Bay Area Artificial Intelligence Meetup Group Message Board › Blog Entry - Sergio Toporek - In Praise of Uncertainty
Blog Entry - Sergio Toporek - In Praise of Uncertainty
| Alex Gaputin | |
|
http://shop.toporek.c...
 Sergio Toporek - Fractal Artist Talks about how our models of reality are arbitrary and in the end we don't know anything that we can really be sure about. ______________________________________ ___My take away / connection to earlier thoughts: The universe is too complex for us to have correct models of it. We need to focus on AI architecture that can feel comfortable with uncertainty. Looking at the probability of anything (0 - 100%) Deductive logic only populates 1 / infinity, while inductive logic is 100% minus (1 / infinity). Deductive logic can be used as a special cast of inductive logic with a probability of 100%. Math/Physics/Logic tools can easily be memorized though habits. Feelings/Language are hard to model through Math/Physics/Logic but are easy to model through habits. Habits don't need certainty to function well, Math/Physics work best with certainty (uncertainty is costly) |
Powered by mvnForum
