AI-Powered Year-in-Review Comes to Podcasting
Online podcast recording platform Riverside has launched its own take on the year-end recap trend popularised by Spotify’s “Wrapped.” The feature, called “Rewind,” generates three short, personalised videos for podcasters, offering a snapshot of their year behind the microphone.
Rather than focusing on traditional metrics such as total recording time or episode count, Riverside’s approach leans into personality. The tool produces a 15-second montage of laughter, stitching together moments where podcast hosts crack each other up. A second video takes a similar approach, compiling repeated instances of hosts saying “umm.”
The third video uses Riverside’s AI-generated transcripts to identify the single word used most frequently during recordings—excluding common filler words such as “and” or “the.”
When Data Meets Personality
The results can be unexpectedly revealing. On a podcast focused on internet culture, the most-used word turned out to be “book.” The likely reasons were practical rather than philosophical: subscriber-only “book club” episodes and frequent mentions of an upcoming book release by one of the hosts.
Across the same podcast network, another show, Spirits, found its most frequently spoken word was “Amanda,” simply because one of the hosts shares that name.
When these Rewind videos were shared internally on Slack, they sparked laughter—particularly the supercut of repeated “umm” moments. There is an inherent humour in seeing conversational habits reduced to short, looping clips.
Fun, But What’s the Real Value?
While entertaining, the videos also highlight a broader issue facing creative professionals. Many digital tools are becoming increasingly saturated with AI features that users neither asked for nor necessarily need. A montage of repeated words or filler sounds may be amusing, but it offers little practical value beyond a brief chuckle.
This raises a bigger question for the podcasting industry: are these AI-driven features enhancing creative work, or merely adding noise?
Automation Versus Creativity
The arrival of Riverside’s Rewind comes at a time when many podcasters and editors are facing shrinking opportunities, partly due to the same AI tools now embedded in production platforms. AI excels at automating mechanical tasks—such as removing dead air or transcribing audio—which can save significant time and improve accessibility.
However, podcasting is not purely mechanical. While AI can generate transcripts quickly, it cannot make nuanced editorial decisions. It cannot judge when a tangent is genuinely funny or when it disrupts the flow of a story. Human editors still play a critical role in shaping compelling audio narratives.
High-Profile Limits of AI Audio
The challenges of AI-driven audio content have also surfaced in news media. Recently, The Washington Post began rolling out personalised, AI-generated news podcasts. On paper, the idea offers clear cost savings: automating research, production, and distribution instead of relying on human teams.
In practice, the experiment exposed serious flaws. The AI-generated podcasts included fabricated quotes and factual inaccuracies—an unacceptable risk for a news organisation. According to Semafor, internal testing at the Post found that between 68% and 84% of the AI podcasts failed to meet the publication’s editorial standards.
This outcome reflects a fundamental limitation of large language models. They are designed to generate statistically probable responses, not to reliably distinguish fact from fiction—particularly in fast-moving news environments.
A Useful Reminder in the “AI Boom”
Riverside deserves credit for creating an engaging, light-hearted end-of-year product. Rewind succeeds as a piece of entertainment and brand engagement. At the same time, it serves as a reminder of where AI adds value—and where it does not.
As AI continues to spread across industries during the current “AI boom,” businesses must be discerning. The challenge is not whether to adopt AI, but how to ensure it genuinely serves users rather than producing novelty features that amount to little more than digital clutter.
0 Comments