How Computers Are Learning To Be Creative
Computers are gaining creative capabilities as we speak. The AI “Benjamin” was trained on a variety of b-movie and sci-fi scripts. Although actors and directors used to interpret the performance, this is no longer necessary. Similarly, children learning how to use computers gain technical vocabulary and become more comfortable with their keyboards and screens. The process of learning to use a computer is similar to learning how to read. Human creativity is different from computer-generated creativity because humans rely on their intuition.
A recent paper by computer scientists at Google reveals a fascinating insight into how computers learn to be more creative than we are. Rather than using the same old boring old programs, computers are now capable of learning new skills. Researchers have created photo-realistic eyes that allow computers to be more creative. The authors explain that the process of creativity involves a psychological model of creativity called the Geneplore model.
Researchers have tried to translate the human mind’s ability to create new objects and understand it using AI models. A baby learns by looking at many examples of the same object. After seeing many examples, the baby associates a particular pattern of windows and wheels as a “car”. The same pattern is used in building an artificial intelligence model. A human might only need a few examples to classify an item, but a computer may require thousands to recognize it.
This trend began as a way of creating new applications. But computational creativity eventually led to gaming. AngELINA, an AI program developed by Michael Cook, created games from news articles. By using current affairs text analysis and hard-coded design techniques, AIs began to mimic human creativity. These AIs are already being used in many fields, although they are not at the same level of humans.
Computational creativity is a field of research that spans multiple disciplines. Computational creativity includes mathematical models, autonomous creative system, and interactive tools that support human creativity. In the late 1990s, computational creativity was established as a formal field of study and hosted its own annual conference. Creative AIs were able to create non-linear artifacts using machine learning and artificial neural networks.
The latest advances in AI are allowing humans to work with their AI. The AI is skilled at performing a variety of tasks and can come up with many concepts and rough drafts. It can also be a co-creator and inspire humans to create amazing artwork. Ultimately, humans choose which of the many AI-generated works are the most successful. These advances are already changing the art world.
