Skip to content

How is AI changing creative work? A radio team’s perspective

Listen to Story · 6 min listen
Quick Takeaway · Key Summary

When an interview ends, a radio team’s work continues. Someone listens to the recording, checks the speakers’ names and decides how a quotation fits its context. An AI tool can put a first transcript on the screen quickly. The work needed to make that text ready for publication may simply move elsewhere.

Imagine a student station feeds an hour-long recording into a tool and receives a draft minutes later. An editor then listens again, corrects a name and checks the sentences around a quotation to preserve the guest’s meaning. The tool might have shortened the first step; this scenario does not measure a time saving. The useful question concerns the whole path to a reliable result and who performs each part of it.

When the parts of a job change

It is easy to describe radio production as one task. In practice, researching a subject, building trust with a guest, recording, cleaning audio and making a publication decision call for different skills. A transcript tool can assist with one part of that chain; it does not automatically assume editorial responsibility for the whole interview.

The International Labour Organization’s research on media and culture describes different levels of exposure to generative AI across creative occupations. It also discusses creative control, consent and fair compensation. The technical possibility of automating one task does not settle the future of an entire occupation.

What does time saved actually mean?

Looking at a first draft, a team might be tempted to say that an hour’s work took five minutes. But between that draft and approved episode notes lie speaker identification, name checks, listening for context and, where relevant, permission review. Comparing two trials also requires attention to differences in recording length, audio quality and the number of speakers. For example, prepare one excerpt of an interview manually and a different, similarly long excerpt with tool assistance. For each, note the total time from preparation through final review and any errors remaining in the publishable version. The excerpts and editors’ experience may differ, so one trial is a starting point for further observation rather than a definitive productivity measure.

📻
RadioTEDU Live Stream

Listen while exploring this story

Tune into campus radio with non-stop independent music and culture.

The ILO’s 2026 review of empirical evidence reports uneven productivity findings across working conditions and notes that reported time savings have not always translated into measured output or earnings. The review also identifies the erosion of opportunities for younger workers as a risk. It does not measure a saving in a RadioTEDU workflow or establish that this risk has occurred at the station. A student team would need to observe its own work and review time to understand its result.

The invisible part of learning

Preparing a first transcript may be repetitive for an experienced editor. For a newcomer, it can also be a way to learn close listening, recognise a useful question and preserve the context around a quote. If a tool takes over the first pass, another question appears: where will the new editor develop those skills?

That does not mean an automated tool prevents learning. A team can listen back together and discuss its corrections. Recording who contributed, clarifying permission to process a guest’s voice and identifying who approves the final publication are also part of production. Sending audio to an external service calls for a review of its data terms and the institution’s own practices.

Experiment in the studio, decide at the desk

RadioTEDU’s about page presents student broadcasting and campus activity as a place to learn by making. In such a setting, a new tool can be judged by what the team continues to learn as well as by its speed. This is a suggestion, not a claim that RadioTEDU has adopted a particular AI policy.

Noticing the work hidden beneath a transcript does not require choosing a side for or against technology. Who listened, who corrected, whose words were published and who made the final decision? Those answers tell us a great deal about creative work.

RadioTEDU monthly

The month in podcasts.

Receive podcast episodes from the latest 30 days and upcoming RadioTEDU ticketed events. No daily mail or unrelated announcements.

This consent covers only the monthly podcast newsletter and ticketed event updates. See our privacy notice.

Live RadioTEDU Pick a station and start listening
0:00 0:00

RadioTEDU · Big5 Türkiye Watch Party

RadioTEDU Live Screening: Big5 Türkiye Watch Party. Thursday 19 November, 18.00, D032. Click for free tickets.