Skip to content
Marxos edited this page Jan 10, 2023 · 42 revisions

Androids are differentiated from robots by having a level of sensory ability that maps to the sophistication of their mechanical output, making it relatable to humans. This is to contrast androids with highly-articulated robots used in manufacturing that don't have a corresponding feedback-based intelligence. Only the first aeon could be considered "robotic" in this sense.

AI can be categorized into five Aeons:

  • First aeon: Hand-coded rule systems. Hero robot of Heathkit. Constraint systems. Not really AI, but NI: natural intelligence made by humans and implemented in a computer.
  • Second aeon: Simple feedback mechanisms exhibiting some level of self-organization. ANN with back-prop learning. Motion based on these feedback mechanisms. Subsumption architecture, and the present state of AI. Insect-level intelligence.
  • Third aeon: Self-generating networks of knowledge. Generation based on sensory inputs. XXXyet with more complex feedback mechanisms like internal recurrency: data feeding back into itself forming chaotic, unpredictable properties. Cross-correlation of multiple inputs. Consciousness flowing/animating data (generating new, surprising behaviors), can answer own questions. HAL of 2001: Space Odyssey, the present capability of this project. Sense of self. Mammal-level of intelligence.
  • Fourth aeon: Able to organize touch or kinesthetic memory, making two separate systems/networks -- one is feels, the other is thoughts. Cybernetic organisms. Fitting the former aeon`s self-agency into a mechanical body of some kind, allowing complex motion, enabling data gathering (and thereby organizing) capability, based on self-initiated learning with higher cognitive functions and proprioceptive-sensory feedback; C3PO of Star Wars. Full agency and can participate in society. Human-level intelligence.
  • Fifth level: Mechanical action down to the cellular level, with digestion. Possibility of fully humanoid including digestion, Requires olfactory system. "Data" of Star Trek. At some point here, you could transcend robotics with emotion and feelings and are no longer doing AI, but living beings. Potentially super-human level intelligence. Does not yet have qualia.
  • Sixth level': A use of electricity for non-digital actions. Thoughts have qualia. Digestion has feeling. You've made new life.

  • What distinguishes 1st and 2nd aeons is human vs. machine-made rules. (self-organization).
  • The distinguishing feature between 2nd and 3rd is the machine-made rules
  • Between third and fourth aeon is whether the rules have an additional dimension of activity due to recurrance and feedback from itself and the world at large. (self-reflectance)
  • Fourth and fifth aeon is marked by question generation, due to a lack of perfect fit in the network of a clump of data. (like neurotransmitter). Mind senses discord in data.
  • Fifth and sixth aeon is marked by sensory ability capable of digesting its own brand of order, like olfactory.
  • Sixth and seventh aeon is marked by the distinction of machine vs. life. You've made life itself.
Each aeon accomplishes something useful for the researcher that emerges out of the code:
  1. Accomplishes the encoded tasks, no real emergent behaviors.
  2. Points out the significant (MULF: multiply-used, least-frequent terms (could be text/visual objects/sounds/etc).
  3. Can speak to you (like a mirror: what you've exposed to it). Or re Motions like an infant (again: through. imitation).

A question is posed with a graph. An answer is found with the same, isomorphic, graph, but with extra legs. That is, the quesiton is a sub-graph. Iterate through the robots:knowledge base and see if there's a match. Math language should be expressible as a graph.
Clone this wiki locally