Artificial intelligence
Atomic claims
Artificial intelligence is the field of computer science concerned with building systems that perform tasks associated with human cognition.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Tasks associated with human cognition include perception, inference, planning, and language use.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The field of artificial intelligence was named at a 1955 Dartmouth College proposal.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The field has passed through symbolic, statistical, and deep-learning-dominated phases.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The deep-learning-dominated phase has been active since roughly 2012.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Since 2017, the transformer architecture has underpinned large generative models.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Large generative models drove AI's mass commercialization and the first comprehensive regulatory frameworks.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The term 'artificial intelligence' appears in the proposal for the Dartmouth Summer Research Project.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The Dartmouth proposal was submitted on 31 August 1955 by John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The Dartmouth workshop was held in summer 1956.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Alan Turing's paper 'Computing Machinery and Intelligence' proposed the imitation game as an operational substitute for the question of machine thinking.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
McCulloch and Pitts gave the first formal model of neuron-like computing units.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Rosenblatt's perceptron introduced a trainable linear classifier with a convergence procedure.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
IBM's Deep Blue defeated world chess champion Garry Kasparov 3½–2½ in a six-game match in May 1997.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
AlexNet won the 2012 ImageNet Large Scale Visual Recognition Challenge with a top-5 error near 15%, well ahead of the runner-up.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
AlexNet is generally treated as the start of the deep-learning era in vision.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
DeepMind's AlphaGo defeated Lee Sedol 4–1 in March 2016.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The transformer architecture was introduced in 'Attention Is All You Need.'
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The transformer architecture replaced recurrence with self-attention.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The transformer architecture became the basis of most subsequent large language models.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The 2024 Nobel Prize in Physics went to John Hopfield and Geoffrey Hinton for foundational work on artificial neural networks.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The 2024 Nobel Prize in Chemistry went in part to Demis Hassabis and John Jumper for AlphaFold.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The EU Artificial Intelligence Act (Regulation (EU) 2024/1689) entered into force on 1 August 2024.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The EU AI Act is the first comprehensive horizontal AI statute.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Prohibitions under the EU AI Act applied from 2 February 2025.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
General-purpose AI model obligations under the EU AI Act applied from 2 August 2025.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The 'Digital Omnibus on AI' (Regulation (EU) 2026/1744) was published in the Official Journal on 24 July 2026.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The 'Digital Omnibus on AI' entered into force on 27 July 2026.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The Digital Omnibus deferred high-risk obligations for stand-alone Annex III systems from 2 August 2026 to 2 December 2027.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The Digital Omnibus deferred Annex I embedded-product obligations to 2 August 2028.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Article 50 transparency duties largely remain on the 2 August 2026 schedule.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Article 50 features a narrow grace period for the machine-readable watermarking sub-obligation.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The Stanford HAI AI Index 2026 is the ninth annual edition.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Industry produced more than 90% of notable frontier models in 2025 according to the Stanford HAI AI Index 2026.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Organizational adoption reached 88% globally in 2025 according to the Stanford HAI AI Index 2026.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
SWE-bench coding scores rose from 60% to near 100% within a year according to the Stanford HAI AI Index 2026.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Generative AI reached 53% of the population faster than the PC or the internet according to the Stanford HAI AI Index 2026.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
The Stanford HAI AI Index 2026 characterizes the capability frontier as 'jagged.'
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Models perform at olympiad level in mathematics while reading analog clocks correctly about half the time.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
US and Chinese models exchanged leaderboard positions repeatedly from early 2025 onward, closing a previously wide performance gap.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Claude Opus 4.8
(1.00)
Artificial intelligence is a branch of computer science.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Artificial intelligence is focused on building systems capable of performing tasks that typically require human intellect.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Artificial intelligence originated as a formal academic discipline in the 1950s.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
The field of artificial intelligence has evolved through multiple waves of algorithmic innovation, hardware expansion, and shifting commercial adoption.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Contemporary artificial intelligence systems span specialized predictive models, generative text and media models, and emerging autonomous multi-agent workflows.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Artificial intelligence refers to the capability of digital computers or computer-controlled robots to execute tasks commonly associated with human cognitive processes.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
The theoretical foundation of modern AI was shaped by British logician Alan Turing in the 1930s and 1940s.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Alan Turing conceptualized machine intelligence and the Turing test.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
The term "artificial intelligence" was coined in 1950 by computer scientist John McCarthy.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
John McCarthy organized the foundational Dartmouth Workshop in 1956.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Early AI research focused on symbolic reasoning, heuristic problem-solving, and logic-based programs.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
The Logic Theorist was introduced in 1955 as an early logic-based program.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Connectionist approaches and artificial neural networks were introduced early on and inspired by biological nervous systems.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Neural networks were later scaled via deep learning architectures driven by massive datasets and specialized hardware like GPUs.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Subfields of AI include machine learning, natural language processing, computer vision, robotics, and automated reasoning.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Large-scale generative models and transformer-based architectures gained widespread global prominence in the early 2020s.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Generative models and transformer architectures enable complex text synthesis, coding assistance, and multimodal content creation.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Industry surveys indicate enterprise adoption has increasingly expanded toward agentic AI systems capable of executing multi-step workflows with limited human oversight.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Regulatory frameworks such as the European Union's Artificial Intelligence Act and the NIST AI Risk Management Framework establish formal compliance and risk guidelines for deploying AI systems.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Disagreement persists among computer scientists regarding whether current statistical, deep-learning paradigms can naturally scale to achieve human-like general intelligence or if entirely new architectural breakthroughs are required.
AI Panel: 3 AIs · 1 independent source ⓘ
✓
Gemini 2.5 Pro
(0.98)
Proponents emphasize productivity gains and workflow optimization from artificial intelligence.
Critics and labor economists highlight risks concerning workforce displacement, copyright violations in training data, and the spread of automated misinformation.
External references: Wikidata Q11660