There's been a lot of buzz about AI in podcasting, and I've noticed some reluctance among podcasters to use it. This reluctance often comes from associating AI heavily with large language models or from concerns about AI replacing creative work. But there are many AI tools out there that can be genuinely helpful for podcasters—tools that might not even seem like AI at first glance. In my opinion, more podcasters should be using these tools to make their work easier and more efficient. In this post, I'll explain when you should consider using AI (and when not to) and why it can be so beneficial.

AI Won't Replace the Human Touch

Before I go into when to use AI, let's start by talking about when not to use it. AI is not going to replace the stories or the human aspects of podcasting. At the end of the day, podcasting is about connecting and telling stories with other humans, and AI can't replace that. Just like in chess, where AI might be better at the game, people still love watching humans play against each other because of the stories, strategy, and connection. The same goes for podcasting—listeners want (and crave) real human experiences.

How AI Can Improve Your Podcast

That being said, there are definitely aspects where you should use AI for your podcast post-production. Instead of keeping it hypothetical, I'm going to concretely go through the places where it can actually be useful for you. Here are some of the use cases where I use AI for all the podcasts we produce, and maybe you'll find some of them helpful:

1. Cleaning Up Your Audio

Especially if you're starting out, you might not have the best studio gear, but you can use AI to make your audio sound polished. A few months ago, these tools were not great—the results were often spotty—but the current tools are much better, especially for cleaning up background hums, buzzes, and noises.

• Noise Reduction: AI tools can remove background noise, making your voice clearer.
• Voice Enhancement: They can boost your voice to sound more polished. Almost like studio sound.

There are both free and paid tools available. I've had mixed results with the latter (voice enhancement), but for basic cleanup (noise cleanup), many free options work great - close to 100% reliability.

2. Fixing Transcripts

If you've generated transcripts for your podcast, you might notice that they sometimes mess up vocabulary or proper nouns. You can use large language models to correct these errors. This is one of my favorite use cases. I kind of almost automated this workflow, so that it's pretty cool. I have perfect transcripts now that we can publish on websites.

• Better Accuracy: You can provide your incorrect transcript along with a "dictionary" of the correct names and terms, and the AI will fix it.
• Organized Formatting: It can structure the text to be easy to read and use.

I use it frequently for many podcasters. Clean transcripts are also helpful for creating show notes and chapter titles - because they will also have the right spellings.

3. Creating Chapter Timestamps and Show Notes

In my opinion, you shouldn't be doing this manually. AI can handle most of the work for you:

• Chapter Timestamps: AI can suggest natural breakpoints in your episode for easy navigation.
• Draft Show Notes: It can generate a first draft of your show notes, capturing most of the important details so you can focus on editing and refining.

I'm not saying it's going to replace exactly what you do, but I think you should get the first draft with AI and then kind of adjust it so that the tone comes through. It's really good at identifying those natural boundaries, and you need to work with it a little. I've automated 100% of the workflows, it's really nice, and I can focus on the fun aspects of it like making the show notes a bit funny

4. Picking Clips for Social Media

Instead of using off-the-shelf tools to create social media clips, I use large language models to help me pick the best segments based on specific criteria. I'll provide the full transcript and ask the AI, "Which clips do you think are good for this purpose?" and it will suggest clips that work well for promotion.

• Highlight Selection: AI can help find the most engaging parts of your podcast to share.
• Caption Ideas: It can even suggest captions to go with your social media posts.

This is a pretty useful use case that I've found for promoting podcasts.

5. Researching Guests

This is something I haven't personally used, but some podcasters I work with have found it incredibly helpful. Before every episode, they used to manually research guests. They still do. But now also, they use large language models and chatbots as research tools—either to generate interesting questions, summarize papers, or summarize the findings sent by the guest. This helps them come up with unique questions that haven't been covered in previous interviews.

Conclusion

AI can take care of the technical tasks of podcasting so you can focus on creating great content. It's here to support you, not replace you. Once the initial friction of learning these tools is gone, the workflow becomes smoother, and the benefits become clear.

By using AI as a helpful tool, you can enhance your podcast while keeping the personal touch that makes it special. I'd love to see more podcasters embrace AI—it really does make the process easier and more enjoyable.

Have you used AI in your podcasting process?