How to Solve AI’s Common Sense Problem
Review of Ron Brachman and Hector Levesque, “Machines Like Us”:
“I don’t think any of the current neuro-symbolic systems account for the difference between common sense and the more methodical, deeper kind of symbolic reasoning that underlies math and heavy-duty planning and deep analysis,” he said. “What I would like to see in this hybrid AI world is really accounting for common sense, making machines take advantage of it in the way humans do, and have it do the same incredible things for machines that it does for humans.”
I feel the same. A couple papers along these lines I know are:
Gabor Angeli and Christopher Manning, 2014. “NaturalLI : Natural Logic Inference for Common Sense Reasoning.” Proceedings of EMNLP, 534–45. https://aclanthology.org/D14-1059/
Hai Hu, Qi Chen, Kyle R., Atreyee Mukherjee, Larry Moss , and Sandra Kübler, 2019. “MonaLog: A Lightweight System for Natural Language Inference Based on Monotonicity.” ArXiv [Cs.CL]. arXiv. http://arxiv.org/abs/1910.08772.
This feels a lot like Description Logic. Has anyone done comparison? Has any recent work followed these footsteps?
Originally posted on LinkedIn.