My talk Toward Semiotic Artificial Intelligence at BICA 2018 in Prague is available online:

Here two slides from the talk containing two examples of how semiotic cognitive grounding can achieve self-reflection (and self-improvement at the level of a reflective interpreter). My paper Toward Semiotic Artificial Intelligence explains how to interpret these code samples.

Reflecting on the output procedure

Reflecting on the successor function

I believe that doing machine learning without grounding, or confining grounding only to robotics implementation of sensory-motion, is an unnecessary burden on our research objectives. Yet we live in a world full of computers which do not know what computation is.

Collodi's "The Adventures of Pinocchio" became a metaphor the quest for artificial intelligence. The story of Pinocchio the Marionette who becomes a real boy began with a magic piece of wood. In my view, semiotic cognitive grounding can be that magical element in our quest.