TikTok’s testing out another generative AI element, with some users now being prompted to use a new “AI Song” option, which, as it sounds, is able to generate original music, based on text prompts, to accompany your clips.
As you can see in these example screens, posted by app researcher Jonah Manzano, TikTok’s new “AI Song” option provides a basic text command process to generate original music in the app.
And the results are… well, I guess the results are relative to your skills in prompt writing:
@imdanielduncan You can now create AI songs directly in the TikTok app, and here’s how! #tiktokupdate #ai #aisong #tiktokai #tech #fyp ♬ original sound - Daniel Duncan ????
Yeah, probably not a great example, but you get the idea.
As explained by TikTok:
“AI Song is an experimental feature on TikTok that uses AI to create songs based on prompts you enter.”
Pretty straightforward.
TikTok further notes that the lyrics generated via the tool are powered by Bloom, which touts itself as the world’s largest, open multilingual language model. The Bloom dataset contains over 176 billion parameters, and has the capacity to generate text in 46 languages.
So there should be a broad range of lyrical options, which should mean that you can produce better results than: “Fulfilling my desires/iPhone eleven glide”.
Though that song is catchy, maybe there is something to it.
The option actually looks to be an expansion of TikTok owner ByteDance’s previous AI music project, a separate app called Ripple, which it launched in selected markets last year.
Ripple enables users to sing or hum a melody into the app, then uses machine learning to build out a track around that initial part.
It seems that ByteDance is using the same music creation backbone to power this new tool, which could help more TikTok creators put together random viral hits, or at the least, unique music that they can then use to accompany their videos.
And if it’s expanded to business profiles, it could also provide a copyright-free way to add soundtracks to your clips.
Generative AI has the potential to completely upend the stock content industry, by providing unique, copyright-free content that you can freely use, however you please, and many websites are already using AI-generated images. And with tools like this incoming, it seems that stock audio will be another aspect set to be disrupted, though it does depend on just how good the outputs are via the tool.
But even if they’re not great, it points to another area of AI development, which will become a bigger consideration over time.
There’ll be various copyright battles to come, especially as more AI content is produced that looks or sounds similar to established artists.
But for now, at least, the potential here is significant.
TikTok's AI Song option is available to selected users in the initial test pool at present.
