 103
• Professional judgment, xiii–xiv

• Raz, Joseph, 109–110
• Regression, 8–10
• Regulation, 159–174
• Reinforcement learning, 19
• Responsibility, 23, 25, 61–78
• AI and, 70–73, 77
• causation and, 64, 68–69
• conditions of, 64
• foresight and, 64, 69–70
• moral vs. legal, 62–63, 65, 66–67
• technology’s effects on, 67–70
• Responsibility gap, 71
• vs. moral crumple zone, 73
• Russell, Bertrand, 155, 157

• Smith, Adam, 127
• On the Wealth of Nations, 127
• Statistics, 1–19
• bias, sampling/selection xv
• correlations, counterintuitive, xx–xi
• history of, 3, 4–6
• vs. human judgment, xiii–xiv, 140
• machine learning, relation to, 3, 4–6
• Supervised learning, 13–14
• error backpropagation, 14
• Surveillance capitalism, 118

• Toffler, Alvin, xii, xxii
• Future Shock, xii
• Transparency, 21–41
• various meanings of, 22–24
• Turing, Alan, xviii, 1

• Unsupervised learning, 18

• Verne, Jules, xii
• Twenty Thousand Leagues Under the Sea, xii
• von Neumann, John, xi, 1

• Wells, H. G., xi
• Wisdom of crowds, 48–49