Cyborg AI Minds are a true concept-based artificial intelligence with natural language understanding, simple at first and lacking robot embodiment, and expandable all the way to human-level intelligence and beyond. Privacy policy: Third parties advertising here may place and read cookies on your browser; and may use web beacons to collect information as a result of ads displayed here.

Monday, March 27, 2017

pmpj0327

Perl Mind Programming Journal

[2017-03-26] The main problem with the ghost184.pl Perlmind is that it does not supply an automatic carriage-return CR-13 if the human user neglects to press the Enter-key. This defect prevents the AI from going back to its own chains of thought when a human user has begun but not completed a message of input.

[2017-03-26] We may be able to fix the problem by supplying a CR-13 carriage-return when the input loop of the AudInput module is making its last loop. We try it, and it seems to work, but we find that ReEntry() does not work after an incomplete human input is supplied with a CR-13 in the AudInput module. We deploy a diagnostic message or two and we learn that the $len variable is not at zero when ReEntry() is called, thus perhaps interfering with the proper function of AudInput. Let us try setting the $len variable to zero at the start of the ReEntry module.

Ghost 185.pl Strong AI enters each input character into memory.

[2017-03-27] In the Strong AI ghost185.pl it is time to change from the re-entry of complete sentences back into the Mind to a more immediate re-entry of each phoneme (character) back into the Mind. In AudInput(), when we start sending each input character directly into AudMem(), nothing gets recognized, perhaps because we need to change the characters to uppercase. We also start incrementing time "t" before AudInput() calls AudMem(), and we start getting auditory recognition of an input word. When we preserve the inner loop of AudInput but we comment out the outer loop for whole words, we start getting a display of the storage of input in memory.

Next we need to implement the switch-over from storing input to generating output. It appears that we are not getting memory-storage of output because time "t" is not being incremented. We also need to turn a nested AudInput() loop into just a one-time sequence. Then in Speech() we set $pho before we call AudInput() for the reentry of Speech() output. Suddenly we begin seeing both the input and the output as stored in conceptual and auditory memory. We tweak ghost185.pl a little and we upload it.