Thanks Lloyd. Of course, we've been working on AI and ML for a long time in this company, when generative AI became something that everybody was talking about. I was very clear that we have urgency and focus on it and we think about it in three ways. One is what we can do to really drive activity in our product. And you've seen us do that with our itinerary builder, and really leveraging our data, the proprietary first party data that we have. Clickstream data, behavioral data to really understand users, and then to deliver an itinerary that is highly personalized, and with the right kind of prompt engineering to deliver a good experience. And that product, of course, is now delivering a really good number of generative AI itineraries. It's double-digit percentage of total itineraries being created and growing. And we're seeing that those users are returning at a much higher rate, as I mentioned, and this is without significant product marketing or further enhancements of the product. So now we look to enhance that product and really drive monetization and further integration coming out of it going forward. We also have been doing a lot with our content, right. We have this incredible vast billion reviews that are from real people. And we have gotten -- we've been improving the product so that we continue to have [indiscernible] and relevance, which gives us an incredible content dataset to be to be leveraging. And, of course, our audience and brand of trust. So we think generative AI around our product is going to continue to be meaningful and we'll look to leverage it across all of our categories, hotels, experiences, and restaurants. We will also leverage generative AI for productivity and we're exploring lots of use cases across how we think about trust and safety and what we're doing to make sure we have the highest quality reviews. Translation and localization as I mentioned, before customer service, engineering productivity, marketing productivity, we can imagine a use case for every functional area and we're going to be leaning in for the long-term. And then finally, we want to be very thoughtful about how we partner with the best of class LLMs and the large tech platforms, as we think about being a part of the generative AI ecosystem, whether we choose to be a direct partner or frankly, select the ones that we feel we're getting the most value from, and choose not to work with and therefore block the others that don't. So we think we've got a good strategy for the long-term and this is a long term proposition. It's very early days but we're excited about where we are.