Over the weekend (no pun intended), an AI voice-generated song featuring Drake and The Weeknd’s voices dropped on TikTok creating quite a stir. From our knowledge, the track released by TikToker, ‘Ghostwriter977’ was neither produced nor approved by Toronto’s Finest. Although doubtful that AI will out-rap Complex’s 2022 top rap verse winner Cardi B, it’s crucial to safeguard the art and industry. From the beginning, we’ve prioritized the protection of AI voice-generated content. We’ve introduced Resemblyzer, which has the ability to identify deep fake audio. Also, in December, we unveiled our neural speech watermarking technology to verify whether audio was produced by Resemble AI. We have additional measures on the horizon to further secure IP. With that being said, let’s investigate how this ‘ghostwriter’ created the track. Then we’ll explore a unique perspective on generative AI as it relates to music writing, production, and editing.
From Studio Gangster to Synthetic Gangster With AI Voice
First, how did Ghostwriter977 pull it off? Assuming he/she isn’t a specialized engineer it’s likely they used an AI voice cloning technology online. Voice cloning involves using a deep learning model to analyze and replicate the nuances of a human voice. By uploading audio files of Drake and The Weeknd to the voice cloning software, the AI voice model was able to train their synthetic replica voices. Next, the AI-generated content was produced through text-to-speech or speech-to-speech. Ghostwriter977 must’ve typed or spoken the lyrics into a computer mic, resulting in the synthetic voices rapping the lyrics. While the tech industry and government must establish guidelines, this misuse of generative AI may have a silver lining.
The Ghost Writer’s Block Is Hot
Similar to any creative project, music artists and writers run into the dreaded writer’s block. There isn’t always someone to bounce ideas off of, however now with generative AI there’s a potential ally in ChatGPT. Instead of going in circles and reaching for a cup of ‘Purple Drink’ (an intoxicating beverage) to stimulate ideas, writers and artists can tap into ChatGPT to break through the creative block. Below is a sample of how ChatGPT can get those juices flowing again.
One of the key advantages of the voice marketplace is the vast array of diverse female and male voices, accents, and languages that are available. Inclusivity is no longer an option but a necessity for any business or developer aiming to create products that cater to a global audience. By leveraging the voice marketplace, you gain access to an extensive collection of natural, high-quality voices from different linguistic backgrounds. This allows your applications to become more accessible and engaging for users around the world. Also, the use cases are robust for female voice over and male voice overs spanning across games, character, IVR, and more. Additionally, the marketplace provides individuals with speech impairments the ability to find an AI voice that suits their commercial voice over needs. Listen to the diverse samples below to get a feeling for the comprehensive range of AI voices.
ChatGPT spitting bars about Drake’s love for Toronto.
Also, kicking off a project brainstorming with ChatGPT is a perfect way to explore potential verse topics. Remember that you get what you pay for. ChatGPT-4 is worth the money.
ChatGPT prompted to brainstorm verse ideas.
Out Shooting, Literally
Beyond writing there are ways production can be streamlined through generative AI voice. Imagine a situation where The Weeknd is out on tour or partying with Champagne Papi in Miami. At the same time, Coca-Cola would like for him to record an audio ad in LA. Through our voice cloning and voice generative AI software, his music label, XO, would be able to give Coca-Cola access to use The Weeknd’s consented synthetic AI voice. Now Coca-Cola’s team can plug into Resemble AI’s secure, real-time API and generate the audio content using text-to-speech or speech-to-speech. Again, ChatGPT can contribute by generating compelling commercial scripts like the one below.
One of the key advantages of the voice marketplace is the vast array of diverse female and male voices, accents, and languages that are available. Inclusivity is no longer an option but a necessity for any business or developer aiming to create products that cater to a global audience. By leveraging the voice marketplace, you gain access to an extensive collection of natural, high-quality voices from different linguistic backgrounds. This allows your applications to become more accessible and engaging for users around the world. Also, the use cases are robust for female voice over and male voice overs spanning across games, character, IVR, and more. Additionally, the marketplace provides individuals with speech impairments the ability to find an AI voice that suits their commercial voice over needs. Listen to the diverse samples below to get a feeling for the comprehensive range of AI voices.
Thought-starter for a Coca Cola commercial courtesy of ChatGPT.
Edits Made Easy With Generative Voice AI
Moreover, the editing process also benefits from utilizing generative AI voice. Currently, an artist records music or a commercial in a studio. If there are any errors or changes in lyrics/script, the artist has to go back into the studio to re-record. Our Resemble Fill solution eliminates the need for any back and forth. Basically, with Fill you can replace any word or verse using the artist’s cloned voice. You just delete the word(s) in our app and replace the word through text-to-speech or speech-to-speech. More savings. Less Doing.
No Need to Empty Out Your Pockets
Along with the time related efficiencies, generative AI can provide songwriters and musical artists with monetary savings. The cost savings associated with travel and hotel as well as manpower resources required to host a celebrity artist is significantly reduced for a brand such as Coca-Cola. Additional savings from editing are accrued by tapping into Resemble Fill. Generative AI gives the Coke team the opportunity to reallocate manpower and budget towards other projects. The takeaway here is efficiency in both time and money, arguably the two most scarce business resources.
That’s A Rap
Finally, as generative AI continues to evolve at light speed, it’s here to assist in human creativity not replace it. In the case of a ghostwriter gone awol, we have to remind ourselves that music is an art and requires a sense of persona. Art can only be felt through the emotions of the artist. Ever wonder why people stare at paintings in a museum? They’re channeling the emotions and state of mind of the artist. Whether it be a conductor in an immaculately tailored suit or the glitz of ASAP Rocky’s gold caps, the performer makes the performance.
If there’s one takeaway from this post, remember that similar to drinking responsibly, please clone consensually.
