This AI Development is Alarming
The Day Has Come.
You can generate music using a text prompt.
Here, let me show you:
I made this song in about 30 seconds (check the description for my prompts).
Literally.
Here’s version 2 for the really curious.
Advancements in AI are continuing to accelerate at an alarming rate.
For the last two weeks, I have spent more than 30 hours experimenting, researching, and freaking out over 2 AI tools that directly affect me and MY wallet:
Suno and Udio
Musicians, we are out of time, and AI technology will only get better from here.
In today’s newsletter, I will show you how the cutting edge advancement of AI technology is interacting with the music industry. I will explore these questions:
What are these tools, and who is this technology for?
Should we learn this tech?
Are we already too late?
Now what?
What are these tools?
Suno
Story
VC-backed company founded in late 2023
All 4 founders worked at Kensho, a tech AI start-up for financial data.
How it works:
2 Basic modes
Basic
Type a text prompt describing that song (subjects, instruments, feelings, and or custom lyrics)
You can chose to make it an instrumental instead of generating lyrics
Custom
Write a prompt for Lyrics
Write a prompt for Style of Music
Delivers 2 examples based on your prompts—up to 2 minutes of music each
Udio
Story
Founded in late 2023 by former Google Deepmind Researchers
Also VC backed
Only recently emerged from stealth mode a couple of weeks ago.
Seems to understand more niche genres
Includes side chaining, tape effects, reverb, and delay
Also delivers 2 examples
Who Is This For?
People who wish they could make music, but don’t know how.
People/organizations looking to drastically reduce their music licensing or marketing budget.
Producers/songwriters who need to create a lot of music in a short period of time.
Trolls who wanna make goofy parodies.
Who Is This Not For?
The faint at heart.
Middle-class musicians. This will directly compete with us in many markets.
People who care about tradition.
Should We Learn This Tech?
Definitely.
I’m an advocate for developing powerful skills.
Let’s say you are sitting on an indie movie that you have been desperately trying to finish.
You raised some capital, but the budget got away from you during production.
Now, you are stuck in the post-production phase. You only have enough money to pay for editing OR music.
Today, you don’t even have to make that choice. You could pay the editors and put together significant amounts of completed music. All for the low, low cost of:
$0
Just write a text prompt synopsis of what you are looking for, and you’re off to the races.
That’s wild.
Will the product be good? Maybe. Prompting methodology is still being definted. Users are experimenting, sharing, and discussing their experiences with these AI apps in their respective Discord channels.
Are We Already Too Late?
No. I think we’re still early. This is an opportunity to learn more!
Join Suno and Udio’s Discord Channels and talk to the community about the apps. These are the early-adopters who are in the trenches. See if you can save time by absorbing their wisdom.
Here are a bunch of videos I watched to research for this.
The AI Music Situation is Insane (RoomieOfficial YT)
Did AI Just End Music? (Now it’s Personal) ft. Rick Beato (Cold Fusion YT)
Why AI is Doomed to Fail the Music Turing Test (Adam Neely YT)
AI Music is Good - So What Do We Do? (Digging The Greats YT)
Here is a resource you can use to get better at prompting on Suno: Suno Prompting
Now What?
History is rhyming once again. The advent of the synthesizer, the drum machine, and the digital audio workstation (DAW) all created tectonic shifts in the process of music creation during their time. While each technological innovation was different from the last, the grim rhetoric around their invention is eerily similar to discourse today.
This new creation process will yield a few interesting outcomes:
More people will be able to create good-sounding music.
This technology will supercharge the possibilities for sampling.
The sync industry is about to get flooded.
Copyright court is about to explode.
Artists will be forced to adapt.
Basic music products like songs will plummet in value. Maybe even to zero.
Takeaways
“With great power comes great responsibility.” - Uncle Ben
Today, humans can use tools like Suno and Udio to create entire pieces of music—using only text prompts. That’s fucking trippy.
This feels like Spotify’s Napster moment.
YouTuber and Professional Musician Rick Beato might agree with me. In a recent Cold Fusion video, the host, Dagogo asked Rick if he thinks that one day we will have an AI-generated song hit #1 on the Billboard charts. In response, Rick chortled,
“Yes, probably two years from now. There’ll be a lot of new stories about it and people will say, ‘boy I really like this.’ And then they will just create more and then there will be more. I can see a time when 9 out of 10 Top 10 songs are AI-generated.”
Well…shit.
Now, going from 0 to a complete musical idea is effortless, quick, and FREE.
I do no believe that tools like these will replace live performance. According to Rick Beato “People have too much fun playing real instruments. I’m not gonna stop playing just because there’s AI-guitar things.”
I couldn’t agree more. Even if there were AI-viola tools, I would never want to use them. Because I spent so many years patiently struggling to get better at viola, playing it gives my life meaning and I draw the deepest satisfaction from continuing my passion.
When I started researching this topic, I was terrified. “This is where the robots take my job” was the inner dialogue keeping me up at night.
I’m still a bit tripped out, but after doing more research on the matter, I am actually feeling a bit more inspired.
I have no control over whether AI is here or not. All I have control over is what I do in response. I choose to grow and evolve.
I wish the same for you.
Stopping by in the middle of reading this to say: those who can’t, those without the willingness to use their pure, raw talent, will use AI…