Can the “conditioned reflexes” of Neural Networks Ruin Creativity?

The 2023 Hollywood strike resulted in a virtual victory for the Writers Guild of America (WGA) over the Alliance of Motion Picture and Television Producers (AMPTP) after reaching an agreement that neural networks would not be used by producers to write screenplays.

If the screenwriters themselves use the capabilities of neural networks in the form of hints for plot moves to the author’s plan, the screenwriters retain all copyrights.

Of course, the struggle of the “Hollywood monsters” was primarily for big money, but there is another side to this victory. The screenwriters managed to protect their rights to professional creativity, at least temporarily, from the producers’ plans to widely use the capabilities of neural networks.

However, the synthetic “conditioned reflex” of neural networks, as a result of their training to average massive initial data to identify common characteristics, is very tempting for producers in terms of speed of implementation and cost.

Of course, this approach interested them in practical applications, for example, when searching for stories based on given parameters. But using this approach to generate a supposedly new one involves averaging what already exists. Speaking in the conventional language of mathematics, the number of combinations from a finite number of initial options is itself finite, although very large.

Any creativity will certainly be lost in the process of automatic enumeration and averaging, which involves merging any stories, characters and everything else with forced simplification. Of course, this was intended solely for the sake of commercial gain from the “conditionally new” instead of the original author’s idea.

Prolonged exploitation of the average “conventionally new” can lead to the loss of attractiveness and the collapse of many creative professions, which threatens to result in the loss of the human ability to create, and which, according to many signs, is already beginning to happen in modern society.

The capabilities of neural networks are undoubtedly useful in optimizing and increasing the efficiency of many production processes. Especially in the manufacture of standard mass products, but not when something unique and especially valuable is created by the hands of a master.

In the same way, the synthesis of “conventionally new” in creativity poses a serious danger in impoverishing and simplifying the development of the cultural and aesthetic level of civilization. Filling with faceless averaging will quickly lead to sketchiness and stereotypes, which will cut off people of creative professions and deprive true art of human originality, uniqueness and individuality.

The screenwriters won the first battle to save their profession. But it’s time for musicians, artists and other creators to prepare for their battles, as businesses will strive to use neural networks in all areas of activity, thereby displacing the human personality under the pretext of moving towards a utopian consumer future.

But will people be needed in this future if they lose their divine abilities to create and become like faceless robots or simply useless creatures?