Thank you, Sebastiano. So one, I'd like to say that AI is probably one of the most transformative things that humanity as ever seen, it has so many different implications. So because of that, yes, something very close attention to it than we across the company are. Two, there are many different ways to sort for the music industry, as well as for other industries, who own copyrighted material. And it really falls into four buckets, one, which is the use of existing copyrights to train generate of AI. The second one is a sampling of existing copyrights, as the basis for new and remixed AI content, AI generating Content. The use of AI to help and support creativity. So an assisted way to do that. And then most importantly, find ways to protect the graft of artists and songwriters from being diluted or replaced by generating AI-generated content, which is what you mentioned. But it's not just that question, it's all of these together that we as an industry and I don't mean just the music industry. But overall, sort of copyright owners need to work together with the AI platforms on. And I want to make sure that everybody understands that, you don't have to be forward-looking in order to address this, which of course we are. But you can also look into today's world, one can benefit our position in the future, where there is a lot of AI-generating content. And what I mean by that is, tracking of content, identifying and tracking of content on consumption platforms that can appropriately identify copyright and remunerate copyright holders, underpins all of this. And different platforms have different capabilities in this regard. Obviously, YouTube is most advanced with Content ID, something that all the others over seen. But there are others who lacked in this department and need to work on that because in the AI future, this would be a serious deficiency. So you can -- you will see us focusing on this quite a lot. Eric?