Hello, PRO user here. I am speaking up regarding AI. I work at BD ByteDance, better known to most as TikTok, in the area of AI machine learning and Python coding. I cannot go into more detail due to contractual restrictions. Damn privacy. I agree.
A current problem with music videos is that prompts cannot be saved. In short, this means AI prompts for visuals, for example how people are supposed to look, are never stored. Sora2 in the US, which is in beta and requires a key, is capable of this. Tunee advertises with “4K Sora²”. It uses it. I know that Sora² can store visual prompts if the prompt is well written.
I think it is obvious to everyone that it is difficult to produce a music video when the artist or the respective actor looks different in every scene.
####################
The AI also needs additional machine learning in the area of execution. In plain terms: if I write lyrics and give the AI a clear command such as “All lyrics that have (malevoice) in front of the line should be sung male” and the AI cannot execute this, or uses female vocals for male lines, or swaps everything, then the IF statements or the IF ELSE logic need to be adjusted.
####################
Aside from these two points, you are on a very good path. You listen to users. Keep doing that. Continue using criticism as guidance. Very good company policy so far. Keep it up.
Thank you very much for sharing your feedback with us.
We want to let you know that we are already actively working on the issues you mentioned, and you should see concrete improvements in the near future.
Regarding your suggestion about controlling male and female vocals through conditional logic, we’re very interested in understanding this more deeply. Could you please describe in more detail how you think the execution logic or code structure should be adjusted? For example, how you would expect IF / ELSE–style conditions to be defined or interpreted when processing lyric annotations. This would help our team evaluate and refine the execution layer more accurately.
Lastly, we sincerely appreciate your feedback and the time you took to share it with us. We’ve also sent you a follow-up email requesting more detailed information—please feel free to check it when you have a moment.
Thanks again for your support and valuable insights.
Hi,since I don’t know your code and don’t have any further information, it’s difficult to make a general statement. The simplest option would be something like this.(This is only meant as an example, please don’t just insert it into your code, because, as I said, I don’t know the entire code that deals with the recognition of female and male vocals and, above all, I don’t know how you try to “force” it to be executed in this way).
