Who is John Galt Trigeron? Trigeron is an AI DJ with a modest and (hopefully after this article) steadily growing listener base. The name is a serendipitous misspelling of Tregaron, a small town in Ceredigion, Wales. It’s a name my parents considered for me at birth, so it felt fitting for my DJ alias.
Since May, I’ve been diving into the world of AI music and today I released six songs. My motivation is twofold: to gauge AI’s artistic skills and to refine my musical taste. Many critics say house music sounds repetitive, and this activity/procrastination technique/leisure activity has helped me appreciate the nuances. Plus, creating something I’d want to listen to is inherently satisfying. Admittedly, most of my output sounds like the generic house tracks I tend to avoid, but the near-zero production cost makes that fact a non-issue.
I use Suno.com to craft my tracks. Thanks to a recent $125 million funding round and a pace of innovation that’s almost dizzying, the platform keeps getting better. In just the past month, the Suno team has launched a new app, revamped the UI, and introduced various competitions (none of which I’ve entered).
However, the lyrics are often lacking, and positioning words and beats precisely isn’t possible.
My workflow involves several tools and platforms. I subscribe to Suno’s Pro Plan for $10 a month and use LANDR’s $20 plan for AI-powered mastering, where I then distribute the tracks I like to Spotify. Creating music with AI feels like a modern alchemy—mixing random words and seeing what the algorithm concocts. Terms like sibilance, arpeggiator, filter sweep, and mid-range harmonics have become part of my Suno lexicon.
I maintain a working document with 120-character prompts that I input into Suno’s custom prompt box, extending them into 4-5 minute tracks. These are then sent to LANDR for mastering and release, a process that takes 2 weeks to a month to appear on Spotify. For visual elements, I use Leonardo.ai, sticking with a space theme inspired by the idea of robots from outer space—my initial creative muse. Occasionally, I use the open-source tool Audacity for minor edits, though I’d like to become more sophisticated with my Audacity usage.
I’ve discovered a surprising affinity for tech house drops, despite generally finding the genre difficult to enjoy. This process of creation has been like an accelerated music humanities class as I can tell the difference of different instrumentation, rhthyms, or beats, since I’m the “creator” (AI is really the creator).
While Lyrebird and Auphonic excel in podcast and voiceover production, few tools specialize in music generation with lyrics. Suno leads this niche market. Loudly offers versatility but lacks lyric support, and Voicemod’s 30-second limit and lyric constraints hinder its usefulness for music creation. Suno provides the most customization options, though its ‘extend’ feature sometimes generates off-target, random lyrics. Despite this flaw, Suno remains my top choice and the clear market leader for AI-driven music and lyric generation.
One clear casualty of this AI music boom is the freelance artist that I paid $50 once to record a song for Mother’s Day, which now can be easily done on Suno.
Exploring these tools and discovering what sounds good takes time and patience. My songs are passable (according to some), but fine-tuning remains a challenge. I want to improve at breaking up parts and incorporating varied beats effectively.
Despite challenges, platforms like Suno make AI music creation exciting with their innovation and ease of use. Low production costs and rapid advancements in AI tools makes this a fun hobby I’ll continue. Despite being completely reliant on the AI tools to make the music, I still believe that creativity transcends the tools we use.
unpublished alpha 