Ideas like entropy and thermodynamic principles have little relevance as fundamental principles for contemporary AI / AGI. The relevance to these ideas are artifacts of Shannon's thinking. They are not central to the ideas that might be abstracted and utilized effectively for creating artificial intelligence.
One mind may make sense of external data that would look like entropic dissipation (or noise) to another. Entropy is therefore both relative and subjective. So, for example, if you were able to measure the 'entropy' of a system of related events in an AGI program and wanted to use this measure to determine causality, you would end up measuring the ability of the system to organize previously learned knowledge around the events. Relative subjective 'entropy' could be reduced by any organizing methods regardless of the usefulness of those organizing methods. Therefore, the thought that this method can serve as a fundamental principle of mind is misdirected. You can seize on any principle of subjective state of mind and try to claim that it is somehow fundamental to mind but it has never taken anyone where we want to go. You might abstract principles from other theories and say that they are relevant to some situation or some method of analysis, but to try to use them as the basis as a fundamental principle of mind is really out there. The fundamental principles of AGI have to be fundamental to thought. I have applied the method of my criticism to my own thinking and I have examined this criticism from different points of view and I still think this is a substantial critical view point. Jim Bromer ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657 Powered by Listbox: http://www.listbox.com
