I like exploring new technology products, even though I am not a techie or do not even consider myself to be technologically inclined for that matter. But I love to explore and dabble with new tools, apps, and products, some of which I have started writing about on my blog.
A few days ago, there was a livid discussion in a facebook group for content marketers about the rise of content creation tools that use Artificial Intelligence technologies. Terms like GPT2 and GPT3 were thrown around, and the more vocal members debated the pros and cons of using such tools. Typically, discussions such as these often give rise to some interesting ideas.
In this case, the thought crossed my mind, “Where is Podcasting headed in an era of AI driven content tools?”
Podbot on ProductHunt
I recalled reading about (and dabbling with) podbot.ai, a podcast creation tool that uses AI. Speficially, it is based on GPT2. I came across podbot on ProductHunt, and thought it was a nice idea. Over the past few months, life and other priorities made me forget about it, but the discussion on facebook group reminded me of this site. If the product was active development, then the tool should have become better. Atleast that was my hypothesis.
On Podbot, my my test podcast is called “Seasons”. It has four episodes published so far, the recent two in or around February 2021.
The first episode is called “Monsoon Rains”, the most recent one is titled “Blogging away endlessly without a Content Calendar- What are the Pros and Cons?”Compared to the first episode, published in July 2020, the most recent one from February 2021 sounded rather accurate.
How to create a podcast using Podbot.ai
Setting up a podcast using podbot, an AI driven podcast creation tool, is pretty easy.
- Once you have set up your account and reached the admin dashbord, you first create one or podcast. I mean the series, not the episode itself that will follow shortly.
- Next, you create an episode. To do so, enter the topic that you like, and the program takes care of the rest.
- The AI tool will generate the text from news or popular blogs or others sources related to the query you have entered.
- It will then fetch images that are relevant or related to the post. This could be from Unsplash or similar sites. That takes care of artwork.
- Finally, using Google’s text to speech technology, the actual audio is created.
First, to check if something had changed in terms of product development user interface. Even the technology itself that bot uses the GPT2 tool for artificial intelligence, while GPT three has been released recently. Next, I wanted to see if the overall podcast listening experience had changed.
The post I created this morning on using a content calendar, seemed quite accurate, and pretty pleasant to my ears.
Do these podcast episodes have an audience?
I was quite surprised to find that the podcast episodes created using podbot actually have an audience! My first three episodes have received more than 100 listens as on date. This is not a perfect technology, but I think this idea holds potential.
Friend, or Foe?
Here comes the important question. As a podcaster do I see this tool as a threat?
Such tools could be an enabler for testing out new show ideas. So binge watching streaming shows happens to be an idea that I’m actually exploring.
Sometimes you may be in a bind and you want to get that episode out there. Using a technology like podbot, you can create and publish the episode, with a disclaimer about how the episode was published. In time, you should then go ahead and replace it with an actual narration, or updated as and when time permits.
Note: this option should only be used in the worse case scenario (treat this as the Don’t Panic! button)
The (Current) Downsides
The narration can sound quite impersonal. Secondly, the advertisement for podbot at the intro is pretty long and rather boring. Next, a recent episode that I created binge watching on Amazon, was actually pretty accurate, only for the first half. The second half was quite literally a copy paste job of the first part. The proverbial Glitch-In-The-Matrix Deja Vu. There are several other oops’es, many of which are highlighted in the reviews on ProductHunt.
Wrapping it up
It is the surprise element that thrills me and intrigues me more than anything else. There may be similar, or better tools out there- I am yet to explore. For now, I am pretty impressed with the idea and the potential it holds.
Resources:
1. The “under the hood” technology of podbothttps://podbot.ai/about
2. FAQ page: https://podbot.ai/faq
