Machines That Think
Description
Artificial intelligence is no longer confined to science fiction. It recommends what we watch, curates what we read, and even drafts what we write. Yet for all its influence, AI remains a mystery to many of us. How does it really work? And what happens when we place more and more of our lives in the hands of machines that "think"?
Physicist and AI researcher Inga Strümke pulls back the curtain on the most transformative technology of our time in this captivating introduction to artificial intelligence. Writing for curious minds from all backgrounds, she traces the history of AI from early chess-playing machines to modern neural networks and language models. Along the way, she explains the principles behind machine learning with a balance of clarity and depth. Strümke explores the promise of AI breakthroughs like AlphaFold, which cracked one of biology's greatest puzzles-and spotlights the risks of biased decision-making, misinformation, and autonomous systems that sometimes fail with deadly consequences. She raises urgent questions about ethics, responsibility, and regulation: Who is accountable when AI systems make mistakes? How do we prevent algorithms from reinforcing prejudice? And how do we ensure that AI serves humanity's highest aspirations rather than its narrowest interests?
Winner of the Brage Prize for Norwegian nonfiction, Machines That Think is an essential guide to artificial intelligence-and a call to shape the future before it shapes us.
More details
Person
Content
... Foreword ... 7
... Preface ... 11
... Part I: The Start of It All ... 15
1 ... A Carefully Selected History ... 17
... The World's Most Successful Hoax ... 17
... Algorithms ... 19
... Analog Computers ... 20
... Gödel, Turing, and von Neumann ... 21
... A Field of Study Is Forged ... 25
... Checkmate ... 26
... The Drosophila of Reasoning ... 31
... The Best Chess Player in the Universe ... 33
... The Age of Machines ... 35
2 ... The Attempt to Make Machines Intelligent ... 39
... Symbolic AI ... 39
... Expert Systems ... 40
... The ELIZA Effect ... 46
... Winter and Spring ... 49
... Machines That Learn ... 51
... Supervised Learning ... 53
... Flowers and Decision Trees ... 56
... Subsymbolic AI ... 59
... Neural Networks Through the Ages ... 63
... Learning Anything ... 67
3 ... The Hunt for Data ... 73
... The True Distribution ... 73
... Data for Machine Learning ... 75
... The Wrong Distribution ... 77
... Bias for Statisticians and Journalists ... 80
... Data Is Expensive ... 85
... Long, Problematic Tails ... 87
... Finding Your Own Data ... 89
... Part II: Artificial Intelligence Today ... 95
4 ... Seeing Is Believing ... 97
... A World of Matrices ... 97
... Machines That Almost Understand What They're Looking At ... 102
... The Simple Answer Is Often Wrong ... 107
... Machines That Understand More Than We Do ... 110
... Artificial Intuition and Chess ... 114
... Mental Health ... 117
... Privacy ... 120
... The Right to an Explanation ... 124
5 ... Is Anybody Home? ... 127
... Artificial Conversation Partners ... 127
... The Ghost in the Machine ... 130
... ChatGPT and Concerned Teachers ... 135
... Talk About Problems ... 140
... A Turing Test for Consciousness ... 142
... The Art of Creation ... 146
... From Noise to Reality ... 153
6 ... Our Artificially Intelligent Life ... 161
... The Machine Revolution ... 161
... The World of Ethics ... 166
... Someone Has to Decide ... 172
... The War of the Currents and the AI Act ... 176
... The Collective Action Problem of AI Development ... 181
... Privacy in Flux ... 184
7 ... Attacking Machine Learning Models ... 191
... Well-Behaved Machines with Good Intentions ... 191
... Optical Illusions for Machines ... 194
... Cutting Grass in Three Dimensions ... 198
... The Curse of Dimensionality ... 200
... The Back Way In ... 203
... Winter and Politics ... 207
... Autonomy and Control ... 213
... Part III: Artificial Intelligence Tomorrow ... 219
8 ... Superintelligence and Other Speculation ... 221
... Artificial Creativity ... 221
... Speech and Acoustics ... 224
... Data Use ... 226
... AI Ethics and Moral Agents ... 228
... Artificial General Intelligence and Superintelligence ... 231
... Machine Goals, Human Goals, and Mesa-Optimization ... 238
... A Brain Thinking About Itself ... 245
... Illustrations ... 255