reasoning depend on motivation
Table of Contents
Reasoning would only be valid if assume a motivation.
1. example - pathfinding
Take a pathfinding agent in 2d grids, with a given starting point and destination.
- reasoning
- would be to decide/generate next action(forward, turn, terminate)
- motivation
- would be to get to the goal; or minimize travlled distance/cost;
- reasoning depend on motivation
Any strategy to generate next action would only be valid if it could get the agent to the goal, or could in some way minimize travelled distance.
- proving
A formal proof of the reasoning strategy would include a mathematical definition of the chosen motivation, and the system arriving at a state where the motivation is satisfied/resolved. the correctness of A* search could be verified by checking that if destination is reachable, it would be reached eventually.
Backlinks
old index
(Academically speaking)
- artificial general intelligence
- artificial child
- kernel learning
- thesis
- discussion
- research workflow
- 2 kinds of academic work
- inovation by using existing technique
- autonomous mobile robotics
- discipline
- customizing reasoning
- reasoning depend on motivation
- motivational reasoning
- biological process is computation process
- perception being solution
- research on new knowledge
- preview before lecture
- knowledge representation
- start state
motivational reasoning
As reasoning depend on motivation, for different motivation, there should be corresponding reasoning, to satisfy/resolve the motivation.
And that sure bring a parameter to intrinsic motivation and generated motivation, in that a higher-order reasoning strategy, or a even higher-order self-programming scheme, for them to be valid, it may have to be proved that such reasoning strategy [scheme] would satisfy/resolve the motivation [scheme].