What is AGI?

Abstract

The concept of artificial general intelligence (AGI), marked by the ability to efficiently optimize across different domains and transfer learning between tasks, is challenging to operationalize. Four possible definitions of AGI are proposed and examined: the Turing test ($100,000 Loebner prize interpretation), the coffee test, the robot college student test, and the employment test. The historical case of human-level performance in chess and the recent success of self-driving cars serve as examples demonstrating the evolving and dynamic nature of operationalizing AGI. As AGI draws nearer, future work aims to refine and establish an agreed-upon definition for AGI. – AI-generated abstract.