Google's NotebookLM takes on The Bootymachine - How AI understand multimodal art

Using the online-offline multimodal art experience of the Bootymachine from 2007, and using the pdf file of the published book named The Bootymachine ( ISBN 978-2-940679-01-0 ) but also some audio, webserver, photo files and URLs from the original YouTube channel, Google's NotebookLM offers a weird deep dive into this experience from before the iPhone years...

Hope you enjoy! This will lead to further rabbit hole travels in the near future, including with other AI tools, still using this weird but easy to control reference for me.

Don't hesitate to interact! :)

NotebookLM is here https://notebooklm.google.com

The Bootymachine is kind of here (original files, but reconstructing the overall site) https://zzz.ch/bootymachine/2007/

Welcome to this episode of AI’s Hit. Today, we’re testing the possibilities of AI by focusing on Google’s experimental Notebook LM system. This is essentially an AI platform, and I thought, what better way to test it than using the full PDF of a book from 2007, which covers an online project called “The Booty Machine.”

The Booty Machine was a collective photography project that involved creating a platform combining photos, audio, and music. It was designed to explore the new possibilities of mobile phones, the internet, and social sharing in both live and online settings. We even integrated real-time elements during music festivals and posted videos. This was ahead of its time, with the first videos appearing on YouTube in 2007, just after we published content on the Booty Machine platform.

Now, if you haven’t heard of Notebook LM, I recommend checking out the website (link below). What I did was feed the entire PDF into Notebook LM’s document synthesis tool. The PDF is dense—almost 400 pages, with photos and text documenting each day of the year, plus a leap year entry. While the book doesn’t include the music, I’ve added some audio files from 2007 for illustration.

Notebook LM generated a podcast based on the PDF, which I’ll share with you shortly. This is an intriguing experiment to see how AI interprets and synthesizes multimedia content. The project itself was quite abstract, blending text, photos, sound, and video—so this is a deep dive into how AI can process and understand such a mixed format.

In the next episode, I’ll walk you through an experiment using ChatGPT on the same content. So if you’re curious, don’t forget to subscribe and stay tuned for the follow-up episode.

By the way, the original Booty Machine project is still online, though the server has had some issues recently due to an attack. However, it’s also archived on archive.org, so it’s accessible despite these technical challenges. Thanks for listening, and I hope you enjoy the podcast. Feel free to speed it up if it’s too long for your liking!