Table of Contents
Introduction: 'I don't expect your children to die of old age' 1
Part 1 Introductions
1 Introducing the Rationalists 13
2 The cosmic endowment 25
Part 2 The Paperclip Apocalypse
3 Introducing AI 33
4 A history of AI 40
5 When will it happen? 45
6 Existential risk 50
7 The cryptographic rocket probe, and why you have to get it right first time 56
8 Paperclips and Mickey Mouse 60
9 You can be intelligent, and still want to do stupid things 66
10 If you want to achieve your goals, not dying is a good start 72
11 If I stop caring about chess, that won't help me win any chess games, now will it? 81
12 The brief window of being human-level 85
13 Getting better all the time 91
14 'FOOOOOM' 95
15 But can't we just keep it in a box? 100
16 Dreamed of in your philosophy 105
17 'It's like 100 per cent confident this is an ostrich' 109
Part 3 The Ways of Bayes
18 What is rationality? 117
19 Bayes' theorem and optimisation 121
20 Utilitarianism: shut up and multiply 131
Part 4 Biases
21 What is a 'bias'? 143
22 The availability heuristic 146
23 The conjunction fallacy 149
24 The planning fallacy 151
25 Scope insensitivity 154
26 Motivated scepticism, motivated stopping and motivated continuation 157
27 A few others, and the most important one 160
Part 5 Raising the Sanity Waterline
28 Thinking probabilistically 165
29 Making beliefs pay rent 173
30 Noticing confusion 176
31 The importance of saying 'Oops' 178
Part 6 Decline and Diaspora
32 The semi-death of Less Wrong 183
33 The IRL community 186
Part 7 Dark Sides
34 Are they a cult? 197
35 You can't psychoanalyse your way to the truth 211
36 Feminism 214
37 The Neoreactionaries 220
Part 8 Doing Good Better
38 The Effective Altruists 227
39 EA and AI 235
Part 9 The Base Rate of the Apocalypse
40 What are they doing to stop the AI apocalypse? 251
41 The internal double crux 260
42 Life, the universe and everything 268
Acknowledgements 273
Notes 275