AI revolutionises the future of radio broadcasting

AI Revolutionises the Future of Radio Broadcasting

The radio industry is witnessing a transformative shift as artificial intelligence (AI), particularly its generative form embodied by tools like ChatGPT, becomes increasingly integrated into various operations. This modern technological wave, which has its origins as far back as the 1950s, is gaining momentum with advancements in AI forms, including multimodal, agentic, and customised enterprise generative models.

The new AI variants, such as Meta’s Llama 2 and Mistral’s Mixtral, are leveraging large language models (LLMs) to process vast datasets for natural language processing, surpassing previous capabilities. One significant development is multimodal AI, which involves three components: input, fusion, and output. This system can process diverse data formats more efficiently, offering a nuanced and intuitive approach to data interpretation.

OpenAI’s recent launch of GPT-4o marks a step forward in multimodal AI, enhancing human-computer interaction across audio, video, text, and images. Meanwhile, agentic AI introduces proactive applications capable of understanding and adapting to their environment with minimal human intervention. This versatility allows businesses to tailor AI functions to specific needs, moving beyond generic models like ChatGPT.

The adoption of AI in radio broadcasting varies globally. Companies like Radio.Cloud are embracing these advancements, focusing on text-based AI to generate unique virtual voice tracks. Andrew Scaglione of Radio.Cloud reveals their extensive use of LLMs, particularly GPT-4o, to innovate in creative production and idea generation.

A contrasting example is Veritone, which has been leveraging AI since 2014. The company has focused on artificial narrow intelligence (ANI), a precursor to current AI models, showcasing pragmatic uses in processes like transcription and translation. Zack Zalon, CEO of Super Hi-Fi, anticipates that AI will significantly reduce the time taken to create and manage radio content, enhancing the listener experience.

The rapid pace of AI development is both exciting and challenging for industry professionals. Joe D’Angelo from Xperi notes the efforts of radio stations to find a balance between efficiency and maintaining their reputation as a trusted media source. Meanwhile, Dmytro Bielievtsov of Respeecher points out that the current generation of AI lacks a complete understanding akin to human intelligence, with capabilities still in development.

In the UK, Bauer Media Audio has started to integrate AI to support rather than replace human roles, streamlining the production and editing process. By employing AI, they can quickly amend recorded content, improving efficiency without needing to recall voice artists into the studio. Bauer Media is also exploring how multimodal AI can assist in audio training and production, offering suggestions to fine-tune recordings.

The BBC is pushing forward with its own initiatives, having launched 12 pilot schemes leveraging generative AI in October 2023. These pilots aim to maximise existing content’s value and enhance user experiences, such as converting live radio commentaries into text for web publication.

As AI continues to evolve, its role in radio remains supplementary, aimed at enhancing creativity and operational efficiency. The industry, with its mix of traditional and innovative approaches, is navigating these changes, seeking to harness AI’s potential while retaining the essential human touch that has always been central to radio broadcasting.

Source: Noah Wire Services