Your browser is out-of-date!

Update your browser to view this website correctly. Update my browser now


Hey ChatGPT! What Can A-I Do for Radio?

Here are the answers it gave me

Are you paying attention to AI? Best get moving.

Artificial intelligence has been the focus of a number of Radio World articles in the past couple of years, including an ebook in 2020 and our news coverage of companies like Veritone, Super Hi-Fi, Adthos, Microsoft and MultiCAM Systems.

But the topic is changing so fast that most of those articles are now dated. And now, with people learning what “generative” and “natural language processing” mean, and with the internet buzzing about DALL-E and ChatGPT and all the fun things they can do, the professional implications for radio of these kinds of tools are really dawning.

So here’s a thought-starter. I asked ChatGPT — the model from OpenAI that interacts in a conversational way — to give me 10 bullet points on ways that generative AI technology might be used in the radio broadcast industry and its technology infrastructure.

Here’s what it spit out (numbered items are from ChatGPT, while the rest of this story is by me — a caution that will become increasingly relevant, I fear):

  1. Content generation: AI can be used to generate news articles, weather reports and other types of content for radio broadcasts.
  2. Music selection: AI can analyze listener preferences and playlists to recommend and automatically schedule music for radio stations.
  3. Voice synthesis: AI can be used to create virtual DJ or host voices that can interact with listeners and present music and content.
  4. Ad targeting: AI can analyze listener demographics and listening habits to deliver targeted advertisements to specific segments of the audience.
  5. Speech recognition: AI can be used to transcribe and analyze spoken content in radio broadcasts for improved search and accessibility.
  6. Quality control: AI can monitor audio quality and signal strength in real time to automatically adjust levels and reduce technical issues.
  7. Network optimization: AI can be used to optimize network infrastructure and manage bandwidth usage for improved reliability and efficiency.
  8. Predictive maintenance: AI can analyze equipment usage patterns and predict when maintenance is required to prevent downtime.
  9. Fraud detection: AI can detect and prevent fraud in ad trafficking and billing systems.
  10. Analytics and reporting: AI can provide real-time analytics and reporting on listener demographics, ad performance, and other key metrics to inform business decisions.

[Sign Up for Radio World’s SmartBrief Newsletter]

This is an image created by the AI tool DALL-E-2 to my written prompt “Cartoon drawing of a broadcast radio tower with a giant threatening figure behind it to represent artificial intelligence.”

As you probably are aware, AI is already doing some of these things within various products in radio and audio. But I think we can look at the list above and bet that the technologies that drive our industry are in for potentially massive changes. 

(An aside: While I was putting this story together, I received a press release from Radio Workflow announcing that the company is using AI to help radio stations create ad copy. It promises the following: “With ‘Write My Ad’ and’ Fix My Ad,’ radio stations can now produce high-quality, compelling ad copy in seconds without the hassle of writer’s block.” So file that under No. 1 in the list above.)

I then asked the chatbot specifically to list five radio engineering tasks that could be performed by artificial intelligence, and got this:

  1. Signal analysis and optimization: AI can analyze radio signals to identify and resolve issues with signal quality and strength, optimizing network performance.
  2. Equipment monitoring and maintenance: AI can monitor and analyze equipment usage patterns to predict when maintenance is required, reducing downtime and improving equipment longevity.
  3. Interference detection and resolution: AI can detect and resolve interference issues in real-time, improving signal quality and reducing listener complaints.
  4. Frequency management: AI can manage and optimize frequency usage for improved efficiency and reduced interference, allowing for more channels to be utilized.
  5. Network planning and optimization: AI can analyze network utilization and performance data to optimize network infrastructure and improve overall system efficiency.

You will probably would get somewhat different answers based on how you ask the question; but it’s clear that the technology has significant ramifications in all of our endeavors.

In Radio World itself, we’re not using language models to create content while our company and the RW team understand them and the broader field of AI better. If we ever do, I’d want that content labeled as such.

But there are so many ways that such technology could be of help in other ways, from identifying potential topics for articles or digging up useful hashtags to performing repetitive administrative tasks like proofing copy. We’re all just learning.

What do you think are the implications of this new tech? Are there specific tasks related to your daily work that would benefit from the support of AI, or that might be replaced by it? Do you have any experience using AI, for instance natural language tools like GitHub Copilot to create code? Email me at [email protected].

And for a further take on this topic, read the commentary “Where Does Chat GPT Fit Into the Radio Industry” (and particularly in radio engineering) by Todd Dixon.