AI Alignment
Problem of ensuring AI systems pursue intended goals and values. Orthogonality thesis: intelligence and goals are independent. Instrumental convergence: AIs may pursue harmful subgoals. Control problem: how to maintain control over superior intelligence. Value learning: AIs should learn human values. Corrigibility: AI should accept shutdown and correction. Existential risk if misaligned superintelligence emerges. Technical and philosophical challenge. Urgency debated.
Your Reaction
Similar Philosophical Concept
Algorithmic Bias
Problem of AI systems perpetuating or amplifying social biases. Training data reflects historical discrimination. Algorithms …
The Singularity
Hypothetical point where AI recursively self-improves, leading to superintelligence. Vinge, Kurzweil: technological change becomes uncontrollable …
Technological Unemployment
Economic concern that automation eliminates jobs faster than creating new ones. Frey-Osborne: 47% of US …
Effective Accelerationism
Techno-optimist philosophy advocating accelerating technological development. Abbreviation: e/acc. Believes technology solves problems technology creates. Opposes …
Longtermism
Ethical framework prioritizing long-term future and potential people. MacAskill, Ord: most value lies in future; …
Object-Oriented Ontology
Speculative realist position that objects exist independently of human thought. Harman: objects withdraw from all …
Consciousness
The subjective quality of experience—what it's like to be something. Phenomenal consciousness is qualitative experience …
Transhumanism
Intellectual movement advocating use of technology to enhance human capabilities beyond current limitations. Transhumanists support …
Explore Related Domains
Discover connections across different types of entities
Eliezer Yudkowsky
American AI researcher and writer who founded rationalist community and works on AI alignment. Yudkowsky's …
Anthropic
AI safety and research company founded by former OpenAI leadership including Dario and Daniela Amodei. …
DeepMind
British AI research lab acquired by Google. Founded by Demis Hassabis, Shane Legg, and Mustafa …
OpenAI
AI research laboratory that created GPT models and ChatGPT. Founded by Sam Altman, Elon Musk, …
Ex Machina
Garland's AI chamber piece follows a programmer testing an android's consciousness. Through the Turing test, …
Factorio
Wube's factory automation game has player crash on alien planet and build industrial complex. Optimizing …
The Talos Principle
Croteam's puzzle game places AI in philosophical trials by mysterious administrator. Portal-like puzzles interweave with …
Demis Hassabis
British neuroscientist and AI researcher who co-founded DeepMind, advancing AI through games and protein folding. …