Learning in embedded systems
Neural networks and intellect: using model-based concepts
Neural networks and intellect: using model-based concepts
Machine Learning
The cog project: building a humanoid robot
Computation for metaphors, analogy, and agents
| Hi-index | 0.00 |
In a previous paper the author posited that a sapient system would be a robot that performs abductions on a classification of concepts. In the present work, I propose the three steps that would be required for a robot to reach an initial state of sapience. Each step begins with improved mechatronics, continues with improvements in pattern recognition and ends with improvements in symbol manipulation. The first step starts with an anthropomorphic robot. I propose that if in the last step we implement adductive reasoning we have arrived at an initial sapient system. A low level robot showing the first stage evolution of a simple robot is presented. The implementation evinces many features of biological evolution such as misfitness, adaptation, exaptation, mutation, and co-evolution.