Intelligent scaling factors for use with evolutionary strategies-based artificial intelligence (AI)

One or more, non-transitory, computer-readable media storing computer-executable instructions which, when executed by a processor on a computer system optimize an application of an evolutionary-strategy-based application of Artificial Intelligence (AI), said application being performed on a pre-determined surface, said application directed to reproducing a human voice using an electronic sound-reproducing device, said application using the AI to help electronically reproduce the human voice to a replicate reality, said processor configured to:

• select a first group of candidate replicate realities;
• determine a mean and standard deviation associated with the first group of candidate replicate realities;
• use a static scaling factor to formulate a size of a population of candidate replicate realities for generation;
• use the mean and the standard deviation to generate, according to the size, the population of candidate replicate realities;
• select a second group of candidates from among the population of candidate replicate realities, wherein each member of the second group of candidate replicate realities is closer to a minimum value of the surface than a remainder of the population of candidate replicate realities, said remainder of the population of candidate replicate realities being formed from a group of non-selected candidates among the population of candidate replicate realities, and wherein the minimum value reflects the convergence of the replicate reality and the human voice, and the minimum represents the location of least quantifiable error.

A method for optimizing an application of an evolutionary-strategy-based application of Artificial Intelligence (AI), said application being performed on a pre-determined surface, said application directed to reproducing a human voice using an electronic sound-reproducing device, said application using the AI to help electronically reproduce the human voice to a replicate reality, said method comprising:

• arbitrarily generating a first group of candidate replicate realities, said first group of candidate replicate realities comprising a number of candidate replicate realities;
• determining a mean and standard deviation associated with the first group of candidate replicate realities;
• formulating a static scaling factor based at least in part on the number of candidate replicate realities in the first group of candidate replicate realities;
• selecting a portion of the first group of candidate replicate realities to form an evolved group of candidate replicate realities;
• and
• using the mean, the standard deviation and the static scaling factor to generate a new population of candidate replicate realities based on the evolved group of candidate replicate realities;
• wherein each member of the evolved group of candidate replicate realities is closer to a minimum value of the surface than a remainder of the first group of candidate replicate realities, said remainder of the first group of candidate replicate realities being formed from a group of non-selected candidate replicate realities among the first group of candidate replicate realities, and wherein the minimum value reflects the convergence of the replicate reality and the human voice, and the minimum represents the location of least quantifiable error.