The short answer is: sort of. The real question is how the final product sounds.

Sometimes, new clients send me links to older podcast episodes they want professionally edited as references. I can often identify two issues right away: an unnatural, “overprocessed” sound that’s clearly artificially enhanced and a disruption in the natural flow of the conversation. When I ask if they’ve used some sort of “AI” editing tools, the answer is always yes.

So, in recent years, artificial intelligence (AI) has made significant strides in various fields, including media production. Among these advancements is the use of AI in podcast editing. While AI can be an incredibly useful tool, relying on it entirely to edit podcasts can more often than you think result in a product that sounds worse rather than better. Why?

One of the most critical aspects of podcast editing is maintaining the natural rhythm and flow of speech. Human conversation is filled with nuances – pauses for emphasis, varying speech speeds, and natural breaks that AI fails to recognize and might cut conversations in ways that disrupt this flow, making dialogues sound choppy and unnatural.

Another significant issue with AI editing is its inability to handle breath ins and subtle speech cues. It can easily cut a breath in half, leaving abrupt transitions. These cuts become obvious and jarring to listeners, significantly diminishing the “smoothness” of the podcast.

If we delve even deeper, there are a lot of other things AI can’t handle perfectly. At least, not yet:

AI may struggle with complex edits that involve multiple speakers, background noises, or require seamless transitions. It may also have difficulty understanding the subtleties of human language, such as humor, sarcasm, and cultural references. While it can detect certain types of errors, it may overlook others or misinterpret sounds. All of these challenges can result in edits that inadvertently change the meaning of the podcast or remove crucial content.

The “AI” speech re-synthesis utilized in Adobe Podcast, Descript, and others is far from exceptional. While it may improve the quality of average-sounding audio, it struggles with poorly recorded audio. It’s definitely not a “studio sound” as advertised, and aren’t replace professional recording setups.

So, to summarize: AI can assist and enhance the editing process, but it cannot magically do it all for you and replace professional human podcast editor because it lacks the intuitive understanding and subjective judgment that human editors bring to the table.

On the contrary, AI can be a fantastic aid, helping with tasks such as noise reduction, room tone suppression, dynamics and spectral balancing, and transcriptions. These tools can save time and streamline the editing process.

However, for better results, content editing, critical decisions, and final touches should always involve a professional human editor. AI should complement the editing process, not take it over.