AI Generated Librarian as a machine editing a podcast.

This is going to be something that all of you will find “interesting,” but maybe not something that you will like. Last week on 3 Geeks, I posted a blog that talked about how to use AI to generate summaries of legal articles. This week, I wanted to expand on that project a little and see if I could turn the summaries into a podcast. The goal was to try to get it completely automated, and completely AI generated. Well… as you can see from the title of this episode, it was almost completely automated, and AI generated. But not 100%.

Here’s the process I created, and I attached the mp3 of the output.
  • RSS Feed that tracks new BigLaw Podcast Episodes.
  • Use a Python script to pull the episode information.
  • Use GPT to create a description of the episode.
  • Use Descript to translate the text summaries into voice output. (I did lightly edit these with an intro and outro as well as tweak the transitions between each review.)
  • Use Soundraw to create an intro/outro music.
  • Combine in Audacity.
  • Output in mp3.
All of these tools are actually free, except for GPT, which is about $0.01 or less per article.
This is far from perfect, but it is kind of cool, and I think there are some uses for these tools.
Whether you love this, hate this, or don’t really care, I’d like to hear what you think!!
(The AI Generated part starts at the 6:50 mark.)

Listen on mobile platforms:  Apple Podcasts LogoApple Podcasts |  Spotify LogoSpotify
Contact Us:
Twitter: @gebauerm, or @glambert
Voicemail: 713-487-7821
Email: geekinreviewpodcast@gmail.com
Music: 
Jerry David DeCicca
Transcript

Greg Lambert 0:07
Welcome to a special episode of The Geek in Review, the podcast focused on innovative and creative ideas in the legal industry. I’m Greg Lambert. So I have a just a special episode that I wanted to play for you today, which is a project I’ve been working on. And I think that a lot of you will be interested in this. And some of you may not like it. But I think it’s something that we wanted to test out and see what the audience thought.

It’s not something that we’re going to be doing with The Geek in Review. But I think it’s just something that with all of the talk about AI, chat, GPT and other tools that are out there, I thought I would expand upon an experiment that I did last week and posted on The Geek in Review about which was to use AI to summarize content.

So this week, what I decided to do was to look at a news feed that I had set up that monitors law firm podcasts. So that’s a interest of mine, I want to know what’s going on with other law firms and what kind of podcasts they’re putting out. So at a minimum, I wanted to kind of just keep up to date on that to know what the topics were, who was putting them out any changes any new additions that were going on? And that’s really all I do with it, it allows me to keep up with what’s going on in the industry.

But as with many things, I started thinking, wouldn’t it be nice if I could just listen to these news reports or these news feeds, rather than having to go on to the computer every morning and then read them. Some folks might like one or the other. But I thought it’d be kind of interesting to see if I could make that happen. And of course, like many things, I don’t have the time to do that. So I looked at the artificial intelligence tools that are out there and the automation tools. And I decided I was gonna give it a go to see if I could create a almost completely AI generated podcast.

And so here’s what I did.

Over this weekend, I worked with a number of tools to see if I could create that output. So here’s the process that I used. And I’m going to actually play the final output that I created. First of all, I have an RSS feed that tracks the big law podcast episodes.

And I then took that feed and fed it through a Python script to pull the episode information.

And then after that, I use the GPT, the actual pay version of GPT through an API to create a description of the episode.

And then the next step was to use a program called descript to translate the text summaries that the GPT tool had created, and put that into a voice output.

Now there is one part and this is why I call it the you know, almost completely AI generated podcast. And that is I did do some light editing on the text summaries so that there was a transition between the articles or the podcast, and I created an intro and an outro. And I kind of tweaked a little bit of the spacing and the timing of the output, but really not a lot of work. On my part.

I used a tool called Soundraw to create an intro and outro for the music.

And then I combined that into Audacity where I did the editing and create an output in mp3. The platform we use is anchor.

And so pretty much everything that I’ve used is a free tool, except for the GPT which is probably about a penny or so per per episode or article that I had a process. I can tell you I’ve been working with it and given and given the paid version of GPT for the last 12 days. And my total bill so far is less than $6. So it’s a really affordable tool for what you get.

There’s actually a much better voice AI tool out there called play h t that I took a look at, although it’s much more realistic, and I would love to have it. It’s not free so I did not use it for this experiment. Now don’t get me wrong. This is not by any stretch of the imagination, a perfect output. But I think it shows the potential that’s out there for doing things like this.

Now I saw a tweet this morning from a really good friend of mine Cat Moon out of Vanderbilt, who talked about how she thought it was interesting how there’s a number of people that look at tools like this and immediately think this is going to end my career there, the the AI tools are going to take over, I think more, we need to get away from the zero sum game, when it comes to technology, that technology is going to be a one for one replacement of human labor, I’m going to tell you right now, I would not do this.

As much as I would enjoy listening to a summary of big law podcast episodes, I’m not gonna do it. I don’t know of a lot of people that have that kind of free time to go through and do that on their own. However, it still is a useful idea, I think, to understand what’s going on in the industry. So instead of looking at it as a replacement for humans, I look at it as an augmentation that I think this these tools are really useful when it comes to providing information in a way that doesn’t take a lot of human time. gives us that time back, but it gives us some really good information that’s available if and when we need it.

So without going into depth any further on on the process. I’m going to let you listen to approximately the next five minutes of an automated well, almost completely automated, almost completely AI generated podcast

AI-Ethan 6:50
Welcome to the Big law podcast update, where we sum up the latest podcasts from am law 102 100 firms.

AI-Emily 6:56
First step we have fastest five minutes, SBIR and contractor minimum wage from Crowell and moring Crowell and moring fastest five minutes is a biweekly podcast that provides a brief summary of significant government contracts legal and regulatory developments posted by Peter Ireland Yuendumu. This week’s episode covers agency labor advisors, data rights, minimum wage for contractors and the strengthening VA Cybersecurity Act of 2020. To tune in to stay up to date on the latest government contracts legal and regulatory developments and access the podcast from pod bean SoundCloud or iTunes

AI-Ethan 7:30
from Wilson Elser, the crazy world of crypto register now February 9 webinar. Join Anjali das partner Chicago, Illinois and co chair of Wilson else’s National Cybersecurity and data privacy practice for her webinar, the crazy world of crypto on February 9 2023. Learn about the crypto ecosystem, the fall of FTX us regulation of the crypto markets and potential silent crypto exposure under traditional insurance policy. CLE credit is available in California, Connecticut, Illinois, New Jersey and New York.

AI-Emily 8:04
Our third summary is Baker Hostetler his podcast. Jarvis Gresham of the American Bar Association litigation section interviews Judge Keith and be frank of the 17th Judicial Circuit of Florida for the sound advice podcast. In the episode Judge Frank provides tips for young lawyers on oral advocacy tune in to hear his advice and gain insight into the legal profession.

AI-Ethan 8:26
More from Baker Hostetler on FINRA signals expanded enforcement priorities and annual report. FINRA is 2023 report on its examination and risk monitoring program highlights the growing risks posed by cybercrime fraud and other bad actors and urges firms to strengthen their internal compliance and reporting structures to combat these threats. The report also covers Financial Crimes manipulative, trading AML fraud and sanctioned and emphasizes the need for firms to evaluate their risk management programs to ensure they are adequately monitoring and reporting on these complex forms of fraud.

AI-Emily 9:00
Ballard Spahr is up next with their mortgage banking update. The CFPB has proposed a registry for non bank supervised by the CFPB that would require reporting and public disclosure of the use of arbitration provisions, waivers and other limits and form contracts. This proposal is a backdoor approach to eliminating arbitration as consumer advocates have been lobbying CFPB director chopper to ban the use of arbitration provisions since he was sworn in. The proposal also covers other contract provisions that the CFPB director dislikes comments on the proposal must be filed by March 13 2023.

AI-Ethan 9:34
Mayor Brown put out its 2023 trends for technology transactions. In this episode of Mayor Browns Tech Talks podcast partners from the technology and IP transactions practice along with host Julian DeVoe discuss eight key trends to look out for in 2023 as companies look to reduce spend and drive revenue topics include cost saving deals digital transformation, leveraging technologies such as AI and open source software and more.

AI-Emily 10:02
Jackson Walker’s podcast talks about value based care and the changing healthcare landscape. In this episode of future ready business, Dr. Lena Anna discusses the shift from fee for service to value based care in the healthcare industry and how his podcast doc to the future is helping to educate rural physicians on the topic. With the help of Glenn Edwards a radio personality Dr. Anna has released three episodes and an intro with plans to release a fourth soon.

AI-Ethan 10:28
Bonds gonna Ken King Latest Episode on Countdown to data privacy day 2023. on cybersecurity and data privacy practices, counting down to world data privacy day with information memoranda, articles, webinars and podcasts to help legal technology professionals stay up to date on data privacy laws and best practices. Contact bonds attorneys for more information on compliance efforts and assessing your organization’s privacy and data protection practices.

AI-Emily 10:56
Our final review today is Willkie Farr and Gallagher discusses US sanctions and export controls 2022 year in review, economic tools take center stage in response to global conflicts in 2022. The US government US sanctions and export controls is its primary tool to respond to global conflicts, particularly Russia’s invasion of Ukraine, the Biden administration and its allies impose novel and sweeping sanctions and export controls to curb the Russian government’s ability to fund its war effort and to prevent the Russian Armed Forces from accessing essential items for combat and maintenance. The spotlight also turned to export controls to limit China’s access to us origin critical technologies, Cryptocurrencies entrance into the mainstream also resulted in a greater risk of misuse prompting a whole of government review. Thanks for tuning into today’s big law podcast update.

AI-Ethan 11:44
We’ll be back tomorrow with more reviews to keep you on top of what issues are happening in the large law world.

Greg Lambert 11:53
So like I said, at the beginning, this is not anywhere near a perfect system. It’s not a completely automated process. But I think it’s something that could be made better. I think it’s something that can be made much more automated. And I think that, you know, this is just scratching the very surface of what these tools can do. And again, I’m looking at this as a way of augmenting the abilities that we have, either as librarians, law professionals, lawyers, there are ways to leverage tools like these to give you the ability to do things that you simply just don’t have the time to do. And the ROI of that time just isn’t worth it for a human to do this. But it’s really interesting, and I hope you got a chance to listen to it.

Whether you loved it hated it, hated me for doing it. I would love to hear more from you. So please reach out to me, you can leave comments here. Or you can reach out to me on Twitter at Lambert to @G L A M B E R T, I would like to hear from you and see what you think about it. So thanks for taking the time to listen to The Geek in Review podcast. If you enjoy the show, or at least find it interesting, please go ahead and share it with a colleague. We’d love to hear from you so you can reach out to us on social media. Marlene can be reached at @gebauerm on Twitter And I can be reached @glambert on Twitter. Or you can leave us an old fashioned voicemail on our Geek in Review Hotline at 713-487-7821.

The music that was not artificially generated and is so much better is brought to you by Jerry David DeCicca .Thanks, Jerry. All right everyone. Thanks for listening.