Skip to content
Mark Janssen edited this page Jul 10, 2019 · 42 revisions

Aeons of robotics:

  • First aeon: crude, generally hand-coded rule systems. Expert systems. Hero robot of Heathkit.
  • Second aeon: sorting data into new information. Some level of self-organization. Simple motion based on either hand-coded rules or simple feedback mechanisms. ANN with back-prop learning, subsumption architecture, and the present state of AI.
  • Third aeon: self-generating, self-organizing networks of knowledge. Recurrency: data feeding back into itself. The present state of this project (formally named "ISAAC").
  • Fourth aeon: consciousness or action with the data, can answer own questions. HAL of 2001: Space Odyssey
  • Fifth aeon: Complex motion, based on self-learning and proprioceptive-sensory feedback, C3PO
  • Sixth aeon: fully humanoid: including digestion, has olfactory system. "Data" of Star Trek
  • A rejected "seventh" aeon of robotics that include emotion and feelings is rejected because you're no longer building robots, but living beings.

  • 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).
Clone this wiki locally