AI Ethics, Coeckelbergh

Review and notes on AI Ethics, by Mark Coeckelbergh.

1. Mirror, Mirror, on the Wall

  • How pervasive is the impact of AI?
  • AI ethics concerns the impact of AI technologies on society and the economy.

2. Superintelligence, Monsters, and the AI Apocalypse

  • Superintelligence - machines that surpass human intelligence.
    • Recursive self-improvement
    • Whole brain emulation
  • Most narrations of superintelligence are premised on some sort of technological existentialism.
  • Transhumanism - we need to enhance the human being (deus ex human, so to speak) to compete/keep up with AI
  • Most researchers interpret ideas in spiritual, religious ways - even if they attempt to reject it.
  • Frankenstein’s new monster - the transfiguration of the creation.
  • The control problem - what happens when science eludes our control?
  • Romanticism attempts to liberate us from the darker sides of modernity
  • The creation competes with the creator, like a child who devours the patient.
  • Frankenstein complex - a deep-seeded fear of the child, of a child which we cannot help creating but is so profoundly different from us we fear it will murder us.
  • Transcendence - God is ontologically and epistemilogically above the mateial world.
  • The transhumanist quest for immortality begins very erly in human history.

3. All About the Human

  • One perspective suggests that experiential knowledge, from the phenomenological being-in-the-world, cannot be formalized.
  • Continential tradition emphasises human exceptionalism.
  • Analytic tradition emphasizes that the human mind really is a machine of sorts.
  • Even if in theory superintelligence is possible, this does not mean it is likely to emerge soon.
  • Enlightenment thinking - optimism on the potential of science to shape human ability and experience.
  • Romanticist thinking - casts science as a pervasive anti-good
  • Humanists vs the transhumanists - who/what is human? What is the materiality/transcendental quality of the human?
  • Donna Haraway and the cybernetic transhumanist project.
  • What is the narrative of progress?
  • Standing upon posthumanism, AI can liberate itself from human intelligence towards nonhuman modes of intelligence and creativity.
  • AI has a strong materiality to it.

4. Just Machines?

  • Should AI be given rights? Responsibilities?
  • Moral agency - can it make decisions and be held responsible? Patiency - does it have protected rights? Is it entitled to certain qualities?
  • Is AI autonomous? Is it intelligent?
  • Do machines have the capacity for mental states, emotions, or free will?
  • Can we give AI principles to follow?
  • We have to give AI partial morality but not complete morality.
  • Trolley dilemmas - a caricature of ethics.
  • Is it morally acceptable to kill AI? Can we torture AI? Is it moral to dismember its mind?
  • Humans need to know little of AI to empathize with it - the Lacanian fantasy.
  • Can robots be others? Or are they relegated to the Big Other?

7. Privacy and the Other Usual Suspects