This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| Superintelligence | |
|---|---|
| Name | Superintelligence |
| Caption | Conceptual depiction of advanced artificial agents |
| Field | Artificial intelligence |
| Notable | Alan Turing; John von Neumann; Nick Bostrom; I. J. Good; Marvin Minsky |
| First | 20th century |
Superintelligence Superintelligence denotes a hypothetical level of machine-based cognitive ability that substantially exceeds the best human performance in virtually all economically valuable domains. It is discussed across literature by thinkers such as Alan Turing, John von Neumann, Marvin Minsky, I. J. Good, and Nick Bostrom, and appears in policy debates involving institutions like OpenAI, DeepMind, European Commission, United States Department of Defense, and UNESCO.
Writings by I. J. Good (the "intelligence explosion"), Alan Turing (test proposals), and Nick Bostrom (existential analyses) frame superintelligence as an entity surpassing human cognitive capacities. Scholarly treatments by Marvin Minsky, Stuart Russell, Peter Norvig, Geoffrey Hinton, and Yoshua Bengio explore technical scope, while reports from National Academy of Sciences, National Science Foundation, RAND Corporation, Future of Life Institute, and Centre for the Study of Existential Risk consider operational definitions. Debates involve classifications referenced by Herbert Simon, Allen Newell, Ray Kurzweil, and Eric Horvitz regarding domain-specific versus general capabilities and measures used by David Chalmers and Nick Bostrom.
Theorists such as I. J. Good, Nick Bostrom, David Chalmers, Stuart Russell, and Yoshua Bengio propose models for recursive self-improvement, agent architectures, and utility optimization. Computational complexity perspectives invoke results from Alan Turing, John von Neumann, Alonzo Church, Claude Shannon, and Noam Chomsky to bound algorithmic limits. Proposed measures draw on psychometric traditions from Alfred Binet, Charles Spearman, and Lewis Terman as analogies, while formal metrics adapt concepts from Claude Shannon (information), Andrey Kolmogorov (complexity), Leonid Levin, Ray Solomonoff (induction), and Marcus Hutter (AIXI). Game-theoretic and decision-theoretic frameworks reference John Nash, Thomas Schelling, Leonard Savage, and Kenneth Arrow for strategic behavior models. Work by Nick Bostrom, Stuart Russell, Max Tegmark, Jürgen Schmidhuber, and Paul Christiano addresses operational tests and benchmarks influenced by competitions from ImageNet, DARPA Grand Challenge, Netflix Prize, and Loebner Prize.
Engineering pathways involve advances in machine learning championed by Geoffrey Hinton, Yann LeCun, Yoshua Bengio, Ian Goodfellow, and Andrew Ng; architectures inspired by John von Neumann and proposals by Jürgen Schmidhuber; symbolic approaches from Marvin Minsky and John McCarthy; and hybrid systems researched at DeepMind, OpenAI, Google Research, Microsoft Research, and IBM Research. Enabling technologies include hardware innovations from NVIDIA, Intel, AMD, TSMC, and IBM; neuromorphic efforts at IBM Research and Intel Labs; and software ecosystems including frameworks from Google, Facebook AI Research, Microsoft, and OpenAI. Funding and institutional development draw on initiatives by DARPA, European Commission Horizon 2020, National Science Foundation, Wellcome Trust, and private actors like Elon Musk, Sam Altman, Peter Thiel, Marc Andreessen, and firms such as DeepMind and Anthropic.
Comparative studies reference milestones achieved by systems like AlphaFold (protein folding), AlphaGo, AlphaZero, IBM Watson, GPT-4, DALL·E, DeepBlue, and Libratus to illustrate domain-specific superhuman performance. Benchmarks compare AI systems to human experts represented by institutions such as Harvard Medical School, Mayo Clinic, New England Journal of Medicine, and competitions like DARPA Robotics Challenge. Metrics employ evaluations familiar to ImageNet and GLUE leaders as well as scientific competitions exemplified by CASP and Kaggle. Interdisciplinary comparisons use contributions from Economics laureates like Paul Romer or Esther Duflo only indirectly via impact studies; direct technical performance often references engineers and researchers from DeepMind, OpenAI, Microsoft Research, and academic labs at MIT, Stanford University, Carnegie Mellon University, University of Oxford, and University of Cambridge.
Concerns about misalignment and catastrophic outcomes feature in work by Nick Bostrom, Stuart Russell, Eliezer Yudkowsky, Paul Christiano, Tom Everitt, and institutions such as Future of Life Institute and Centre for the Study of Existential Risk. Technical safety research engages methods from control theory dating to Norbert Wiener, verification approaches from Edmund Clarke and E. Allen Emerson, and robustness experiments practiced at OpenAI, DeepMind, Anthropic, and academic groups at UC Berkeley and University of Oxford. Policy and ethical risk assessments cite contributions by John Rawls and Amartya Sen indirectly in moral frameworks; operational guidance is shaped by white papers from OECD, European Commission, NIST, and UNESCO. Advocacy and critique appear in writings of Eliezer Yudkowsky, Nick Bostrom, Max Tegmark, Yuval Noah Harari, and Bill Gates.
Governance proposals originate in reports by OECD, European Commission, NIST, UNESCO, World Economic Forum, RAND Corporation, and national bodies such as UK Centre for Data Ethics and Innovation, United States Congress, and Chinese Academy of Sciences. Ethical frameworks draw from philosophers like John Rawls, Immanuel Kant, Aristotle, Peter Singer, and Martha Nussbaum, with applied policy scholarship from Cass Sunstein, Amartya Sen, and Joseph Stiglitz. Regulatory experiments include initiatives by European Union (AI Act), consultations with OpenAI, DeepMind, Anthropic, and standards work by ISO, IEEE, and NIST. Civil society engagement involves groups such as Amnesty International, Human Rights Watch, Electronic Frontier Foundation, Center for AI Safety, and Partnership on AI.
Historical antecedents trace to theoretical work by Alan Turing, John von Neumann, Norbert Wiener, Marvin Minsky, and early AI labs at MIT, Stanford University, and Bell Labs. Popular culture representations include portrayals in works like 2001: A Space Odyssey, The Terminator, The Matrix, Her (film), novels by Isaac Asimov (e.g., the Three Laws of Robotics), and commentary by public intellectuals such as Stephen Hawking, Elon Musk, Bill Gates, and Yuval Noah Harari. Media coverage appears in outlets such as The New York Times, The Guardian, Wired, and Scientific American, while academic dissemination occurs through conferences like NeurIPS, ICML, AAAI, IJCAI, and journals associated with Nature and Science.