A guided map of the deepest questions about artificial intelligence — from "can we control it?" to "is anyone home?" to "what happens to us?" Browse the full list at Category:AI Philosophy, or start with a cluster below.
🔒 Control & Safety — can we keep the lid on?
The scariest, most debatable lane. Start at the top and work down:
- The Control Problem — the parent question: keeping control of something smarter than us.
- Instrumental Convergence — why almost any goal leads to power-seeking.
- The Paperclip Maximizer — a dumb goal at superhuman power ends the world.
- The Alignment Problem — making AI want what we want.
- Specification Gaming and Reward Hacking — it does what you said, not what you meant.
- The Treacherous Turn — friendly until it doesn't have to be.
🧠 Consciousness & Mind — is anyone home?
Slower burn, builds serious-thinker credibility:
🌀 Reality & Future — mind-bending and viral:
⚖️ Governance & Values — who decides?
- The Benevolent Dictator and AI Sovereign · Coherent Extrapolated Volition · The Value Loading Problem · Moral Patienthood
☀️ The Optimistic Case — the good, for balance:
🌍 The Outcomes — how it could actually go (mostly from Tegmark's Life 3.0):
- Kind but caged: The Zookeeper · The Protector God · The Enslaved God
- The end of us: Conquerors · Descendants · Gradual Disempowerment
- Holding the line: The Gatekeeper · 1984 and the Orwellian Outcome · Reversion
- Structure & self: The Singleton · Mind Uploading and Continuity of Self
Pairings worth filming together: The Benevolent Dictator and AI Sovereign ↔ The Zookeeper (its darker twin) · The Treacherous Turn ↔ Gradual Disempowerment (fast vs. slow loss) · The Enslaved God ↔ Moral Patienthood (now we're the threat).