ARTFEED — Contemporary Art Intelligence

AGI Debate Intensifies as Chinese AI Firms Pursue Human-Level Intelligence

ai-technology · 2026-09-12

Chinese AI developers are publicly declaring ambitions to build artificial general intelligence as the country races to close its gap with the United States, even as Beijing keeps a cautious posture toward the concept. AGI lacks a single agreed definition, but broadly describes systems that equal or exceed human-level intelligence. US physicist Mark Gubrud coined the term in 1997, and computer scientists Ben Goertzel, author of the book Artificial General Intelligence, and Shane Legg, co-founder of Google DeepMind, popularised it during the 2000s. Recent advances in large language models have renewed efforts to pin down what AGI actually means. A widely cited paper published in December, whose authors include Dan Hendrycks of the Center for AI Safety and Dawn Song of the University of California, Berkeley, defined AGI as AI matching the versatility and proficiency of a well-educated adult across 10 cognitive components, among them reading and reasoning. The debate has been further fuelled by OpenAI Astra.

Key facts

  • Chinese AI companies are declaring ambitions to pursue AGI as China races to close the gap with the US
  • Beijing appears to maintain a cautious stance towards the concept of AGI
  • There is no single definition for AGI, though it broadly refers to systems matching or surpassing human-level intelligence
  • US physicist Mark Gubrud first coined the term in 1997
  • The term was popularised in the 2000s by computer scientists Ben Goertzel and Shane Legg
  • Ben Goertzel authored the book Artificial General Intelligence
  • Shane Legg is co-founder of Google DeepMind
  • A widely cited December paper defined AGI as AI matching the versatility and proficiency of a well-educated adult across 10 cognitive components, such as reading and reasoning
  • The December paper's researchers include Dan Hendrycks from the Center for AI Safety and Dawn Song from the University of California, Berkeley
  • Recent advancements in large language models have triggered new efforts to define AGI
  • OpenAI Astra has sparked debate

Entities

Artists

  • Mark Gubrud
  • Ben Goertzel
  • Shane Legg
  • Dan Hendrycks
  • Dawn Song

Institutions

  • OpenAI
  • Google DeepMind
  • Center for AI Safety
  • University of California, Berkeley
  • South China Morning Post

Locations

  • China
  • United States
  • Berkeley
  • California

Sources