A theory of mind (TOM) for artificial intelligence (AI) is presented, with links to on-line records of the creation of the TOM and to implementations of the theory in AI software.
The above diagram was created ca. 1991 on the Commodore Amiga
and was published on
Fred Fish public-domain disk #411.
The AI Mind recognizes an image of a dog. An associative tag carries a red spike of activation from the visual engram of a dog over to the neuronal concept-fiber of the idea of “dog”. Syntax for thinking about “dog” flushes out the concept of “dog” and fetches the Latin word “canis” for “dog” in the auditory memory channel. Then the AI Mind attaches the proper grammatical ending to the stem of the word for “dog”.
Although only a single, perpendicular concept-fiber for "dog" is seen in the MindAnim, we must assume that each concept in our brains is not a so-called grandmother cell, but rather a redundant assemblage or associative accretion of concept-fibers all holding or mediating the same concept.
3. Original Theory Documents
4. Software Implementations
5. Discussion Forums
6. Associated Web Materials
Table of Contents (TOC)
cover image of "Artificial Intelligence in Ancient Latin"