Design of the Artificial: lessons from the biological roots of general intelligence
This addresses the fundamental challenge of designing AGI for researchers and developers, but it is incremental as it builds on existing bio-inspired approaches.
The paper tackles the problem of achieving Artificial General Intelligence (AGI) by analyzing biological systems, concluding that evolutionary tinkering and hierarchical architecture for contextual information processing are key requirements.
Our fascination with intelligent machines goes back to ancient times with the mythical automaton Talos, Aristotle's mode of mechanical thought (syllogism) and Heron of Alexandria's mechanical machines. However, the quest for Artificial General Intelligence (AGI) has been troubled with repeated failures. Recently, there has been a shift towards bio-inspired software and hardware, but their singular design focus makes them inefficient in achieving AGI. Which set of requirements have to be met in the design of AGI? What are the limits in the design of the artificial? A careful examination of computation in biological systems suggests that evolutionary tinkering of contextual processing of information enabled by a hierarchical architecture is key to building AGI.