Who Needs a Ghostwriter When You Have AI? 🤭

What’s better than one Drake? Two. Especially if they’re from the same country as you (word to Vancouver’s finest, Chad 🇨🇦). Btw, if you read that to the same tune of “Family Feud” by Hov, you’re a real one. 

Now you may be like what are they talking about? Two Drakes? Can’t be. 

Well… believe it. We’ll call this era Real Drake vs. AI Drake. ICYMI, over the past few weeks AI generated songs that sound just like the artists they’re designed to imitate (from beat selection to tone to cadence to overall vibe) have taken Twitter and TikTok by storm. Check out an example of AI Drake below:

C’mon… The sample? The word play? The flow? Respectfully, we need Drake to hear this, recreate it, and release it ASAP because it’s giving Summer 2023 anthem. 

In this post, we’re going to break down how it’s possible for AI to create music like this and what this could mean for the music industry. Let’s dive in!

How can these hits be produced using AI?

Back in November, we wrote about how Generative AI is taking the creative industries by storm. And in the biggest plot twist of the year, we’re seeing this come to life with AI songs inspired by some of the music industry’s faves. As we covered in the previous post, music generated by AI is produced by using algorithms and computer software that can analyze existing music and also generate original songs. Here’s a six step breakdown, courtesy of ChatGPT (look at how all of these tech advances are coming together!), that shows how these programs create quality music:

1. Data Input: AI algorithms are fed with large amounts of existing music data, such as melodies, rhythms, chords, and lyrics. This data serves as the training material for the AI model.

1. Pattern Recognition: The AI model analyzes the input data to identify patterns, trends, and relationships within the music. It learns to recognize various musical elements, such as melodies, harmonies, and rhythms, and understands their relationships.

1. Composition: Once trained, the AI model uses its learned knowledge to compose new music. It can generate melodies, harmonies, rhythms, and lyrics based on the patterns it has recognized in the input data. The AI may also be programmed with specific music styles or genres, allowing it to create music in a particular style.

1. Iterative Process: The generated music is evaluated by human composers or music experts who provide feedback, which is then used to further train and refine the AI model. This iterative process helps improve the quality and creativity of the generated music over time.

1. Output: The final step is the output of the AI-generated music, which can be in the form of sheet music, MIDI files, or audio recordings that can be played by virtual instruments or real musicians.

TLDR: if you were to put an artist’s full collection in, the AI model would analyze every aspect of the music and train itself on how to generate music that sounds similar to the inputs given. Then an actual person would go in and give feedback to the model to further improve its outputs. We don’t know whether to be excited about the possibilities of this or scared for what this means for the current state of artists… Either way, it’s something interesting to follow!

What does this mean for the music industry?

So it’s fair to think that artists like Drake might be feeling a little nervous. Someone stepped on “his” track with “his” voice and went more viral than his Toosie slide. And they did it FAST. Here’s a few reasons why AI-generated music thing works so well:

• Creativity: AI-generated music can push the boundaries of traditional music genres and create new and innovative sounds

• Efficiency & Democratizing: AI can produce music quickly and at a lower cost than traditional music production, which can make it more accessible to independent musicians and emerging artists. This has the potential to democratize the music industry and open up new opportunities for aspiring musicians.

• Diversity: AI-generated music can draw from a wide range of cultural and musical influences, which can help to promote diversity in the music industry.

But at the same time, Drake (and his ghost writers) should find solace in the fact that AI-generated music will never be one thing: human. There’s limitations to what AI can do, so Drake can breathe easy knowing it probably won’t be coming for him anytime soon. Here’s a few more reasons why:

• Legal and copyright issues: it can be difficult to determine who owns the rights to the music, and there may be legal challenges related to the originality of the music produced by AI.

• Lack of emotional depth: while AI can produce technically impressive music, AI just isn’t human, so it might lack the emotional depth and human expression that make music truly engaging and relatable to listeners. (Think about how Marvin’s Room hit you right in the soul in the 2010s. Can we trust AI to do that?)

• Risk of bias: AI is only as good as the data it is trained on. So depending on what it learns from, there’s a risk of bias towards certain genres or styles of music, limiting the diversity and creativity of the music produced.

Even with these concerns, in a short time span we’ve seen that AI-generated music has the potential to revolutionize the music industry and create new and exciting opportunities for musicians and producers. As with any new technology, it's important to weigh the benefits and risks carefully and consider how we can use AI in a way that is ethical, inclusive, and empowering for all.

Thanks for reading, and we'll see you next week! *Chad & Kendall drop mics*