Artificial Intelligence For Podcasters — Podcasting Strategy Insights - Issue 2

Greetings from Juergen!

A.I. tools have been on the rise, accelerated by ChatGPT. This has led to a flood of recently introduced podcasting specific A.I. tools. My overall impression of the A.I. space in general:

Let's quickly launch some tools and see what sticks.

So I am having to balance my excitement about the potential in this space with the reality of innovators launching trial balloons of sometimes short-sighted or (as of yet) incomplete solutions.

But I definitely think A.I. in podcasting has tremendous possibilities, and I love seeing the progress!

Guide Additions

Nomono |A self-contained 4 person recording kit for capturing interviews in the field.

A self-contained recording kit for capturing interviews in the field. Completely wireless from microphone to cloud. How it works:

Hand out mics Attach them using the clip or included magnets.
Press record No levels to set. No settings to worry about.
Upload and enhance Auto-upload to the Web App, enhance and review. -Download your recording Use the original or enhanced version in post-production.

AI-powered audio enhancement and team collaboration in the cloud. Built specifically for podcasters.

My take on this: This entire platform integrates AI and hardware in an interesting way. The cloud and AI portions of their offering focuses on audio processing and podcast collaboration.

But the hardware is interesting and points to a trend in the way AI can help simplify a recording setup: A few years ago I was hired to travel to record podcast episodes for a large corporation during a conference. I traveled there with about 75 lbs of gear. High quality mics, a Rodecaster Pro mixing console, boom arms for each speaker etc. I think we are approaching a time when we will no longer need all that gear.

The AI algorithms that identify background noise and can make a clip on condenser mic sound great are improving at a rapid rate.

Ultimately I feel this tool is directed at the enterprise or mid-sized business podcasters, and difficult to afford for indie podcasters

Fathom: The podcast player of the future.

Discover podcasts at the speed of thought with mind-blowing Search, Transcripts, Clipping, and Highlights.

It was only a matter of time before A.I. features were going to appear in podcast listening apps, and here is an example.

My Take: I am particularly interested in the chapter feature, because I consider this the most valuable feature. It implies overcoming several challenges for an A.I. The content of a podcast episode must not only be understood, but separated into themes, each with a chapter sub-heading.

For us, the quality of this process has been low in most A.I. solutions we've observed so far. But the importance is huge, since auto-generating great summary show notes would save podcasters huge amounts of time.

On Artificial Intelligence

I hate reading content or watching YouTube videos that were obviously generate by A.I. — I can "feel the Matrix" at times, for example reading a thinly developed travel article.

So how should I ethically approach the use of A.I.? For me it involves guiding the A.I. and ensuring not only useful, unique and factual content, but also aligning that content with a point of view of it's human author. In testing various platforms this month, I am not sure that I achieved this entirely, but my own thinking is still evolving.

My prediction is that we will start placing even higher value on personal videos and content with obviously human qualities.

What follows are two examples and some hopefully thought provoking articles.

Example 1: Guided A.I. Writing about Podcasting Marketing

Here is a recent podcast marketing tips article from our blog that was written by an A.I., however with careful guidance and step-by step instructions.

Example 2: Show Notes - How Alex Sanfilippo is Helping Podcasters Succeed

Disclosure: Alex was kind enough to help me test a number of A.I. tools as part of this interview. We feature and review these A.I. tools on our Podcasting Resources Guide and the latest Podcasting Strategy Insights newsletter.

Rumble Studio: We conducted the interview using Rumble Studio, which is an A.I. asynchronous remote recording tool where I pre-recorded questions, which Alex then answered in his own time. The entire episode audio was mixed and generated using 1 click.

Swell and Melville: The show notes you see below were generated by Swell.AI and Melville. The A.I. generated content was left un-edited, something I do not ordinarily recommend, but left in place for the purposes of evaluating these tools. (thank you Alex!)

ChatGPT and Other Chat Bots Are a ‘Code Red’ for Google Search - The New York Times

A new wave of chat bots like ChatGPT use artificial intelligence that could reinvent or even replace the traditional internet search engine.

The question I ask myself is how integration with ChatGPT could be deployed in podcasting apps and platform interfaces. I think there could be future applications for both podcast listeners and podcast creators. Podcast and episode search and discovery Podcast editing tools Podcast ad placement and targeting

Artificial Intelligence and the Modern Podcaster

“Media production is now entering a phase where if you can dream it, it can happen,” says Jay LeBoeuf, head of business and corporate development at solution provider Descript. “And you no longer need to have an expensive studio or decades of training to accomplish your goals.”

How tech publisher Heise Online created their own synthetic podcast speaker - Aflorithmic

Built in just a few weeks, the cloned voice of a popular podcaster unlocked Heise to double the frequency of their successful daily tech news podcast. This didn’t only lead to increased advertising revenue, it also enabled the podcaster to create audio content without having to record anything.

In the first week of the “Kurz Informiert” podcast doubling its frequency, the overall plays increased by 37% from 92,000 to 126,000.

Microsoft's VALL-E AI can mimic any voice from a short audio sample | Engadget

Microsoft has shown off its latest research in text-to-speech AI with a model called VALL-E that can simulate someone's voice from just a 3-second audio sample..

A note on ethical use: Microsoft has chosen not to open source the code for its AI model VALL-E, citing potential risks in misuse such as spoofing voice identification or impersonating. The company stated that it would follow its "Microsoft AI Principals" in any future development of the model.

The Emerging AI Tool Landscape

This article by the team at Ghost (A CMS like Wordpress) has created a chart of emerging AI tools. It offers a bit of organization to the rapidly changing AI landscape. The areas you will likely want to pay the most attention to are Text, Image, Video, and Speech tools.

So, where should you start with all this? Their suggestion:

Pick one tool and incorporate it into your process for 30 days. Whether that means launching a podcast or YouTube channel with AI content, doubling your output, or adding one AI image to every post — just try a few experiments and see what happens.

Alarmed by A.I. Chatbots, Universities Start Revamping How They Teach - The New York Times

With the rise of the popular new chatbot ChatGPT, colleges are restructuring some courses and taking preventive measures.

My take on this: I think in an educational context A.I. is most challenging. But I'm old enough to remember a time when a calculator in class was not permitted. Or taking notes on a laptop. Or having a mobile phone.

Detecting and punishing the use of A.I. will not stem the tide of progress. Would it not be better to teach incorporation of A.I. in meaningful and creative ways?

A.I. Tools Testing

Platforms that automate SEO savvy show notes creation is what I am most excited about when it comes to Artificial Intelligence. At my company Polymash we specialize on Podcast growth through SEO optimized show notes, which leads lead to discovery and website traffic via Google. I would love it if an A.I. could take this process over.

Here are some new tools we tested and added to the Podcasting Resources Guide this month. I used the same podcast episode to upload and test with each A.I. tool.

Spoiler alert, IMHO: These are a super helpful start — and we are still a long way from "set it and forget it".

Melville | An (A.I. Powered) Podcast Copywriter

Save time and money by letting Melville write everything you need for your podcast episodes.

This is a very interesting A.I. development, and a solution I have been researching for for close to 12 months. The idea is to use A.I. to take your podcast episode transcripts and turn them into usable show notes. This is the closest to "usable" example I have seen, and I'm excited where the Melville team will take the platform next.

Currently it provides multiple-choice "Optimized Episode Summaries" based on AI or human transcripts. It suggests several "Click-Worthy Episode Titles", also identifies "Keywords for Better SEO", and perhaps most impressively generates "Time Stamped Key Points".

Right now the platform is in BETA and you can apply to try it for free. Currently the platform assembles all of the content, but does not generate a single document from the A.I. output. It's pretty easy to do this yourself, but next on their roadmap is a drag and drop interface to allow users to select and assemble the best output into a single show notes document.

Capsho - AI Copywriting Software for Podcasters

Among the A.I. tools we tested for this issue of the Podcasting Strategy Insights newsletter was Capsho. Here are our observations:

Like many other up and coming platforms, this has potential, but is hampered by some current flaws, as I see it:

No "search and replace" feature to replace misspelled host or guest names.
No ability to add SEO savvy H2 headings
A.I. translation seems "middling" when compared to tools like Jasper and ChatGPT

Swell AI - Saving The Best For Last

Out of all the show notes tools, this is by far the most advanced I have tried. And by advanced I mean simple to use with high quality results. I tested several other tools using the same podcast episode.

This one got pretty close to filling my needs, perhaps the quality comes from ChatGPT integration. Important note: ChatGPT is integrated into the interface, and processes the podcast transcript before answering any of your questions. It stays "on-topic" and so I was able to generate a pretty accurate show notes article, complete with sub-headings, paragraphs that made sense and followed the topics we covered on the show.

I also got Instagram, LinkedIn and social media posts just by asking ChatGPT for them. If you are interested in example output, visit our Podcasting Resources Guide entry .

Signing Off For Now...

14 Podcast Conferences to Attend in 2023 | Articles on Podchaser

If you are curious about the conversation around Artificial Intelligence, it’s ethics, tools and implication for podcasting, consider attending one of the many upcoming conferences.

Our own favorite podcasting conference PodFest Expo kicks off today, so I want to give a shout out to the team at Podfest. For example, Larry Roberts will be presenting "Podcast Like A Pro with ChatGPT" at Podfest!!

Podfest team members have been using it every day for one reason or another.

"Give me 20 YouTube video ideas. Write a script for suggestion #5. Now write 6 social media posts promoting that video. These are great, but make them wittier. Number 3 doesn't make sense, can you do another one?"

So check out the agenda and grab your virtual conference tickets here.

So that's it for me this this month!

Please drop me a line with any comments or if you have some tools or processes we should be adding to the Podcasting Resources Guide. I had fun testing A.I. tools and curating this newsletter for you, so thanks for reading and being subscribed! And please don't hesitate to opt-out if this content does not hit your sweet-spot :)

Cheers for now,

Juergen