The sound of change; AI versus AI

The interaction between humanity and technology in music and audio post-production has become a highly debated topic. 

Artificial intelligence [AI] has emerged as a transformative force, reshaping our approach to audio and various creative outputs. Whether we believe in its merits or not, it’s here to stay. Even my 75-year-old mother is embracing AI assistants for her everyday writing tasks.  

Artificial intelligence [AI] has emerged as a transformative force, reshaping our approach to audio.

The rapid integration of AI into everyday life mirrors its influence in the professional realm, where it has the potential to revolutionise workflows and enhance creative processes. This evolution makes me want to explore how AI can boost efficiency while preserving artistic integrity across different creative fields.

Above: AI can help the industry without replacing human jobs. 

In audio post production, AI algorithms provide significant assistance by automating repetitive tasks such as noise reduction — removing unwanted noise from audio signals — and EQ matching, which adjusts the level or amplitude of specific frequencies to achieve a clear and balanced mix. This enables sound designers and mixers to focus more on the intricate and creative aspects of their work. Some of my preferred tools include Izotope and Accentize for denoising, ChatGPT for drafting sound design briefs, and Eleven Labs for generative speech and bespoke background content.

Generative speech tools can also be incredibly useful for dialogue editors and mixers.

Generative speech tools can also be incredibly useful for dialogue editors and mixers. I have been in countless commercial sessions where a client asks for a version of a line that ends with a downward inflection instead of an upward one. It would be incredibly easy to train AI to model the voice actor and generate the new ending. Instead of the current, extensive process of bringing an actor back into the studio, we could resolve the issue in just five minutes. This would benefit the actor as well because, by saving time, they could book other paying jobs instead.

By using AI in this way, users treat it as a tool rather than a method for creating original content that could potentially replace human jobs.

Above: Nike used artificial intelligence to showcase two versions of Serena Williams playing against each other.

Recently, Nike's campaign Never Done Evolving featuring Serena Williams utilised AI to showcase a match between her younger self at her first Grand Slam in 1999, and a more contemporary version from the 2017 Australian Open. This innovative approach not only highlighted Serena's enduring legacy and evolution but also resonated strongly with audiences worldwide, demonstrating the creative potential of AI in storytelling. Without AI, achieving something like this would have been more challenging, and likely costly.

While companies can establish guidelines, the biases, morals and creative preferences of the individuals coding the AI inevitably shape its outputs, making it inherently biased.

However, embracing AI’s potential involves more than just technological innovation. It also requires a commitment to ethical practices to ensure fair compensation for actors and protect their creative rights. Clear guidelines on usage limits, such as implementing single-project-based agreements or capping instances of generated speech, are crucial for safeguarding the integrity of artistic expression and maintaining ethical standards in the industry. 

Despite the remarkable advancements of AI in the audio and creative fields, it is crucial to acknowledge that AI is fundamentally influenced by its programmers. While companies can establish guidelines, the biases, morals and creative preferences of the individuals coding the AI inevitably shape its outputs, making it inherently biased. This inherent bias highlights the importance of maintaining an objective perspective and not relying entirely on AI without critical scrutiny.