Listen to Some of the First Single-Text Prompt AI-Generated Songs

Let’s face it, we’ve all known this day was coming. As soon as generative AI made its way into images and video, we knew that it was going to make its way into music as well. The only question that remains is how soon will AI music be good enough to replace (or use in tandem) traditional human-made music.

Well, ElevenLabs, an AI software company founded by ex-Google and ex-Palantir engineers and strategists appears to be now leading the cause to provide AI-generated music for all of us to enjoy and use in our various creative projects.

Let’s take a look at some of the first AI-generated songs that were generated from single text prompts with no edits and explore how they might be scarily useful for your film and video projects.

---

A Quick Look at ElevenLabs

We’ve actually covered ElevenLabs in the past as the AI software company focuses on developing all types of natural-sounding speech synthesis and text-to-speech software with artificial intelligence and deep learning.

ElevenLabs has already made great strides in providing lip-syncing features for your AI-generated characters that could make AI-powered animation projects more of a thing. We also covered how ElevenLabs was one of the first AI audio players on the scene to help add AI sound effects to the AI videos demonstrated by OpenAI’s Sora.

But these new AI-generated songs, which only require a single line of text to prompt, are going to perhaps be the company’s biggest splash.

ElevenLabs AI-Generated Music

— (@)

Shared on their Twitter (aka X) account, ElevenLabs has provided four completely AI-generated songs for us to listen through and review. ElevenLabs Music touts that these songs were generated from a single text prompt and include no edits, and even share with us the prompts they used to get the music.

You can check each one of them out here below.

— (@)

— (@)

— (@)

Questions and Concerns

Now, there’s a lot to unpack here.

We could discuss the quality of the music itself, which does sound quite… good? The songs do sound like they were professionally recorded and are honestly able to tap into the genericness of what you might find on Top 40 radio. Which, for those who are a bit afraid of AI technology replacing artists, should be quite concerning indeed.

We also don’t know a lot about how ElevenLabs has sourced and trained its AI to create this music. Were these just the first four songs they put into their AI model? Or are we getting the best picks out of hundreds or thousands?

Also, the biggest question that should be asked for any AI technology going forward is where did this AI get the sample music and tracks needed to “train” on? What music library (or libraries) provided their fully licensed and covered music for this AI?

Until we get a lot more of these questions answered, we’re left with the rest of you out there, simply feeling a bit in awe as well as a bit scared about how quickly these AI models are developing—and what they could mean for the future of music, film, and art itself.