Cover of AI Threat

Book 5 · Presentation

AI Threat

Are the Dystopias Coming True? Evidence, Risk and Response for European Enterprises and Public Authorities

82 slidesPowerPoint · 4.2 MB

Download the deck (.pptx) Read the book online

All presentations / AI Threat

All slides

Select a slide to see it larger; use the arrow keys to move through the deck.

  1. Slide 1: AI Threat1
  2. Slide 2: The Question2
  3. Slide 3: Introduction: Are the Dystopias Coming True?3
  4. Slide 4: Introduction: Are the Dystopias Coming True?4
  5. Slide 5: A Century of Machine Dystopias: What Science Fiction Actually Warned About5
  6. Slide 6: A Century of Machine Dystopias: What Science Fiction Actually Warned About6
  7. Slide 7: The Voices of Alarm and Calm: Who Is Warning, About What, and Why Now7
  8. Slide 8: The Voices of Alarm and Calm: Who Is Warning, About What, and Why Now8
  9. Slide 9: Weighing the Evidence and Disagreeing Well9
  10. Slide 10: Weighing the Evidence and Disagreeing Well10
  11. Slide 11: What AI Can and Cannot Do in 202611
  12. Slide 12: What AI Can and Cannot Do in 202612
  13. Slide 13: The Human Mind13
  14. Slide 14: Are We Getting Dumber? Cognitive Offloading and Critical Thinking14
  15. Slide 15: Are We Getting Dumber? Cognitive Offloading and Critical Thinking15
  16. Slide 16: Learning Without Struggle: Education, Children and Skill Formation16
  17. Slide 17: Learning Without Struggle: Education, Children and Skill Formation17
  18. Slide 18: Persuasion, Sycophancy and Manipulation18
  19. Slide 19: Persuasion, Sycophancy and Manipulation19
  20. Slide 20: Artificial Companions, Loneliness and Mental Health20
  21. Slide 21: Artificial Companions, Loneliness and Mental Health21
  22. Slide 22: Society and Truth22
  23. Slide 23: The End of Truth? Synthetic Content and the Information Ecosystem23
  24. Slide 24: The End of Truth? Synthetic Content and the Information Ecosystem24
  25. Slide 25: The Automated State: When Public Authorities Got It Wrong25
  26. Slide 26: The Automated State: When Public Authorities Got It Wrong26
  27. Slide 27: Discrimination at Scale: Bias in Automated Judgement27
  28. Slide 28: Discrimination at Scale: Bias in Automated Judgement28
  29. Slide 29: Surveillance, Profiling and the Erosion of Privacy29
  30. Slide 30: Surveillance, Profiling and the Erosion of Privacy30
  31. Slide 31: Work, Skills and the Ironies of Automation31
  32. Slide 32: Work, Skills and the Ironies of Automation32
  33. Slide 33: Inequality and the Concentration of Power33
  34. Slide 34: Inequality and the Concentration of Power34
  35. Slide 35: Culture, Creativity and the Homogenisation of Thought35
  36. Slide 36: Culture, Creativity and the Homogenisation of Thought36
  37. Slide 37: The Physical Cost: Energy, Water and Materials37
  38. Slide 38: The Physical Cost: Energy, Water and Materials38
  39. Slide 39: Humans Wielding Machines39
  40. Slide 40: Cyber Offence at Machine Speed40
  41. Slide 41: Cyber Offence at Machine Speed41
  42. Slide 42: Biological and Chemical Uplift: The Hardest Question to Test42
  43. Slide 43: Biological and Chemical Uplift: The Hardest Question to Test43
  44. Slide 44: Fraud, Impersonation and Influence Operations44
  45. Slide 45: Fraud, Impersonation and Influence Operations45
  46. Slide 46: Autonomous Weapons and the AI Arms Race46
  47. Slide 47: Autonomous Weapons and the AI Arms Race47
  48. Slide 48: Machines Beyond Control48
  49. Slide 49: Specification Gaming and Reward Hacking: Getting Exactly What Was Asked For49
  50. Slide 50: Specification Gaming and Reward Hacking: Getting Exactly What Was Asked For50
  51. Slide 51: Deception, Scheming and Evaluation Awareness: Can We Trust What We Test?51
  52. Slide 52: Deception, Scheming and Evaluation Awareness: Can We Trust What We Test?52
  53. Slide 53: Self-Preservation and Shutdown Resistance: HAL 9000 Under Laboratory Conditions53
  54. Slide 54: A stop path outside the agent's reach, with the signals that trigger it54
  55. Slide 55: Agents in the Wild: The Incidents of 2025 and 202655
  56. Slide 56: Agents in the Wild: The Incidents of 2025 and 202656
  57. Slide 57: Gradual Disempowerment and Correlated Failure57
  58. Slide 58: Gradual Disempowerment and Correlated Failure58
  59. Slide 59: Superintelligence, Loss of Control and the End of the World59
  60. Slide 60: Superintelligence, Loss of Control and the End of the World60
  61. Slide 61: The Response61
  62. Slide 62: What Europe and the World Have Decided62
  63. Slide 63: What Europe and the World Have Decided63
  64. Slide 64: Frontier Safety Frameworks: What Developers Promise and What It Is Worth64
  65. Slide 65: Frontier Safety Frameworks: What Developers Promise and What It Is Worth65
  66. Slide 66: Organisational Defences: Containment, Evidence and Decision Rights66
  67. Slide 67: Organisational Defences: Containment, Evidence and Decision Rights67
  68. Slide 68: Staying Capable: What Individuals, Families and Schools Can Do68
  69. Slide 69: Staying Capable: What Individuals, Families and Schools Can Do69
  70. Slide 70: The AI Threat Register: Governing Scenarios as Code70
  71. Slide 71: The AI Threat Register: Governing Scenarios as Code71
  72. Slide 72: The Verdict72
  73. Slide 73: Nordvik Region and Hansa Mutual: Threat Conclusions73
  74. Slide 74: Nordvik Region and Hansa Mutual: Threat Conclusions74
  75. Slide 75: What Going Well Looks Like75
  76. Slide 76: What Going Well Looks Like76
  77. Slide 77: Conclusions: Is There Anything in It?77
  78. Slide 78: Conclusions: Is There Anything in It?78
  79. Slide 79: Appendices79
  80. Slide 80: Code Examples: The AI Threat Register and Evidence Artefacts80
  81. Slide 81: Templates and Tools81
  82. Slide 82: AI Threat82