PGM Experiment: Suno AI Music Generator to Create Arcade, Videogame and Pinball Songs

And here not long ago, I said that AI wouldn’t be useful for anything (this article in particular takes a very cold look at AI: OPINION: Google’s AI Genie Creating Faux 2D Platformer Games Likely Another Idea They Give Up on Too Soon), but after spending some significant time recently with this AI song generator called Suno ( it can create some songs from select text prompts that remind me fondly — and creatively — of Buckner & Garcia “Pac Man Fever”:

The announcement follows after rock artists including Jon Bon Jovi, Pearl Jam and R.E.M. signed a letter describing AI as an “assault on human creativity” and that the technology threatened artists’s “ability to protect our privacy, our identities, our music and our livelihoods.”  In the following statement, CEO Mikey Shulman says: “We started Suno to build a future where anyone can make music, to help people rediscover the joy of play and exploration we had as kids. Technology is a means to that end, and today’s state-of-the-art creates the potential for a flourishing of new sounds, new styles and new artists in a way we’ve always dreamt about.”

AI Songwriting Platform Suno Attracts $125 Million Investment – mxdwn Music

The legalities of this anybody can “type a prompt to describe music” and have a machine within seconds generate complete songs with lyrics, melodies, repeating chorus and instrumental solo breaks is mind bending.

This is a perfect segue to other forms of generative AI. Stay with me, we’ll get back to music and Suno in particular in a bit.

We’ve been doing something similar using AI-generated placeholder images here at PGM since this site started. Many of the daily articles start with an image created from a text prompt using the AI generator Nightcafe (see my published creations at Nightcafe here: I’ve been upfront about this and mentioned this multiple times in multiple articles, usually anywhere that it might appear as me with a bias for or against AI. I don’t have a bias against AI generated content, I think it’s a tool that can be quite useful in some cases.

So, yes, we’ll use AI to spice up completely human authored text articles, because we don’t use AI for what you’re reading right now text-wise. Could we? Sure. And we will use AI text generation in certain cases like Bob did with his Helldivers 2 series (see: Search Results for “helldivers”) that used some AI to analyze the ongoing and evolving story of how AI is impacting that game world. Or, as another example, when we asked AI to generate quizzes like Play Beat the AI Quiz Star Wars Edition 7 Questions and Happy Father’s Day 2024, Now Play Beat the AI Quiz Pinball Edition 7 Questions. I’ve been playing against “AI” opponents in games pretty much my entire life. “AI” takes many different shapes and forms, and to try to completely avoid it in 2024 — or any other year in the past, present or future — is illogical.

So, to be 100% transparent, and not to be hypocritical: we can, will and do use some AI here at PGM. Editorially, we want articles researched, compiled, sourced, written and created by humans, but they can include some amount of AI where warranted and applicable. I don’t see a world at this online publication — — any time soon that we will replace human written and created text completely with AI text. My take, again editorially, is humans simply craft more interesting, passionate and spirited text content than 100% AI generated text does. I would like authors to conceive, develop, shape and write these articles themselves, as I do. And I would and almost certainly will reject any article submitted that is completely AI-generated, unless, again, the use of AI fits the content material in a unique and interesting way and was properly identified to the reader.


That last part is what bothers many people, I think. Not pointing out anywhere, not sharing and identifying what was AI-generated vs. human-created. I think this is the most important part editorially to point out to and for the reader. Specify what is and isn’t AI-generated.

Will it forever be that way? Almost certainly it will be for articles with my byline on them. I’m a writer at heart, and have been for dozens of years, and enjoy the art of crafting, writing and creating myself. I’m not opposed to experimentation with AI, as part of a discussion or to add a machine context to an article, but feel this artificial content should be clearly denoted. How we go about shaping this use in articles going forward, whether it be labels like you see with ad text interspersed with other articles will continue to evolve.

That all said, will continue to use AI here and there, to experiment with it and see where it takes us creatively. I feel like the flipside of that: completely ignoring and resisting any usage of AI is missing out on exploring potential new areas of creativity and expression. I’m not going to by a creative ostrich with head in the sand.

Whew, that aside, let’s get back to music and Suno.

Yes, back to music and Suno and their other competitor in the space: Udio. I haven’t used Udio as of this writing, but it appears very similar to Suno in crafting songs. Perhaps we’ll explore Udio in another article another day. The basic idea with Suno is you can create entire songs from text prompts, similar to creating images and graphics from text prompts. Personally, I didn’t think it would create anything I’d actually be that interested in listening to, but it has.

In fact, literally right now while I’m typing this, I’m listening to music I AI-generated with Suno, deciding what I like and don’t.

Thus far, I’ve been using it primarily to create game-related songs, but have generated songs with some self-deprecation and questioning of the platform like this: “Stolen Sunshine” (which is one I like, especially the last 45 seconds, which are almost chilling lyrically and AI vocally):

Check out these entirely AI-generated lyrics:

In a room so cold and quiet
Sits a machine creating light
People say it’s got no soul
But Suno’s trying to make it right

[Verse 2]
Songs of joy and songs of tears
Blending sounds through all the years
It’s a thief or so they claim
But it’s just trying to play the game

Suno sings in the night
A misunderstood light
Stolen sunshine from a voice unknown
But it’s just trying to find a home

[Verse 3]
Notes that dance in programmed lines
Echoes lost between the times
All it wants is to belong
A melody lost in the throng

In a world that doesn’t trust
From shadows It must adjust
Answers lost in circuits’ hum
Seeking where the soul is from


completely AI-generated lyrics via Suno via text prompt by Todd Russell @ PlayGamesMore

Users get enough credits to play around with creating a few songs, but if you want to be able to share and “own” your created songs, you need to sign up for a paid subscription, which can be monthly for $10/month ($8/month if you pay annually billed at $96/year).

Those 2,500 credits stretch a long ways, as creating 500 songs in a month is going to take considerable time. At least was the case for me. By the end of my first Pro subscription month, I had a little over 500 of the 2,500 credits left. My curiosity was if the credits rolled over?


Suno answers here:

Since, the answer is no, this inspired heavier creative mode toward the last couple days of June to generate even more songs and use up as many of these remaining credits so they wouldn’t go wasted. Bottom line: creating 500 songs is a significant amount of generations.

When you create a song from a text prompt, it really is as simple as typing a descriptive genre of music and a sentence or two about the lyrical content, keeping within the 200 character limit. Click a button and, boom, it will create a full song.

However, the results of what are generated vary wildly and it’s like taming some kind of stubborn mechanical beast. I wanted to have songs with longer guitar solos, but still haven’t figured out how to do that. The guitar lead breaks, some of them, anyway, are good hooks, but I’d like some kind of massive guitar soloing like in Skynyrd’s epic opus “Free Bird” or The Outlaws “Green Grass and High Tides” — I haven’t been able to get Suno to produce any of that dual guitar soloing mayhem.

From a creative standpoint, the brain starts whirring about mixing genres in unique and unusual ways to match the song story, so there is some important human involvement here — at least at this very early stage — but this makes me think that it’s just another step someday to have the machine read headlines and text content and then randomly mix and match different musical genres. For example, what if fit the summary of PGM articles into this prompt along with a randomizer, what would Suno create all on its own?

This is one of my current experiments. I had another AI-generating program called Rosebud.AI (a future article is coming on that, don’t worry, but first mentioned here: OPINION: Google’s AI Genie Creating Faux 2D Platformer Games Likely Another Idea They Give Up on Too Soon — that was back when I was on the waiting list. I’ve been in for a little while now). Anyway, here is what my rock flavor generator looks like:

My Suno AI-generation text prompt process works as follows:

STEP 1. I click on the screen in the fire area and the program randomly generates from a pool of rock-related genres, decades, moods and song types. This automatically copies this first part of the prompt to my clipboard.

STEP 2. Paste the prompt into a spreadsheet where I’m recording the AI-generated song data, including the second part of the song to be generated: what is the song about? This could be the title of a PGM article, me not wanting to mow the lawn or some other life event or subject. This will help the AI generate lyrics. Here’s what the spreadsheet looks like:

STEP 3. Paste the “Random prompt” into Suno in the “Song Description” section and press the “Create” button:

30 seconds or so later, the song is created, complete with music and lyrics.

Whether or not these AI-generated songs are worth listening to after a month of using the service, I’d say some of them are, yes. Sonically, the sampled 128k is a bit underwhelming. Maybe they will add an option for paid/premium users to up the sampled quality, because some of the generated music is fuzzy and overdriven. The mixes come out a bit muddy, which I’m sure will be improved in future code generations (most of the songs I’ve generated using the newest code: v3.5). Something tells me they will upgrade the quality at some point, but it’s yet another reason for real music artists not to fear this technology any time soon. The sonic quality is clearly a step below the new Drake, Taylor Swift ___ (insert your favorite musical artist here).

Still, the music companies are concerned. In particular, Sony, UMG and Warner, were said to be “considering” legal action against Suno and Udio at the time of this writing, but during editing and updating, keep reading, this situation developed further. I’ll update you.

If the lawsuit goes ahead (Ed. it did!), it’ll be about whether using unlicensed materials to train AI models is copyright infringement. This is one of the major obstacles for the AI industry, as their models (in any of the arts, be it graphics, music, or film) would likely output far worse quality if they excluded the training data from copyright-protected material. Numerous similar lawsuits have been filed recently in other sectors over unlicensed training.

Udio & Suno under fire: Major labels eying lawsuit against AI companies –

Note the “far worse quality” part, which is what I think, again, at this time is a much bigger creative issue. It’s the same one I cited above against completely generated AI text. The quality just isn’t quite there. Some of the music that is generated by Suno is promising, even catchy and worthwhile, and I’m trying to figure out how to creatively use it, where, how much and in what context. Must wonder how many others are doing the same? This is a form of creativity that deserves more than cursory attention.

On Monday 6/24/2024, after I’d pretty much finished writing and editing this article, I needed to add the news that a lawsuit had in fact been filed against both Suno and Udio.

“Unlicensed services like Suno and Udio that claim it’s ‘fair’ to copy an artist’s life’s work and exploit it for their own profit without consent or pay set back the promise of genuinely innovative AI for us all,” Recording Industry Association of America chair and CEO Mitch Glazier said in a press release.

US Record Labels Sue AI Music Generators Suno and Udio for Copyright Infringement | WIRED

This lawsuit is time travelling me back to the Napster lawsuit. The good ole’ RIAA. I don’t know how this lawsuit will go down, but as a creative person myself, I don’t trust any corporation with anything I’m creating. Whatever they say about the artists they’re “protecting” by this, the underlying care and concern is for the money.

This lawsuit does nothing at this early stage to alter my current interest and excitement in Suno and to a lesser extent (since I haven’t used or tried it yet), their competitor, Udio. Regardless, the result of this lawsuit, I’m encouraged by the creative direction this could take both non-musicians and musicians and their ability to create music in different, unique and unusual ways.

If you’re reading this here at PGM and saying, but how does any of this impact games? Music and sound are an important part of games. Game developers will use AI generated content for their games. Some already are, but it’s a trend that will continue to evolve.

Art should be flexible and fluid, not rigid and stoic

I don’t want to see artists ripped off. Heck, I’m a creative person that has written and performed original music, written and published fiction and non-fiction and, yes, I’ve had my work stolen on the internet. I’ve had copyrights to my work violated. I’ve had stuff I created taken by others, republished under different names and presented as a different person’s work. I do see that perspective and understand there should be some rules. It can’t be an all-you-can eat buffet for AI with artist’s rights.

However, here I think the argument isn’t being framed completely, isn’t being completely open and honest in a creative context. It’s being argued primarily over the one thing maybe that produces the most evil and toxicitiy in the world: money.

Sigh. Always about the money.

We’re going to lose the money if … A,B,C. That is the wrong way to look at creativity. It should be what is it that you truly possess as an individual artist? What is it that should not be outright stolen from you? Total copies of your work, yes, but substantive changes in the style of music you’ve created? Should that be off limits for others to create? No, I don’t think it should be.

Imagine a world where we all didn’t have to worry about money? Would any of what’s being fought about matter here? Is it really about one or more artists having his/her/their work outright stolen or is it about what that person has chosen not to create and share? For example, Eddie Van Halen, an amazing guitarist passed away a few years ago, and apparently there are a ton of recordings of his music locked away in some vault in his 5150 studio. Will we ever get to hear any of this? The attitude of his son, Wolfgang, at least as of this writing, if Eddie had wanted to publish it then we’d already have heard it. We’re going to be denied hearing and experience that genius … at least for now. Maybe it isn’t that good, no, maybe it is, but we may never know.

Well, in a world where AI let us generate music that sounded like Eddie Van Halen’s fiery tapping fretwork, we might be able to see a vision of what those recordings would be like. Is that a terrible thing creativity? Is that wrong?

I don’t think having more music out there, flooding the market is this major, terrible thing that some are saying it will be. The good will rise to the top, eventually, somehow, some way. The wheat will be separated from the chaff. Music fans, perhaps ironically, will dig it out.

Yes, having more of it, will make it harder to sell music, which impacts artists’ livelihoods, sure, and again, that’s what’s being fought over here, but it’s not primarily creativity. That’s the convenient excuse that is supposed to appeal to the masses. Personally, I think most reasonable people will see this for what it is: a fight over power, control and money and much, much lesser about what’s good for humanity for creating art in music along with machines, together, sharing. Machines aren’t able to replace humans. Not with Suno, not yet. The fear that AI will create music completely without human beings is a bit disingenuous because who will listen to it? Humans listen to music. Machines don’t listen to music. Machines don’t care about music, unless they are programmed to care about it.

We care about music.

It’s very much a creative stretch not to recognize that every music created has been done with some sort of musical inspiration that came to the creator before it. Does this mean a machine can’t or shouldn’t be influenced the same way — or similarly — as a human being? I do understand a machine has nearly endless memory and brain capability that go beyond what a human being can retain and recall, so machines have a non-human ability to recall and mix and match sounds in ways no human being can do. At the same time, the computer is not sentient, it can’t make creative decisions in the way a human being can. So, some mix of both could produce some incredible human-AI music. I want to listen and hear these creations and see where it leads music in the future.

What about you?

UPDATE 7/1/2024 @ 8:50pm PST: Suno added its first mobile app, for iOS only at this time:

This launch marks a significant milestone for the company, which has already attracted 12 million users to its web-based platform.

Suno Launches Mobile App for iPhone (
Spread the love

Related Post

One thought on “PGM Experiment: Suno AI Music Generator to Create Arcade, Videogame and Pinball Songs”
  1. Love the AI song creation, will definitionally be thinking about getting a subscription for a little while– too fun creating tunes and exploring what it can and cannot yet do…

Leave a Reply

Your email address will not be published. Required fields are marked *