AUDIO + ARTIFICIAL INTELIGENCE

SUMMARY.

Undoubtedly, AI will be an integral part of our daily lives in the coming years. Machine learning, a branch of computer science, enables AI to acquire the ability to learn on its own. Programmers achieve this through machine learning algorithms.

The big question is: how will this impact the audiovisual arts?

The term "machine learning" is often confused with "artificial intelligence," although the former is just a subfield of the latter. Machine learning refers to a computer's ability to learn without being explicitly programmed. In its most basic form, machine learning uses algorithms that process input data and predict output values within an acceptable range. As it is fed more data, the system adjusts its operations and improves its performance, developing a sort of "intelligence" over time.

A notable example is Google, which has developed "Magenta," an experimental project based on neural networks. This synthesizer learns the fundamental characteristics of an instrument and then combines its sound with others, creating completely new and unique compositions.

However, the applications of machine learning are not limited to music alone. At the MIT Computer Science and Artificial Intelligence Laboratory, an algorithm was presented that can "guess" sound from a video. This system analyzes the sound properties of each frame and pairs them with the most similar sounds in its database. Once gathered, the algorithm automatically assembles them to generate a sound coherent with the image.

Will this be the end of post-production?

Certainly not. Immediate technological advancements will provide experts with new tools to refine their work rather than replace it.

There are many other fascinating applications worth exploring…