If you’ve been keeping track, you might have noticed that I stopped updating my blog back in February 2024. Since then, I haven’t posted much. Why? Well, life just got busy, and the blog kind of fell off my radar for a while. But now, as we reach the end of February 2025, an entire year has passed. So, what have I been up to? Quite a bit, actually.
For a while, I was really into AI image contests, and if you were following along, you might remember that in April 2024, I decided to stop making AI images for WireStock. The reason was simple—they wanted my Social Security number for tax purposes, and I’m not comfortable giving that out. Besides, I had already created more than enough images—at last count, well over 50,000, all neatly categorized. I was starting to see myself making the same types of images over again, so it felt like the right time to move on to other projects.
The first thing I picked back up was my VR 360 tours, but this time with a purpose. Back in September 2023, I had joined a community of people who follow Tom Campbell’s work, My Big TOE (MBT for short). His theory suggests that we live in a simulated virtual reality, and I had come across his ideas right around the time I wrote my blog post, Are We Light Beings?. I first heard Tom on the Next Level Soul podcast, and what he was saying really resonated with me, especially since I had been contemplating similar ideas.
One thing Tom mentioned was Tom’s Park, a book he wrote about an imaginary park where people can go during meditation to meet up. That immediately reminded me of the kind of experiences I’d been having with my VR tours, so I thought—why not create a virtual tour to help people who struggle to visualize what’s in the book? The book includes some illustrations and floor plans, but I had met people who found it difficult to picture the spaces in their minds. Over the past year, I’ve been working on this tour in bursts, adding to it whenever inspiration strikes. If you’re curious, here’s a link to check it out.
As for my YouTube channel, I had put it on pause for a while since I wasn’t getting much traffic. That changed when I discovered Notebook LM, an AI-powered language model that turns books, videos, audio, and text into podcasts. That kept me engaged for a while—I ended up making over 50 podcasts, hosted on Podomatic and posted to my YouTube channel. At first, these podcasts were getting a lot of attention, but as AI-generated podcasts started flooding YouTube, the views dropped off. Eventually, I decided to take a break from it, at least until Notebook LM updates to allow more customization, like choosing different hosts or switching voices.
In the meantime, I found another service called Jellypod that lets you create unique podcast voices and even have multiple hosts. It had some great features, but by the time I discovered it, I was feeling a bit burned out from the whole AI podcasting thing. Plus, Jellypod had a paywall, and I have to be mindful of where I invest my resources.
Just as I was wrapping up my podcasting experiment, something new caught my attention. Someone on Consciousness Hub—a platform where MBT followers post messages—shared an AI-generated song about Tom Campbell. It was surprisingly good. I had dabbled in AI music before but had set it aside because, at the time, the quality wasn’t great—songs were short, and the sound had a lot of issues. But this one was different. That got me wondering if AI music had improved, so I decided to give it another try.
And I’m really glad I did. The technology had come a long way. Around the same time, I watched a podcast from another MBT community member who had done a TEDx talk on kindness and another podcast interview on the same subject of kindness. That sparked an idea—what if I created a song about kindness?
I started by downloading the podcast transcript and feeding it into ChatGPT, asking it to generate lyrics that kept the focus on kindness. Of course, AI isn’t perfect, so I tweaked the lyrics to refine them. Then I moved on to the sound and tempo. I asked ChatGPT to craft a detailed prompt describing the song’s vibe—upbeat, singable, with a memorable chorus—so I could use it in Riffusion, the AI music instrument I’m using.
Yes, I called it an AI instrument—that’s how Riffusion describes itself. In Riffusion, I uploaded the lyrics and prompts for the sound, then generated 10-15 different versions of the song until I found the one. It’s a process—listening to each version, making tweaks, regenerating—but I enjoy it. These days, I can easily spend half a day to a full day crafting a new song, and I feel like the results are worth it.
Sure, anyone can go into Riffusion and type a basic prompt like, “Give me a hard-rocking song about an alley cat,” and get something decent. But those songs tend to lack depth, and the lyrics can be hit-or-miss. Instead of using the simple prompt method, I prefer the compose method, where I supply my own lyrics and refine the sound through detailed prompts. That gives me way more control over the final product.
Once I land on the best version of a song, I download the “stems”—the separate tracks for vocals, drums, bass, and other instruments—and import them into Logic Pro. There, I fine-tune everything, enhancing the vocals, deepening the bass, and adding a compressor to improve the overall sound.
At this point, I’ve created over 50 songs, and I’ve been rocking out to all of them. If you’re interested in hearing what I’ve been working on, head over to my music tab or click this link to check them out! I’ve also created a couple of videos to go with the two of my new songs, “The Superhero of Kindness” song and a song about stamping out hate “One Beat One Heart” I also forgot to mention that a lot of my new songs are coming from a lot of the writing I’ve done here on this website as well as various posts I’ve made on other sites. If you go to my Music Tab you’ll see a list of songs and descriptions that are associated.
Riven Waters 