Nanograte: I do not want to be rude, but if you have interpretable AI networks
figured out then just apply those rules to an actual AI program. I cannot
understand what precisely you are saying, but the graph or model is never
mature (in a constrained sense of graph theory) if you introduce hypergraphs,
which are necessary. It is not only Godelean self-referential step-state
contradictions, but there are an extensive range of automated referential
logical issues (like circular reasoning) that would create complications that
need to be managed. I have circular reasoning "all figured out" but there are
other more subtle problems that cannot be foreseen by the programmer at a
purely abstract level. I am not claiming to be an expert (at graph theory or
hypergraphs and so on) but I believe that I have thought about these issues at
a more intuitive level.
------------------------------------------
Artificial General Intelligence List: AGI
Permalink:
https://agi.topicbox.com/groups/agi/T1675e049c274c867-Ma1e5122b12a75539dacd7068
Delivery options: https://agi.topicbox.com/groups/agi/subscription